JP2018198039A - Contactless key matrix controller - Google Patents

Abstract

To provide a contactless key matrix controller that is an adhesion interface type/induction type of concurrent processing, which abolishes key scan that causes problems, such as time delay in response and burden of battery driving due to constant current consumption, common to a conductive rubber contact of a TV remote controller, a membrane contact-pantograph mechanism of a PC keyboard, and a keyboard of a non-contacted capacitance switch.SOLUTION: Instead of performing key scan, parallel detection of the pulse waveform that propagates through X and Y lines is performed when spontaneously generated/power-generated at the intersection of a matrix, so that the contact life becomes semipermanent. Spontaneous pulses due to adhesion or separation of charged parallel plate electrodes, or power generation pulses by magnets passing through coils, are processed concurrently by a key matrix.SELECTED DRAWING: Figure 22

Description

The present invention is an operation panel and a calculator that must be mounted and connected to a wall remote from the product body such as a remote control for a TV, a controller for a game machine, a PC keyboard, an air conditioner, etc. using infrared rays and radio waves, and must be operated reliably. Key matrix scan, which is shown in Fig. 1 as a whole, is used exclusively for the operation panel of home appliances, fixed button / cell phone, etc. Or which button is pressed so as to be coupled by capacitance. Above all, in remote controllers / game machine controllers / general operation panels, the X and Y electrodes are short-circuited by a floating Z electrode, so the two contacts need to be connected at the same time. There is a higher probability of end-to-end conduction failure. Design intention to maintain many conduction points by absorbing irregularities on the X / Y electrode surface, such as by using flexible conductive rubber for floating Z electrode so that two contacts are conducted at the same time. Although there is a design intent to avoid the bonding capacity just by approaching with a stronger force of contact, the rubber will be deformed over time, or carbon filler is mixed for conductivity The conductivity of the conductive rubber material itself is lost due to the loss of continuous contact of the carbon black inside, or conduction is lost due to adhesion of an oil film or the like. In addition to this, a leakage current that cannot switch off the contacts may occur. These various requirements overlap, and it can be said that they are destined to break down if overuse, but they are not accepted by society, so it can be said that every effort has been made to resolve contact failures so far. Again, fundamentally permanent contact life is required. In addition to this, the battery-driven remote control is conspicuously hidden from the problem that the AAA dry battery runs out in more than a year due to current consumption such as key scanning that always flows even when it is not operated at all. This current consumption can be reduced by switching from sequential key scanning to parallel processing, but it can be said that the technical system has been progressing in one direction without any doubt at first due to the key scanning.

The key matrix controller consisting of multiple buttons has two X- and Y-lines at the same time by attaching a soft-landing conductive rubber to the back of the button part that is made of spring with the integral rubber sheet shown in Fig. 2 (a). Whether the key matrix scan that conducts by contact is used for a remote control, etc., or the membrane type / pantograph type shown in FIGS. 2B / C is used for a PC keyboard as a membrane 1 contact point method that can be expected to be more durable Fig. 2 (f) is a historical key matrix scan that turns on / off a one-contact point type reed switch in which a magnetic substance and metal contacts are sealed in a glass tube with an inert gas. It was a mold.
The capacitance switch type shown in FIG. 2 (e) is a relatively new method compared to the conduction type, and is provided with an X electrode at the driving end and a Y electrode at the receiving end on the substrate, and a conical metal spring on the top. The outermost peripheral portion of the electrode is AC-grounded to the X electrode, and the counter electrode area increases or decreases due to the expansion and contraction of the spring. When the button is touched with the fingertip, the charging of the fingertip and the disturbance via the fingertip are transmitted to the conical spring as the floating Z electrode. If the Z electrode is AC grounded to the low impedance X electrode, Although it can be said that the Y electrode is always locally shielded from the top, constant voltage driving of the X electrode cannot be performed for the following reason. In addition, since the X line and the Y line drawn for a long time pick up a disturbance, a scan pulse amplitude (for example, 3 V to 5 Vpp) that overcomes the disturbance is given. The conduction type can perform constant voltage drive, and the amplitude of the scan pulse is directly transmitted to the receiving end. However, in the capacitance switch type, the amplitude of the CR differential waveform transmitted to the receiving end does not change depending on the coupling capacitance. Discrimination must be made possible by increasing the impedance of the signal so that the coupling capacitance changes the magnitude of the transmitted pulse. For this reason, the amplitude transmitted to the receiving end is reduced, and this is a system with many problems to be overcome, in which sufficient attention must be paid to disturbance. The drive may be a sine wave burst instead of a rectangular pulse.
In any of these methods, as the key scan, the X line is driven by an inverter in the IC chip with an output impedance of ˜several KΩ, and the resistive / capacitive impedance of the switch is inserted in series, and the bias resistance is 1 KΩ to 100 KΩ. The operation is performed by slicing the waveform transmitted to the Y line that is terminated at the end. The bottom of the PC keyboard is often laid with a weight-stabilized iron plate, which serves as an electrostatic shield that is grounded to the PC body via a USB cable and a magnetic shield that utilizes the magnetic property of the iron plate. ing.
2 (f) / (g) / (2) gives new possibilities to these common systems that perform key matrix scanning in the form of a crossbar switch that conducts the X and Y lines at their intersections. h) is a format showing the basic concept, in which voltage pulses appear independently on each of the X line / Y line at the intersection of the matrix, and N + M pieces of information transmitted to the IC chip are calculated in parallel. Based system. In short, the permanent magnet moves steeply and the coil generates power to generate a voltage pulse of several mVpp, or the parallel plate electrodes charged in advance to a constant potential difference through a bias resistor are in close contact with each other, or are in close contact The positive and negative pulses of several hundred mVpp appearing depending on the distance from the above state are transmitted to the IC chip. These are called parallel processing for conventional key scans, but they can be said to be hidden means that have been excluded from the beginning by the fixed idea that key scans are inevitable.

The remote control for TV, which is often driven by AAA batteries with a capacity of about 900 mAh, consumes a current of about 80 μA at all times due to the key matrix scan type scan operation. The problem is that the remote control in the place that is visited once a year is always near the end of the battery. Even PC keyboards have the same problem with wireless driving, but in most cases, power is supplied via USB wired connection, so there is no particular problem. The current remote control is universally a key scan type, but the current consumed at all times is reduced from about 80μA to about 30μA just by adopting parallel processing, and the battery life due to the current consumed constantly with the battery in the battery is naturally discharged. It can be close to about 3 years determined by With infrared and radio remote controls, if the button is pressed frequently, the ratio of infrared and radio transmission power consumption will increase, but if it is infrequent, it will not necessarily be recognized by the designer that the battery will run out due to the power consumption of key scanning. It has not been prevalent, and it has been left without being seriously addressed.
Furthermore, since the remote control for TV first presses the power button and then other buttons are pressed, in parallel processing, the bias current flows only to the comparator on the IC chip connected to the X line and Y line across the power button of the key matrix. Others can be turned off so that the current drawn from the remote control batteries can be ignored when the remote control is not in use.

One of the major nuisances of performing a key matrix scan is to cause internal electromagnetic interference to the circuit of the device in which it is incorporated, or to electromagnetic interference to peripheral devices. A shield is required. From this point of view, if it is a spontaneous pulse system that does not perform scanning, an electromagnetic shield that does not provide scanning is unnecessary. On the other hand, in order not to malfunction due to electromagnetic interference from the outside, an electromagnetic shield is required both in the scan format and in the generation / power generation pulse format. In particular, in the parallel processing generation / power generation pulse method, the electromagnetic shield becomes more severe because the pulse is small. When these two aspects are considered comprehensively, it is clear from the following description that the spontaneous / power generation pulse has the possibility of expanding the degree of design freedom.

In the key scan method, a stroke of 1 mm to 2 mm for pressing the button is performed at about 100 mS and reaches the bottom (bottoming), and the intersection of the key matrix is turned on whether it is a conduction type or a capacitance switch type, as shown in FIG. As described above, the rectangular scan pulse or its CR differential waveform is transmitted from the X line to the Y line. Sequential processing key scan results in a time delay and uncertainty due to sampling with a scan cycle of about 10 mS, but parallel processing using voltage pulses generated and generated independently without scanning is realized. In this case, the problem of time delay and inaccuracy due to the scan pulse period does not exist. If something that moves on the screen has to be pressed at a certain point, this time delay and uncertainty is a problem, but it is essentially unavoidable as it is unavoidable with key scanning. No attempt has been made to reduce the error by shortening the scan period as necessary.
In addition, as a method of embedding an analog amount that can be said to be halfway until now, in the scan method, the speed of pressing or hitting the button, or the pressed depth, is the size of the pulse transmitted from the X line to the Y line. In principle, the conduction type cannot be reflected, and the capacitance switch type, as shown in FIG. 3, determines the size of the transmission pulse by the depth at which the finger is pushed down. After passing through the dome bending point, it was difficult to control the position and scan pulses arrive irregularly, making it difficult to embed analog quantities precisely in the transmission pulse. On the other hand, in the parallel processing, the analog amount is reflected in the height of the spontaneous pulse without any distortion from the beginning, and no new preparation or precondition is required. Since the analog amount is already embedded, a finer operation that is convenient for the controller of the game machine can be performed. In addition, the PC keyboard has a new possibility that the optimization of key touch that reduces malfunction and provides lightness is promoted by a learning function that compensates for individual differences.

The conduction type remote control has a three-pole structure of X electrode / Y electrode / movable Z electrode, and the output impedance r of the X line inverter drive is set so as not to cause malfunction due to disturbance even when the Z electrode is in a floating state. Consideration is made to reduce the termination resistance R of the Y line together. This low impedance is one of the factors that increase the scan current consumption. In the conduction type, the electrodes are close to each other immediately before conduction and immediately after separation, and have the same capacitive coupling form as the capacitive switch. Therefore, the differential waveform of the rectangular scan pulse is transmitted to the Y line, which may cause malfunction. is there. To avoid this, the impedance must be reduced so that capacitive coupling and conduction can be distinguished. Also, the larger the floating capacitance of the Y line, the fewer malfunctions. In order to help this, the counter electrode area is made relatively small, so there is a vicious circle that the possibility of poor conduction increases accordingly.
The capacitance switch type shown in FIG. 4 has a tripolar structure of X electrode / Y electrode / floating Z electrode, and a fixed capacitance between the X electrode and the Z electrode and a variable static capacitance between the Z electrode and the Y electrode. A series value of capacitance is inserted as a switch. In order to prevent that the X electrode and the Y electrode are directly coupled without going through the Z electrode and the capacitance on / off ratio cannot be obtained, a guard ring grounded is provided between the X electrode and the Y electrode. Since it cannot be said that the electrostatic capacitance switch type has an electromagnetic shield, the impedance is reduced as much as possible, or mixing of disturbance is suppressed by interleaving that reverses the polarity.
As shown in FIG. 5, the conical metal spring is an opposing capacitance between a surface and several spiral spring wires. Compared with the opposing capacitance between the surface and the surface, the capacitance when separated is as follows. Although there is no big difference, since the capacitance when approaching is smaller, the variable width of the variable capacitor, that is, the ratio of the maximum capacity / minimum capacity cannot be increased. The basic problem that the size of the differential pulse transmitted to the Y line from the rectangular scan pulse that drives the X line does not change depending on the variable capacitance is that the Y line has a stray capacitance and the driving impedance r of the X line. It is not until zero becomes a non-zero magnitude that the transmission pulse changes according to the variable capacitance and its change ratio. Similarly, FIG. 5 shows an equivalent circuit of the capacitance switch.
Equation 1 is a transfer function from the X line to the Y line, but the change in the magnitude of the transmitted pulse cannot be known unless the step response is obtained instead of the frequency response.
Capacitance switch type that combines conical metal spring and rubber dome is not used for remote control, and it is used only partly for PC keyboard, but it can be said that reliability is already high in membrane type and pantograph type. However, the combination of conical metal spring and rubber dome is not different from other methods in the end, but the market price is 8 times that of other standard products. Only 10 times is differentiated. However, the reason why it is sometimes used for ATM 10-key input, which requires a high level of reliability, is to make a selection that is less likely to cause contact failure than the membrane type or pantograph type.

Almost all current remote controllers / keyboards are key matrix scans as shown in FIG. 1, and their shapes are low-impedance driven N-line X-lines and low-input impedance terminations in order to suppress disturbances. If the M lines of Y lines are orthogonal, N * M intersections are created, and at one of them, the X and Y lines are exposed, and the X and Y electrodes are directly shorted. Or if a short circuit is mediated by the floating Z electrode of the conductor, only one of the combinations of the N * M matrix of the X line and the Y line becomes conductive, and the transfer function from the X line to the Y line becomes The common principle is that the remaining N * M-1 combinations have the selectivity of not being so. This is the idea that scan pulses are applied to one X line and not applied to the other. An alternative in terms of reliability / manufacturing cost / comfort is to actively generate a pulse at the intersection of the X and Y lines and transmit it in parallel to both the X and Y lines. This is a new idea, and if this is by far the most advantageous in some respects, including reliability, it could replace key matrix scanning in all areas. In order to examine the potential, it is necessary to understand the difference from the key matrix scan in a more fundamental principle.
The deeper advantage of parallel processing is that key matrix scanning collects on / off information focusing on only one intersection, but in parallel processing, all information from N * M intersections is stored in N + M X / Y is superimposed on the Y line. If this information is used as a reference without truncation, there is a possibility that discrimination with higher accuracy can be performed.
It's no exaggeration to say that the continuity type rubber contact remote controller and the rubber contact game machine controller will always suffer from poor contact, but until now there has been no way to replace the continuity key scan. The reliability of the PC keyboard is higher than that of the remote control because it is not a rubber contact. The contact is surrounded and protected by a rubber dome. This is because it is protected from being pressed and cannot be pressed as hard as the buttons on the remote control. PC keyboards also break down if they are unlucky, but they are standard, so they are inexpensive and can be said to be not a big problem if you can easily replace them with new ones. On the other hand, TV remote controllers often break down with a higher probability and are often replaced with new ones rather than ordering new remote controllers at prohibitive prices.
Taking the current PC keyboard as an example, the three types of conduction type pantograph type / membrane type / mechanical type shown in Table 1 have been established, but the deterioration of contact conduction over time has lowered the reliability. . The relatively new 4th capacitance switch type can be said to be contactless, so it is expected to be more reliable, but due to the high price, the spread is only a small percentage, and those who are professors who continue to hit the keyboard There are only things that can be bought. Quantitatively speaking, most general external keyboards are of the membrane type, and the pantograph type is often used for notebook PCs that require a thin keyboard. Until now, it can be said that there is no format that satisfies all of low manufacturing costs, ultimate reliability, lightness and certainty of key touch, and thinning. It can be said that the requirements of the controller of the game machine are the same as those of the PC keyboard.

JP 1997-204846 JP2009-302690A

For example, if you live next to the school, you can pass through the school gate as soon as you leave the house, but if you use a 10-minute route bus in a congested environment, you may have an irregular bus waiting time of up to 10 minutes. If you have a full bus, you have to wait 10 minutes. In the key scan that would be the same situation, N * M pins are not necessary for the IC chip, and it has been taught in school that N + M pins are enough, but the scan fate waits for a scan pulse of 10 mS ~ Produce. What I couldn't tell was the misunderstanding that N * M pins and wires are required for parallel processing without scanning. In fact, N + M pins and wires are used in the same way, and there is no waiting time. In key scan, the current consumption increases not only due to the scan current but also due to the suppression of disturbances, but the current consumption of parallel processing without scanning is smaller, and it has been believed that a fixed amount of constant scan is inevitable so far. Therefore, it can be said that none of the IC chip designers had the opportunity to make steady efforts to save power in other parts. If the wasteful current in the other parts is reduced, the current consumption is dramatically reduced by stopping the scan.
To stop the key scan, which is clearly difficult in the key matrix, is the voltage pulse generated by striking the piezo element with the key, or is the voltage pulse generated by changing the magnetic flux of the permanent magnet linked to the coil? As long as the distance between the charged parallel plate electrodes changes, a transient change pulse in which the potential difference obtained by multiplying the constant electric field by the distance changes and rebalances can be used while the charge distribution does not change in a short time. These changes are given to both the X line and the Y line from the intersection with the button almost in parallel and transmitted to the pins of the IC chip.
In the case of the key matrix scan, as shown in FIG. 3, the button is kept pressed for a time longer than the cycle in which the pulses come, and a redundancy is provided during which several pulses are transmitted, so that chattering and the like are not confused. In this way, it is determined that two or more on-pulses have arrived, and the button is kept pressed. Instead of removing this redundant restriction if it is replaced by parallel processing, it is necessary to make sure that access is performed with a single operation of hitting a key, whether it is an inter-electrode distance change pulse or an electromagnetic induction power generation pulse. It must be accompanied by severe disturbance suppression, but at the same time, it is possible to do this at around 150mS with no significant individual difference in the round-trip time of TAP strokes where a human strikes a key with 10 fingers. It is an understanding of the mechanism of the living body. In addition, it is important to understand that if the generated / generated pulse is much smaller than the key scan pulse, the problem can be solved by passing the amplifier if there is a signal / noise ratio.

As another method for reducing standby current that is always consumed by not performing key scan while being conductive, the X line, which is a voltage output, is multi-valued between 0V and 3V as shown in FIG. There is a form in which a threshold window is provided to discriminate from the fact that it is divided and placed and transmitted to the Y line through the conductive contact. If there are N three-input comparators on the Y line on the detection side through the conductive switch, it can be determined from which X line the voltage comes. Since the number of comparators on the IC chip connected to the Y line is N * M, and the outputs are only combined and decoded, the number of circuit elements is reduced as compared with the key scan, and the current consumption during standby is reduced. In addition, since the sequential processing of scanning becomes parallel processing, immediate detection without a pulse waiting time becomes possible. In this method, when the number of X lines is about N = 8 to 10, the resolution for identifying the voltage of which X line is selected is not so problematic. Even when the voltage of two AAA batteries decreases from 3V to 2V with a remote controller or the like, the resolution can be prevented from being impaired by a device. However, simultaneous pressing of two buttons on the same Y line cannot be identified. In any case, if the conducting button becomes non-conducting due to the formation of an insulation film, the point that the button does not work is not solved by multi-valued, but the battery is long-lasting and improved in immediacy etc. Just do.

As shown in FIG. 7, the coil connected to the X line formed on the printed circuit board and the coil connected to the Y line are concentric spirals, and the metal plate is attached from the top in conjunction with pressing the button. The inductance of each X / Y coil is reduced or the magnetic coupling coefficient between the X / Y coils is relatively reduced depending on whether the metal cylinder passes through a hole formed in the central portion of the concentric coil. A key matrix that uses things can be configured, and contactlessness can still be realized. This method detects a voltage dip that occurs when the distance between the parallel plates is brought close to the contact, or a method of detecting a power generation pulse in which a voltage is induced by bringing a magnet close to the coil, or a rectangular wave or sine wave burst. It is necessary to scan with. In conclusion, this method cannot be said to be superior in terms of effect / cost compared to the non-scanning type in which power is generated by a permanent magnet shown in FIGS. 2 (g) / (h). In this way, it is highlighted that the problem is solved by switching to parallel processing that eliminates key scanning.

Compared to a membrane contact in the form of a single contact point where the X and Y electrodes of about 3 square mm face each other and the two conductive plates are in contact with each other, the conductive material mixed with carbon directly below the button of the integrally formed elastic rubber sheet The floating Z electrode with a conductive rubber is connected between two X- and Y-electrodes with the same contact area at two contact points. Just malfunction. The essential reason for making this one contact point between the upper X electrode and the lower Y electrode is that contact failure is unlikely to occur. At the two contact points, the Z electrode is parallel to the surface position where the X electrode and the Y electrode are aligned. This is because if one is not approached, one side does not reach conduction and may end in capacitive coupling. The conductive rubber contact is flexible to deform and has the meaning of absorbing the irregularities on the surface where the X and Y electrodes are arranged. However, when the conductive rubber is deformed over time by repeated pressure bonding, there is a possibility that the conductivity is deteriorated as a material.
The two-contact point type mediated by the Z electrode is abandoned, and the one-contact point type membrane contact that connects the lower Y electrode and the upper X electrode in close contact with each other is generally used in PC keyboards and is not necessarily reliable. It can not be said. On the other hand, even if the same thing is done with the TV remote controller, the upper X line cannot be formed so as to be individually wired on the integrally formed rubber sheet. If conductivity is given to the entire integrally molded rubber sheet, the X line must be in one line and the Y line must be in the M * N column, and the pulse driven X line will be exposed to the outside. . In the case where the integrally formed rubber sheet is made conductive, there is only a possibility that it is grounded, but this means that the key scan is abolished.
If the upper crimp electrode plate is made of metal, a metal-to-metal problem occurs with the lower copper foil. Even if the non-metallic conductive rubber is deformed and adheres closely to the uneven surface of the lower portion to make the contactable area as large as possible, it is inevitable that the non-conductive rubber becomes non-conductive due to the formation of an oil film or the like. The unevenness of the X and Y electrode heights in the X / Y / Z three-electrode type maintains the parallelism with a pantograph or absorbs it with the flexibility of an integrally molded rubber sheet, resulting in close contact with both contacts. It becomes a complicated process.

In designing a PC keyboard correctly, you must understand the fundamental characteristics of the living body. The person is conscious of trying to move the fingertip lightly, and the finger bends together from the base and the fingertip descends about 10mm quickly. Following this movement, the fingertip returns to its original position. Get up and complete. Moreover, since this reciprocating motion is continuously performed even if the consciousness is separated, it can be said that the main consciousness has shifted to the selection of which button to drop the finger on the keyboard typing. Since the time required for the reciprocating motion is about 150 mS, and the return is a little faster and the return is a little slower, it may be assumed that the reciprocating time is about 70 mS + 80 mS. If this reciprocating motion is continued, it can be performed 6 to 7 times per second, and the natural shape of the dominant hand is that the amplitude at that time is about 10 mm. If you are not a dominant hand, the round trip time may be slightly slower. If you hit the key in this way, the number of reciprocations will decrease if the barrier until the rubber dome is bent is large, and if the stroke stops after about 1mm instead of about 3mm, Your fingertips will not come back naturally and the number of repetitions will be even smaller. Therefore, from the point of view of wanting to hit at high speed, the stroke to the lowest point that does not reach the bottom should be as natural as possible, the distance to the bending point of the rubber dome should be as short as possible, and the drag should be as small as possible. It is done. The only thing the current keyboard has to set apart from these conditions is that the membrane contact common to the membrane / pantograph type crosses the barrier, opposes the drag, and the protrusion in the rubber dome with excessive force This is because it does not conduct unless the film is struck, and in notebook PCs, it is desired to make the thickness when the liquid crystal panel is closed as small as possible compared with competitors, so only an unnatural short stroke can be taken.
Despite being conscious only of bending the finger, the return movement is always accompanied by the forward muscles that normally strike the fingertips downward and shrink in the opposite direction with little use on the opposite side The muscle is in a relaxed state, and this muscle is the same as an elastic rubber band. When stretched, it is considered that the muscle is in a relaxed state and then contracted. Even if the consciousness is only to bend to strike the key with the forward muscle, it naturally returns to the home position. The energy consumed when the forward muscles contract is basically used for the kinetic energy that depresses the key and the energy that extends the hook spring, but the extension energy is absorbed by the reverse muscle in the relaxed state and becomes a fever. . In other words, the energy of the smallest unit is used in a single movement, but the rubber string that stretches and contracts passively in a relaxed state generates heat even if it does not consume energy.
Another thing to pay attention to is that you can not hit with a fingertip on a hard surface as it is, so you are hitting with one reciprocation that lifts the fingertip and then swings it down. In this case, the role of the contracting muscles and the muscles contracting by the passively pulled reaction are interchanged with those of the TAP. That is, when we place a finger on the key of ASDF / JKL: at the home position and hit one of them, it is usually pushed down as it is, and it is rare to swing it down after lifting. These two are clearly separated as unconscious movements and cannot be in between. Keys other than the home position can be said to swing up. Considering this, the current threshold value of the drag of 50 to 60 g is too large to make sure that the key of the home position is always pushed down as it is. If this is, for example, an initial drag of ˜20 g, it will definitely push down directly from the beginning.
From the observation of a general nerve pulse, it is likely that a single nerve pulse is caused by a natural and smooth movement of the smallest unit of a finger. A short 1mm push-down exercise that comes back home, staying stationary in the depressed position, a slowly moving exercise, and a consciously smooth exercise all involve extra concentration and muscle energy consumption. The brain circuit on the commanding side also seems to consume a lot of electrical energy due to frequent nerve pulses.
Only humans can skillfully walk on two legs, and can move horizontally and move horizontally with low energy consumption in animals. It is reconsidered that he has been able to support his heavy head by using self-sustaining. In fact, when making a doll that reciprocates like that, there is a surprising simulation result that the doll continues to walk forward for 10 steps without power. If the muscles on both sides continue to reciprocate passively in the state of rubber bands, if the nerve pulse synchronizes and contracts the muscles to compensate for one loss, the energy consumption is minimal. Yes, the round trip speed is the fastest. It seems to be said that the PC keyboard should aim for this natural state.
In this way, the unique reaction of the finger to become a self-supporting reciprocating movement is inherent in the person, not learned by habits and training, and there are few individual differences. The handedness you always use has a shorter self-supporting round-trip time, but with the new keyboard, if the other hand is also released from bottoming out, it will keep up with the speed of the handedness's round-trip time. it is conceivable that. The key scan method is a system in which the button is pressed down and held for a while, and several scan pulses go too long in the meantime. Therefore, it is better to complete it as a simple one-way exercise. Unlike this, the reciprocating movement that is lightly touched and retracted, rather than the reverse muscle that is extended by the forward muscle working in the minimum movement unit, only returns to its original shape It can be thought of as a simple passive reaction before the process of evolution, for example, to shrink by itself. Without recognizing this fundamental principle, we were not on the starting line to pursue the ultimate key touch.
With the exception of some keys, the current keyboard is designed to allow continuous character input when held down. This is obviously uncomfortable, and it is better to input by detecting the transition to the bottoming operation or the operation of returning from the bottoming to the fixed position. FIG. 8 is a conceptual diagram for that purpose. The current keyboard assumes that the movement of the parabola is convex downward in the relationship between time and stroke position, and that the membrane contact is conducted at the lowest point where it stops. At that time, it seems that there is a misunderstanding that if you push your finger strongly or fast, the time to reach the conduction contact will be faster, but in reality it is about 70mS, which is the forward time when tapping lightly. Is the fastest, and the rest of the consciousness increases the time to reach the students. All current keyboards are in key scan format, so when you reach the lowest point, you can't input unless the scan pulse circulates, and wait for the arrival of two or more pulses to overcome chattering. It needs to stay at the lower point for a while. On the other hand, the stroke distance until the fingertip exceeds the judgment threshold is short, the malfunction of the fingertip meat compression is smaller than the initial drag, avoiding malfunction, the ON judgment exceeding the threshold is made immediately, and chattering is in principle Subsequent sinking of 1 to 3 mm that does not occur exists for comfort that does not fatigue the operator, and there is a method in which the repetition speed of hitting the key is the same as that of the present or faster.
Unfortunately, it is hard to say that all key touch designs are made with a basic understanding of this. However, in the process of pursuing as light a key touch as possible, it is well known through experience that the most important factor of comfort is that the stroke is turned on when it reaches the bottom, which is the membrane type / pantograph type. It was not possible to achieve it, and it was possible to approach the place where it should go 10 steps with the capacitance switch type, but it was a fire, and this time the membrane type was comfortable by using a metal spring etc. In terms of key touch, we have taken one step forward, so it can be said that we have overtaken the capacitive switch type by half a step. But somebody has to pack 9 steps left. The ultimate key touch, however, is not 8-10 times more expensive, but in the pursuit of the ultimate cost reduction, the best form may exist with ultimate reliability.

While the PC keyboard is struck with 10 fingers, the TV remote control lifts the remote control with the palm of the left hand and moves it to the right hand. The habit of pushing is familiar, and the stroke is about 1 mm. There is no option to push lightly because of the high drag. There is no process of returning the pressed thumb, and if you release it, the finger will be released naturally together with being pushed back to the button. On the other hand, in the case of a calculator, you can hold it with your left hand and use your fingertip of your right index finger to lightly push a small, dense button, but bend the base of your finger and move your entire arm to select the button. It is a quick movement using care so as not to make a mistake in pressing, and the fingertip is accompanied by a self-reciprocating movement away from the button without being conscious.
If you slowly bend your fingertips over a few seconds, you will not be able to move smoothly, and you will end up with a discontinuous movement. This is considered to be unnatural because the combination of the command to move in the forward direction / the movement of the muscle and the combination of the command / the movement of the muscle that tries to push it backwards cancel each other out. . Using two contradicting muscles is wasteful energy consumption. If the smallest unit that moves about 10mm once is 1, use 1 to advance 1mm and use 1 to stop it. If it continues 10 times and advances 10 mm, it will take 20 steps to travel, and more energy and concentration will be required to maintain that position, and energy will be used again on the way home.
If you are conscious of moving your fingertip slowly, it is not difficult to bend your finger and move it about 40 mm, even if you need consciousness, and it is much more difficult to move slowly about 10 mm. The most natural and fatigue-free movement is that the reciprocating motion of about 10 mm inherent in human beings is performed in an isochronous time of about 150 mS, which is a single-shot process that is neither fast nor slow. It can be said that it will be more comfortable to be able to confirm by listening to the uninterrupted electronic sound.
This dream-like rational movement that the PC keyboard should respond with respect is a different movement from pressing the buttons on the current remote control slowly and strongly. Considering the ideal of the remote control button again, it does not operate when the lowest point is reached as in the current PC keyboard, but it is the original ideal of the PC keyboard in that it should be activated immediately after the push-down starts. It can be said that it is the same. Even after switching to this ideal, there will still be some ambitions that have been acquired in the history of remote control televisions over the past 40 years and that will push down the button strongly to make the switch somehow conductive. If the lowest point is no longer connected with the on / off of the switch, it will not be pushed hard, and the edge will soon be cut off from the fear of deterioration.
An alternative to turning the contact on for the first time at the lowest point is that there is a way to turn the contact from on to off near the starting point of the push down, but only one button is open and the remaining N * M-1 It is not easy to find that all the buttons are connected by key scanning. To do that, you must first eliminate the key scan. The necessity of using the spontaneous pulse generated near the starting point of the push-down or the returning end point is seen instead of the lowest point.
Currently, there is a side that strengthens the drag of the rubber contact button so that it will not be accidentally pressed by touching any of the buttons when holding the remote control, instead the stroke is short. When a new start point or end point is detected instead of the lowest point of the stroke, the threshold value of the drag is at the start point, so that the possibility of erroneous operation by touching any of the buttons is reduced. At present, it is thought that both the PC keyboard and the remote controller will accidentally press the button lightly. However, it is the initial drag that determines it, regardless of where it is sliced during the stroke process. It is the current state that associates this.
The conduction type / capacitance switch type is a switch that depends on the position of the electrode, but the contact interface type / induction type does not respond if the button is pressed down, but only changes quickly. Depends on. The concept of the single finger movement required for a PC keyboard is expressed well in the English word TAP, but there is no term in Japanese that corresponds to such movement, and the light movement of a single pulse. It can be said that this idea is missing. On the contrary, the slow and large movements seen in Tai Chi focus on the value of penance that seeks to approach smooth movements with a mental concentration that controls the difficulty that positive and negative movements conflict, but the keyboard is rather light Therefore, smooth TAP is suitable. The composition ratio of slow muscles and fast muscles varies greatly among the races of the world, and TAP is an exercise of fast muscles only. Absent.

The movement of the fingertip is composed of a response of a nerve pulse command and muscle contraction. As shown in FIG. 9, both the transmission of nerve pulses and the reaction of muscles are performed by using one sugar molecule unit (H—C—O = H), which is considered to be mainly used for fast muscles, as oxygen. It is considered that electromagnetic energy is used when one molecule (OO) is sandwiched to generate one implosion water molecule (HO = H), and (O-CO) ) The combustion energy of one molecule is smaller than that of water molecule (H-O = H) generation, and there is a fact that is clear from the calorie calculation that it is only a by-product.
When this is written as a structural chemical formula representing the bonding of atoms, Formula 2 is obtained.
In contrast to the fast muscle, the slow muscle is a fatty group in which one oxygen atom (O) is extracted from a sugar molecular unit (H-C-O = H) and the molecular weight is reduced from 30 to 14 and stored. It can be understood that what has become H—C—H) is used as a complementary, but as shown in Formula 3, one oxygen atom (O) is added to the fatty group to form a sugar molecule unit (H—C—O). = H), the carbon atom (C) is sandwiched between oxygen molecules (O—O) and terminated to water molecules (H—O = H) through a process of returning to the oxygen molecule (HO). Although it requires more oxygen, it can be understood that it has a feature that the sustained energy per body weight is large because O is extracted for storage.
It is well-known that key touches mainly use instantaneous sugar molecule units instead of persistent fat units. Nerve pulses are generated by one reversible replacement of Equation 2 common to animals and plants, and are transmitted to the muscles through the nerve cords. It is considered that the core part of ATP is accompanied by a plurality of terminations to the molecule (H—O═H). The muscles of the trunk are controlled by a number of nerve pulses, so the noise / signal ratio of pulse density variation / number of pulses is small and smooth movement is possible, but if you move only your fingertips, one to several nerve pulses In addition, the frequency of whether or not there is, and the proportion of uncertainty as quantization noise increases, so only jerky movements can be made. It is considered that the TAP of the fingertip is controlled by one nerve pulse, and if only the timing of its generation is controlled by the brain, the round trip time of about 150 ms and the stroke of 10 mm are determined as the minimum unit. Can be interpreted as a smooth movement. The use of muscles close to the trunk is the rule of thumb for accurate and smooth control in sports, but this is simply because the quantization noise ratio of nerve pulses is smaller in the trunk. It can be said that there is.
Combustion of coal, which is a plant fossil, is considered to be a change in which carbon (C2), which is the main component, is oxidized to (CO2). Combustion of petroleum containing hydrogen in animal fossils is more explosive because the amount of heat generated by the implosion generation of the main product (H-O = H) is higher than the generation of by-products (CO2). It is not easy to burn 100% carbon material, it can only be heated and oxidized from the surroundings, and the word usage is wrong to call it combustion. It can be said that the explosive power of explosives containing nitrogen N has been overlooked in the science of chemistry, with the exception of the fact that this (H2O) is generated. The fact that water drains or falls from the exhaust pipe of a car that runs on oil is that the (HCOH) group in the oil takes in oxygen (O) and becomes a sugar molecule unit (HCOH = H). After that, it is the core part that (C) is extracted by (OO) and (HO = H) is implosion generated, and in that sense, gasoline car / diesel car / hydrogen fuel The common principle that both cars are implosion generated (HO = H) is overlooked. The generation of (H—O = H) generates a large energy, although the positive and negative energies of whether (O = H) is generated from two atoms or separated into two atoms are small, but (O = H) When (H) and (H) are present, it is understood that (H-O = H) becomes a magnetic path short circuit of spin magnetism, so that the largest energy after the nuclear energy is thermally radiated or directly generated. The In other words, (O = H) or (H) is an ordinary existence that has no energy, but it is the only chemical in the universe that causes implosion and becomes (HO = H). It is an abyss (low energy level), and when it falls there, an extraordinary amount of energy is released, and the same energy is required to separate into (O = H) and (H) by electrolysis helped from the abyss. In the universe, there is no molecular bond that has large energy in and out, and it is at an ordinary energy level. So far, science has made a mistake in this regard, focusing on the fact that there is one rich man, but it has never been possible to pay attention to the strength of one poor man with no bottom. .
Plants make sugar molecular units (HC-O = H), have no neurotransmission, and root the water molecules (H-O = H) necessary for photosynthesis of sugar molecular units (HC-O = H). The water molecule (H2O) in the atmosphere cannot be used. The proportion of plants containing (H2O) in the enclosure is much smaller than in animals. Animals that take in and use sugar molecular units (HC-O = H) made by plants by taking food into the body carry water molecules (HO = H) into the body, while oxygen molecules from the atmosphere. (O-O) is a group of power generation units that extract the carbon atoms (C) of the wedges from the sugar molecule unit and impregnate (HO) one by one. The nerve pulse that contracts and controls this is transmitted to the nerve cord as an electrical signal in units of decomposing one sugar molecule unit to produce one (HO = H). This one nerve pulse causes the fingertip to perform a TAP operation, and is a quantization unit that cannot be subdivided further. It is reasonable to use a TAP operation generated by muscle contraction by one quantization unit command as a single pulse waveform for parallel processing in a key matrix. (H2O) produced by animals can be reused in the body, but even with intense muscle use, intense brain use, the fatty group is the starting point, and the sugar molecule unit is the starting point directly, the result is a large amount Carbon C is sandwiched between sugar molecular units (HC-O = H) by (OO) and becomes (HO = H). (H2O) becomes surplus, and the body weight is reduced by 2-3 kg due to the discharge. If you exercise, you can drain by sweating, but if you use just the brain, it will be another drain. Brain power generation requires instantaneousness, so a large amount of sugar molecule units are sent through blood vessels, but fatty groups are never used.
The fast white muscles that make instantaneous movements and thicken muscles when trained consume oxygen molecules in units of sugar molecules instantly, whereas they operate continuously and are difficult to strengthen even when trained. The red slow muscles cause the fatty group (HCH) that has been stored in a reduced weight to interrupt the single oxygen atom (O) generated in the activation process shown in Formula 3 into the sugar unit, The carbon atom (C), which functions as a wedge, is sandwiched between oxygen molecules (O—O) to generate (HO—H), but this method has high fuel use efficiency per weight. Sustainable but lacks flashiness. If the fast muscle does not collect hemoglobin with oxygen in advance in the muscle fiber, it cannot be instant, so it will soon run out of breath.
While the fingertip moves about 10 mm by TAP, it can be said that the key stroke is limited to 1 mm to 2 mm. Due to this difference, every time the key is hit, the durability acceleration test with a load that gives an impact to the contact is repeated until failure occurs. The standard reliability standard is 10 million times. Recently, a durability test of 50 million times or more may be imposed as a sampling inspection. The surprising thing is that human fingers greatly exceed this durability. From another point of view, it can be said that the operator continues to polish the surface in order to force the electrodes to conduct with great durability. In other words, if a human finger first develops tendonitis, the keyboard will not break. When the key is struck, the upper electrode collides with the lower electrode and the stroke stops, but if the key itself is stopped by a mechanical buffer stopper before and after that, the impact is weakened. However, due to the secular change of this buffer mechanism, the critical condition is imposed on the conduction, and such a buffer stopper already has the membrane contact itself or the pantograph mechanism. The buffer stopper and its subtle impact sound are speculative confirmation means that the key will surely be pressed, and the bottoming is unpleasant, but the awareness that click sound confirmation is supplemented as comfort Is a habit. However, it can be said that it is impossible to perceive as comfort that the rubber dome is bent and the drag is reduced.
In order to eliminate the dilemma caused by these situations, the collision does not occur and the slice is made in the middle of the stroke, and the confirmation that the key has been accessed can be heard by an electronic sound, or the repulsive force of the magnet never causes the collision. There is a method of preventing the occurrence of damage, or because both electrodes are in close contact with each other over a relatively wide area, soft landing with a compressed air flow reduces damage. When pressing a button on the remote control, it is a form that moves slowly by increasing the repulsive force of the rubber button instead of TAP. The action of pressing the button on the remote control quickly accelerates the deformation of the conductive rubber. When pushing a seal, you don't know how hard you should push it, but you have to push it again after seeing the result.

Conduction between metal electrodes results in malfunction as a switch because the surface of the metal is oxidized, but when a large current is passed, it is interpreted that the formed insulating film is broken and re-conducted repeatedly. Yes. When the flowing current is very small, the effect of re-conducting is reduced, so that it is common to apply gold plating that is difficult to oxidize to the surface of the metal electrode. Even in a conductive rubber contact that requires conduction between a metal electrode and a conductive rubber electrode, it cannot be said that the deterioration of the contact is accelerated when the current flowing during conduction is small. If it is empirically considered that the deterioration is slower when a large scan current is applied, the termination resistance is reduced, and the current consumption of the scan increases. If the contacts are not electromagnetically shielded against disturbance, a small termination resistor must be selected to avoid malfunction. Conduction between metal electrodes does not occur over the entire opposing area of the electrodes, but as a collection of multiple contact conduction points. In general switches, pressure can be applied to increase the total area of contact points, but in a key matrix, this pressure cannot be expected. Therefore, the structure is intended to adjust the contact surface by using conductive rubber on one side of the electrode. This deformation works well in the early stages, but promotes contact failure later in the period when the deformation of the rubber is fixed or the conductivity inside the conductive rubber material is degraded.
To compare metal-to-conductive rubber and metal-to-metal contact failures as two options, what is conductive rubber, what is a metal lattice, what is a work function, molecular level / atom It is essential to have a deep understanding at the level. As shown in FIG. 10, the carbon atom C is considered to be a triangular atom having a two-layer structure, and interleaves to form a hexagonal ring (C6) autonomously. This is the benzene ring. It can be said that the presence or absence of hydrogen atom H is irrelevant to this basic skeleton structure. When carbon atoms are placed vertically so that the two-layer structure is up and down, if the spins of the three orbital electrons in the upper layer are counterclockwise, the spins of the three orbital electrons in the lower layer are clockwise. Call it. Conversely, if the spins of the upper three orbital electrons are clockwise, the spins of the lower three orbital electrons are counterclockwise, and this is called R-type. The L-type is L-type even if it is turned upside down, and the R-type is R-type even if it is turned upside down. Therefore, the atoms in the even nuclear layer that cause the distinction between L-type and R-type are L-type / R-type is present at a ratio of 50% / 50%. This is considered to be the truth of what was called optical anisotropy, and the mirror image symmetry itself that has been said to come to mind is not the cause of optical anisotropy, but only one of the results. From this point of view, optical rotation does not occur in the topology of molecular bonds, but occurs as a difference in L / R of atomic units. The benzene ring and graphite are L-type and R-type alternately arranged to form (C6), so the relative position does not change and is not flexible. On the other hand, in diamond, a single crystal is constituted by only C atoms of either L type or R type. It can be understood that a single crystal crystal is also composed of a combination of either type of Si atom and O atom and becomes a right crystal or a left crystal.
Even if a single-layer graphene with a benzene ring formed into a net bond is a perfect L / R type interleave, the carbon nanotubes of a seamless pipe can be rolled to create an L / R type complete interleave as shown in FIG. It is maintained and has become a useful material. On the other hand, the fullerene (C60) is not a hexahedron but a 32-hedron, but is composed of 20 hexagons and 12 pentagons like a soccer ball. If six hexagons surround one hexagon, the connection maintains a complete L / R interleave as LLRLRLR to form a plane. However, if a curved surface is forcibly created by surrounding a single pentagon with five hexagons and connecting them one after the other, the entire soccer ball is managed to maintain a spherical shape that is not connected as an LLRLR. It turns out that the ring is incompletely shaped like an incomplete sphere. If the pentagon is a simple cavity and is composed of 20 hexagons, as shown in FIG. 11, the entire sphere has lattice defect lines at the fifth and tenth, and the twelfth and seventeenth locations. Become an account. In nature, such improper things in the non-living world and in the living world are not stable and significant, and all are appropriate, where some rational relationship is maintained. In most cases, this rationality is achieved by repeating the L-type and R-type in a complementary manner (adjacent spin directions are reversed in the horizontal direction and aligned in the same direction in the vertical direction). In other words, it is claimed that a method for producing fullerene with an ultra-low yield has been found, but unlike carbon nanotubes, no practicality has been found yet because of its lack of rationality. It can be understood that it is only there.
In the rubber molecule represented by the repetition of (C5H8), the hydrogen atom (H) is irrelevant to the formation of the skeletal structure, and a torsion spring is formed by the connection of the carbon atom (C) having a triangular nucleus. It can be said that flexibility, elasticity, and elasticity are expressed. The six carbon atoms that should be closed as a hexagonal ring are LL or RR step coupling where one side of the hexagonal ring should be LR repeats, and the opposite side is also LL or RR. I can't close it. This shape is repeated for a long time to form a zigzag spring having elasticity in a chain as shown in FIG. Since the biological DNA spiral structure is originally created with this LR step, natural rubber and muscle expansion and contraction spirals can be easily replicated and mass-produced.
In natural rubber, since H atoms are bonded to C atoms, the orbital electrons cannot be ionized, and the C atom pseudo hexagonal ring skeleton suppresses the ionization of orbital electrons, so that the number of free electrons supplied is low. Is the body. The conductivity of matter means that there are abundant orbital electrons that are ionized first and supply free electrons, and then, when free electrons travel, they are taken into the orbital electron sphere and held by the attractive force of the central nuclear proton. It is thought that it is made up of two factors: a small proportion of accumulation. Of these two requirements, the natural rubber, which is an insulator in the former sense, is made conductive by mixing a filler made of carbon black, which is a conductor, into a needle shape. Instead of increasing the number of free electrons, as shown in FIG. 13, a conductive chain is established by contact conduction between fillers having free electrons. When one of the contact conductive chain paths is no longer in contact chain, the conductivity is lost and returns to the insulator.
If the free electrons move continuously from the filler with free electrons to the natural rubber region, conductivity is obtained as a whole, but there is a work function when the free electrons move from the filler to the natural rubber region, Free electrons remain in the filler, and eventually there is no conduction path other than the movement of free electrons through the filler. The problem is that mixing the filler gives conductivity while at the same time losing the flexibility of the natural rubber, so if the conductive chain between the fillers becomes solid due to deformation due to overuse, the surface will be cleaned somewhat. Even if scraped, the contacts will not be restored, and since the flexibility is lost, the number of surface contact points is reduced. That is, it can be said that the combination of the metal electrode and the conductive rubber electrode is not the final solution that human beings should select as a conductive contact, and the basic structure of the molecule can be said clearly. In this case, praise is essential.

When the counter electrode is brought close and brought into close contact, the conduction switch that maintains several contact points that have become familiar and fixed for a while will actually function as a capacitance that suddenly increases the process until contact conduction. Depending on the distribution of impedance, there may be a state where it can be said that the switch is turned on before it becomes conductive. Strictly speaking, the conductive type and the capacitive switch type cannot be clearly distinguished. This change in impedance between the electrodes is a state (for example, in terms of series insertion impedance) in the process until the capacitance becomes maximum, if a pure numerical understanding such as a vector display in FIG. Changes significantly, and there is no noticeable change in conducting further after close contact. It can be said that the continuity type has caused many problems due to the fact that it does not try to discriminate by exaggerating this unchanged part.
Capacitance switch type can be said to use a linear change with respect to the separation amount as an impedance proportional to the reciprocal of the electrostatic capacity, depending on the ON / OFF discrimination, but the change of the variable capacitance value The fact that the critical dependence on the driving impedance of the X line is used for discrimination is not correctly explained. This is not a measure of variable capacitance value, but it cannot be advanced unless it is handled in the dimension of impedance change. If the driving resistance r is small, a relative impedance relationship cannot be obtained, so the magnitude of the transmitted differential pulse does not change depending on the coupling capacitance.
On the other hand, the contact interface utilizes the fact that the spontaneous pulse is actively generated when going from contact to separation or from separation to contact, so that the transmission pulse is not further reduced, and the pulse generation mechanism The difference is that it is almost completely electrostatically shielded. Discrimination has already been completed due to close contact from the distance, and even if close contact finally leads to conduction at the contact point, there is no effect on the discrimination operation, and there is no need to press the electrode strongly as in the conduction type / scan type. Therefore, the deterioration of the electrode surface is alleviated. Even if conduction (= leakage) occurs and the insulating film is repeatedly covered again, there is a feature that the operation continues without any problem.
Despite the fact that there is no substantial difference between the conduction type and the capacitance switch type from the point of view of the contact impedance vector diagram, the two are different principles and no one refutes today. Bringing the state. If this is correctly interpreted as being similar, then the opposite will show itself, and the problems will go to ease. In other words, since it is contactless, durability is not excellent, and the core process that brings about the greatest change in reality does not include whether or not it is conducted. Therefore, the conduction type that is based on whether or not it is conductive and the capacitance switch type that is based on the change width of the capacitance value miss the sweet spot as the core.
When this core part is understood by the close contact interface type, if the distance between the electrodes is changed faster than the charge is redistributed faster than the one charged in advance by applying a voltage difference between the electrodes, A negative pulse voltage dip is generated. To explain more precisely, the dipped voltage change immediately returns to the original voltage with the CR discharge time constant, but as shown in the overall view of the parallel processing type contact interface type in FIG. Immediately before the contact when the capacity increases rapidly and the CR discharge time constant increases, the discharge is not completed and a negative pulse remains.
Similarly, as shown in FIG. 15, the IC chip of the contact interface type key matrix is commonly used for the on type / off type, and the external circuit is simplified because a clock for scanning is unnecessary. A slightly smaller package with fewer pins is used, reducing power consumption and simplifying the circuit inside the IC chip. The same IC chip can be used to use a rubber dome or a coil spring. It is no longer necessary to use the pantograph mechanism to reduce the height of the keyboard, and in that case, the same IC chip is also assumed. The analog amount when each key is pressed is reflected, the slice level optimized for each key is given from the host side, the latch circuit corresponding to individualization, and the eight keys of the home position and one selected key Incorporation of a circuit for exchanging latch information with the host for performing more accurate and immediate determination by comparing 101 patterns with the stored pattern makes the original circuit of the IC chip easier Can be assigned so that the cost does not increase.

In understanding the operation of the capacitance switch type PC keyboard, which is often described as utilizing the change in the capacitance value between the counter electrodes, the relational expression Q = CV is cited. In that sense, it is understood and explained that the amount of accumulated charge Q increases as the capacitance C increases due to close contact. However, it is often correct to say that Q is rather unchanged.
The idea of using the capacitance C originally appeared as an alternating current theory, and the impedance is associated with voltage and current as an imaginary parameter of Z = 1 / jωC. However, this concept of parametrics is a hindrance when directly understanding transient phenomena. Even if the key scan pulse passes through the capacitance switch, the height of the differential pulse does not change depending on the capacitance value. The actual step response cannot be obtained correctly in the frequency domain, and it is necessary to perform a transient analysis in the time domain. That is, the relational expression Q = CV is not useful for understanding the capacitance switch type.
On the other hand, in the contact interface type, for example, when a variable capacitor is connected to a voltage source of 5 V through a termination resistor R, a sudden change in the distance between electrodes cannot cause a change in the charge amount Q. Since the electric field E between the electrodes is determined by the amount of charge distribution per unit area, the electric field E between the electrodes is constant regardless of the distance d unless Q changes. Since the potential difference V between the electrodes becomes a relational expression of V = E * d obtained by multiplying the electric field E by the distance d, the potential difference V between the electrodes suddenly increases when the distance between the electrodes suddenly increases. Since this is different from 5V of the voltage source, the charge moves through the termination resistor R, and the potential difference between the electrodes of the capacitance eventually returns to 5V. At this time, the coulomb attractive force works between the positive and negative electrodes, and the work of widening the distance between the electrodes is performed. It is an understanding of the principle similar to adiabatic expansion that d is changed from d = 1 mm to d≈0 contact, so that the potential difference V becomes smaller in proportion to d. Inspire engineers with new ideas on behalf of education.
Even if the capacitance C changes, no change occurs unless it is charged in advance. In this case, the concept of capacitance becomes virtual. If the distance between the electrodes is changed, the counter electrode is precharged. However, if the current I is not taken out (even if there is no power of V * I), work is given or received, and the potential difference V changes. . In other words, if d = 0 is called conduction, there is no reason to distinguish it from immediately before d> 0. The current distinction between the conduction type and the capacitance switch type is free from this essence. The reason for approaching conduction is that a barrier called a contact potential difference is the difference in work function when free electrons jump out from the surface of the crystal lattice of one metal plate and enter the crystal lattice of the opposite metal plate from the surface. There is, however, no barrier before d = 0, and it is merely a meaningless, notional distinction between conduction and close contact capacitance at the atomic level. It is not the essence of the operation to forcibly distinguish the case where d = 0 is conductive and d> 0 is close, but it is not the essence of the operation, but the change from the separation of d = 1 mm to the close contact of d≈0 is equally divided. Understanding is the essence. The concept of the capacitance switch type deviates from the optimum design in order to overestimate immediately before contact, and the conduction type also overestimates immediately after contact and creates the problem of conduction failure.
There is a possibility of chattering when the boundary between close contact and conductive state is generated in both the conductive type and the capacitive switch type, but this is caused by a change in the work function of the interface before and after the close contact. Since the change that equally divides the process from the separation of d = 1 mm to the adhesion of d≈0 is the core part, the chattering is achieved if the separation of d = 1 mm, which is 90% of the change width, is set to d = 0.1 mm. Never wake up. Also, electrical deterioration of the interface does not occur. Even if the change width becomes 90%, there is no problem in detection. In the contact interface type, since both electrodes are separated by a dielectric insulating layer of about 10 μm, no chattering actually occurs.
In this way, apart from the classical electromagnetic point of view of whether or not it is conducting, from the viewpoint of the atomic / molecular structure, the viewpoint of the adhesion interface is actually at the center of various phenomena, and it is clear that it forms the boundary of change. The This can be placed at the heart of the key matrix. On the other hand, placing the orbital electron spin at the core is also a fundamental method. This is the induction type. The present invention builds a system based on such fundamental elements.

As shown in FIG. 18, the basic idea of the capacitance switch type is that a waveform is transmitted from the X line to the Y line. When the capacitance is close and the capacitance value increases, the waveform that is transmitted is sliced, but it is easy to inadvertently look at the frequency characteristics from zero to the frequency of the sine wave burst you want to use. Since it changes relatively linearly, it is assumed that analog information can be embedded. However, since the frequency is actually fixed and the distance between the electrodes is changed, it can be seen that the change is biased when d is taken on the horizontal axis and the transfer function is viewed. That is, in order to embed analog information, the drive waveform is a sine wave burst instead of a rectangular wave, but this makes sense because of this bias. In the driving of the rectangular wave, the amplitude of the CR differential waveform does not change even if the capacitance is changed. Therefore, it is necessary to configure the LPF with the driving resistance r and the stray capacitance so that the amplitude of the transmission waveform changes. However, the bias due to this compromise is about the same as the bias for using sine wave bursts, and there is no point in using sine wave bursts. Therefore, in the following description of the capacitance switch type, all are represented by rectangular wave drive, but this does not necessarily mean that the capacitance switch type product does not use a sine wave burst. It's just that it makes no sense to use bursts. In the end, it is not possible to use the scan method to correctly embed an analog quantity and read it, but to switch to a parallel processing coherent interface type.
Compared with the conduction type / capacitance switch type performed by key matrix scanning, the change between the contact state where the parallel plate electrodes are separated by a dielectric insulating layer of about 10 μm and the state where the surface is separated by about 100 μm is shown. As shown in FIG. 15, the contact interface type received as the amount of charge transfer is one of the intersections of the X line of N rows and the Y line of M columns, and the distance between the X electrode and the Y electrode facing the G electrode. In the on-type that focuses on the close contact, the IC chip detects that negative pulses are generated at approximately the same time on the X and Y lines forming the intersection. As shown in FIG. 16, the close contact interface type includes an off type that pays attention to the moment when the X electrode and Y electrode that are in close contact with each other are separated from the G electrode.
In FIG. 17, the capacitance switch type that detects the change in the capacitance between the X / Y electrodes as an impedance by the transfer function of the scan pulse and the equivalent of the two contact interface types of the on type / off type The difference in the circuit is shown. Only one place where the intersection of the X line of N rows and the Y line of M columns is N * M is in close contact, and the remaining N * M-1 places are all 100 μm or more away from the surface. In addition, only one part is 100 μm to the surface, and the remaining N * M−1 parts are all in close contact with each other, and when using a conventional key matrix scan, the on type can be detected. However, the off type is difficult to detect. On the other hand, since the close contact interface type is used in parallel processing, an on type and an off type can exist equally.

When comparing the contact type on-type and the off-type in terms of key touch, the on-type detects a change occurring in the second half of the 1-2 mm stroke of the button, and an impact at the time of contact is applied to the interface. On the other hand, in the off type, the change immediately after the start of the 1-2 mm stroke of the button is detected and the interface is only separated from the contact, so the impact does not necessarily occur, and the voltage that should rise rapidly is the contact capacity value of the same line Is absorbed by the sum C1 * (N-1). Since the impact when the fingertip comes back from the button and returns to the close contact due to the spring property of the rubber dome is gentle and constant, it does not impair the semipermanent reliability of the close contact interface type. The point at which the switch should operate during the stroke of 1 to 2 mm is no longer meaningful, and the contact interface leaves and the positive pulse is generated at the moment when the resistance of the rubber dome is bent by the compression of the fingertip meat. This is a simple operation, and should be an ideal form with little time delay.
In the on-type, even if one of the contact interfaces becomes conductive due to aging, it is possible to select an operating point that can continue to operate as it is, and the continuity is canceled and the capacitive coupling is achieved by subsequent aging. It will continue to work as usual. On the other hand, in the off type, it is difficult to design such that one of the adhesion interfaces continues to operate even when conduction occurs due to aging. In summary, it can be said that there are merits and demerits in the off-type that can reduce the drag with an immediate response without delay and the on-type that operates at the end point of the stroke and can allow conduction due to deterioration.
Rather than change the distance d of the adhesion interface over the entire stroke of about 3 mm which gives ideal lightness, in reality, the change of d in the range of 100 μm is used, but it is about 3 mm This gives you the option of whether it is at the beginning or end of the reciprocating stroke. In other words, ~ 2.9mm is a combination of whether it is just play, run-up, or a section that receives momentum, so far the membrane contacts and capacitance switch type were struggling under conditions that could not move. Different from the taste. For the reliability / comfort / character input speed / reducing mistakes / reducing fatigue / reducing tendonitis, the designer can provide a framework that allows the user to select a combination that suits the design intention. This is the purpose of the present invention. It is desirable that one IC chip can handle any of the combinations.

Strictly speaking, it is 101 keyboard, but it is omitted and assumed to be a 10 * 10 matrix. One digital TV average remote is also assumed to be an 8 * 8 matrix for simplicity. In the conduction type, the disturbance picked up by the wiring is superior to the disturbance picked up by the contact point at the intersection, and the smaller the termination resistance R of each Y line, the smaller the noise amplitude constantly applied to the IC chip due to the disturbance. Further, the driving resistance r of the X line is added in parallel to the termination resistance R while the contacts are conductive. The fingertip touches the button and relays the disturbance to the contact portion rather than the wiring. Similarly, the smaller the termination resistance R and the driving resistance r, the smaller the noise amplitude due to the relayed disturbance.
When the capacitance switch type is used for a PC keyboard, reducing the termination resistance R reduces the influence of disturbance, but at the same time, the amplitude of the scan pulse transmitted from the X line to the Y line is also reduced. The drive resistance r of the X line cannot be reduced because an LPF (low-pass filter) needs to be formed between the X-line and the Y-line stray capacitance. At the same time, the X / Y line is surrounded by a GND pattern on the PCB substrate as shown in FIG. This also helps to shield against transmitted pulses that are unavoidably attenuated by the LPF and r / R ratio. At the point where the fingertip touches the button and relays disturbance, even if the metal spring is AC grounded to the X electrode, r cannot be sufficiently suppressed because r is not driven at a constant voltage.
Compared to this harsh condition, the contact interface type can place the G electrode on the upper side to make an electrostatic shield on almost the whole surface, so the influence of disturbance can be greatly reduced, and an iron plate is also placed underneath If electromagnetic shielding is used, perfect shielding is achieved, but on the other hand, the generated pulse signal is also reduced, and overall, the S / N ratio is not much higher than the conduction type / capacitance switch type.
In the capacitance switch type, as described with reference to FIG. 5, the capacitance between the conical metal spring and the Y electrode has a small variation width because the spring material has a circular cross section, and Since the fixed capacitance between the outermost circumference of the conical metal spring enters in series, the change width is further reduced. For this reason, even if it is devised, for example, the maximum series capacitance is about 4 pF and the minimum capacitance is about 1 pF, that is, the change width is about 4 times, and the floating capacitance of the Y line is about 5 pF. If the premise is applied to the capacitance switch type equivalent circuit shown in FIG. 19, it can be seen from the step response that it is quite difficult to secure the on / off ratio of the transmission pulse. When the magnitude of the differential pulse when there is no LPF effect is simply seen with respect to the capacity division ratio, the difference is about three times.
The interval between the buttons of the PC keyboard is about 19 mm, and the area of one button is 19 mm * 19 mm = 361 square mm, but as shown in FIG. 16, 44% of the area is about 80 + 80 square mm. It can be said that assigning to the counter electrode area of X and the counter electrode area of Y is a design that is reasonable for a mechanism that makes the inter-electrode distance variable. If the thickness of the adhesion interface type dielectric insulating layer is 10 μm, the relative dielectric constant is 2.5, and the counter electrode area is 80 square mm, the capacitance at the time of adhesion with the 10 μm dielectric layer sandwiched is C1 = 177 pF, the surface is The capacitance at a distance of 100 μm is calculated as C2 = 6.8 pF. When the number of columns is M = 10, the CR time constant with the termination resistance R = 1 MΩ is (177 pF + 9 * 6.8 pF + 5 pF) * 1 MΩ = 0.243 mS in the on-type as shown in FIG. If the change in adhesion between the surface separation of 100 μm and the dielectric layer of 10 μm occurs instantaneously, the charging voltage of V = 5V applied initially becomes 0.2V, and 10 C2 of the same X line The capacity 10 * 6.8 pF = 68 pF is hanging, and the degree of voltage drop is small. A change of ΔV≈−5V occurs instantaneously and returns to V = 5V with a time constant of 243 μS. The actual stroke generates a negative pulse of hundreds of mV because the voltage drop overcomes the CR time constant immediately before the close contact of the stroke of 100 ms.
On the other hand, in the off type, as shown in FIG. 20, the time constant is (10 * 177 pF + 5 pF) * 1 MΩ = 1.78 mS. If the surface separation of 100 μm instantly occurs after the 10 μm dielectric layer is sandwiched, the initially applied V = 5V charging voltage is 130 V, but nine C1s are hanging on the same X line. Since the voltage rise is absorbed by the entire 1655 pF, a change of ΔV≈ + 5 V * 1/9 = + 0.55 V occurs instantaneously, and returns to V = 5 V with a time constant of 1.66 mS. Since the actual stroke requires about 70 mS, the initial voltage change of several tens of μm away from the contact occurs in a time of 1 mS or less, and discharge is performed by applying 1.66 mS, but the voltage change in the latter half is small and immediately at 1.66 mS. As a total of the discharge process, a positive pulse having a necessary and sufficient magnitude of several tens mV is generated.
It can be said that the terminating resistance R = 1 MΩ is a value sufficient to overcome the leakage current caused by the mixing of dust and the like and to provide durability. Actually, there are not N + M termination resistors R of 1 MΩ in the IC chip. As shown in FIG. 20, a negative feedback circuit composed of CMOS transistors only appears to be equivalent to 1 MΩ. The sum of the adhesion interface capacitance and the stray capacitance outside the IC chip plays a role of capacitance for circuit stabilization (preventing local oscillation). When such a mechanism is compared, it can be said that off-type is somewhat advantageous, but there is no great difference as a product, and both can be used properly in the right place. When comparing the on-type / off-type parallel processing type close contact interface type and the key scan type capacitive switch type PC keyboard, industrially, the capacitive switch type can survive in all aspects. It's hard to say.
In the case of a TV remote control composed of an 8 * 8 matrix, as shown in FIG. 16, the contact interface type counter electrode area can be about 40 + 40 square mm, which is half of the PC keyboard. When this is used as an off-type, the CR operation time constant is 0.7 mS for the termination resistance R = 1 MΩ, and there is no big difference in the essential operation principle. However, unlike a remote control so far, the operation of pressing strongly may slow down the speed and prevent the threshold from being exceeded. Instead, lightly pressing the thumb will naturally increase the speed and provide the required generated pulse magnitude. If this practice of proper pushing is acquired, consumers are freed from fear that the remote control will always malfunction. In the coherent interface type, a slight Coulomb attractive force acts between the charged electrodes, but the force is negligible compared to the conventional key drag force of 50-60 g.

Capacitance switch type and contact interface type can be understood accurately only by considering them in the time domain instead of the frequency domain. As shown in FIG. 21, when considering the uniform contact process in a stepwise manner, first, as the inter-electrode distance d becomes smaller, the voltage step change amount becomes 1 / d for the same differential step amount. It grows in proportion to. Next, since the capacitance C increases in proportion to 1 / d, the CR time constant of discharge also increases. Due to such a synergistic effect, negative self-sustaining pulses are concentrated immediately before contact. This is on-type. On the other hand, in the direction of separation, positive self-sustained pulses are concentrated at the moment of leaving the contact. This is an off-type, which is a form in which the order of the on-type, the contact and the separation are interchanged.
For simplicity, the waveform combining the case where the inter-electrode distance d defined without sandwiching the dielectric insulating layer of about 10 μm between the electrodes approaches the contact from the separation at a constant speed and the case where the contact d moves away from the contact is expressed by Equation 4 It is represented by the differential equation shown in
Rather than spending time to solve this, practically iteratively calculating, as a first-order approximation, the bias voltage is 5V, and the on-type of the adhesion interface, as shown in FIG. When the terminal resistance is sufficiently large, the voltage generated when the terminal is in close contact with each other and immediately reverses at a speed of 1 mm is proportional to the voltage d. Therefore, a broken-line triangular wave that reciprocates between 5 V and 0 V is generated. If the terminal resistance is reduced practically, the change will be from a solid line to a dotted line. In the case of the on-type contact interface, the contact is actually made at the apex of the downward parabolic motion, so the occurrence in that case is also noted on the right.
On the other hand, as shown in FIG. 20, in the off-contact type of the contact interface, the generated pulse that is 1 mm apart from the contact at a constant speed and immediately inverts the contact and has a sufficiently high termination resistance, for example, A broken line waveform reciprocating between 5V and 7.5V is generated. If the terminal resistance is reduced practically, the change will be from a solid line to a dotted line. In the case of the off type, since the interelectrode capacitance of other addresses of the same X / Y line is hung at the maximum value, the voltage change is shared and becomes smaller, so the separated waveform becomes larger. There is nothing. Compared to the on-type, it can be seen that the waveform changes sharply at the moment of leaving the contact. When the initial drag is small, a determination is made at the moment when the key is hit, and the response becomes a one-touch prompt response that is completely different from the keyboard of the conventional key scan.
Looking at such voltage changes, it is difficult to understand why key matrix scans have been performed. If there is a parallel processing pulse that does not do that and it can be read separately, it is understood that scanning is meaningless. However, the key scan sometimes has the meaning of serialization (sequentialization) when data is transmitted from the keyboard to the PC. However, in actuality, which 101 was pressed is only 7-bit information, so there is no contribution in this sense.

As shown in FIG. 22, the membrane-type PC keyboard is an intersection of about 3 square mm at the intersection of the upper sheet with the X line drawn on the back and the lower sheet with the Y line drawn on the surface, and the lines are perpendicular to each other. By bending the counter electrode with a relatively large pressing force and making it conductive, a key matrix is formed only by a simple two-sheet configuration without a jumper and a spacer, and the manufacturing cost is conspicuous. On the other hand, if the contact interface type of the two counter electrodes that changes the distance between the X / Y electrode and the GND electrode is realized only by the configuration of the two sheets and the spacer, the GND sheet is a powerful electrostatic shield on the entire surface. However, it is necessary to provide a jumper on the X / Y electrode sheet to cross the X line and the Y line. The parallel processing makes the IC chip simpler, but the sheet manufacturing process becomes complicated.
A simple method of avoiding this and avoiding the addition of the jumper process is a single-electrode type of about 100 square mm in which the lower side is an X-line sheet and the upper side is a Y-line sheet, and the distance between the electrodes is 100 μm to As an on-type or off-type that detects a change in a close contact state and a close contact state through a dielectric insulating layer having a thickness of about 10 μm, each X line is set to a high voltage by a terminating resistor R in the IC chip. The Y line is connected to the Low voltage by the terminating resistor R to give a steady bias potential difference between the opposing electrodes, and a positive / negative pulse due to a transient change from close contact to separation is generated between the X line and the Y line straddled by the hit button. A parallel process that does not require a key scan to identify the address by detecting that a large pulse is extracted from only one differential amplifier that appears in the differential direction on both sides. As a matter of fact, a matrix can be constructed. This method can increase the area of the counter electrode by 30 times compared to the membrane-type conductive contact, so the impedance is reduced and the influence of disturbance is reduced, and it is sensitive to small pressing pressure as a large area bending, so a light stroke It can be operated with. This method minimizes the modification of the manufacturing process of the current membrane type PC keyboard, and the manufacturing cost is almost the same as the IC chip simplification and the dielectric film formation cost offset, and the reliability is simply semi-permanent. In addition to the great progress that makes it softer and the key touch becomes softer, it is possible to acquire analog information and to provide a learning function for personalization with the keyboard that I just bought for the future. I will put it.
Regarding the suppression of disturbances, the component inducing the in-phase voltage of the X line and the Y line as antennas is canceled in the common mode. The number of differential amplifiers is the same number of N * M for a matrix of N * M size. Can be further reduced.
Since the area of the FPC (flexible substrate) to be pushed down is large, the contact is made with a force much smaller than that of the membrane type, so that the ON determination is made with a touch smaller than 50-60 g, and the time from the start of hitting / pressing until the ON determination is made short. The key height can be lowered by replacing the pantograph type of the notebook PC. When a large area is pushed down, the sheet under the adjacent button is pulled to increase the tension, making it difficult for the sheet to be in close contact with each other. In other words, a membrane contact type opens up new possibilities by simply giving up conducting and ending with simple contact.

As shown in FIG. 23, the buttons of the remote control formed by integral molding of the rubber sheet and the membrane type / pantograph type / capacitance switch type of the PC keyboard can cause problems when the buttons are tilted and pressed. There is a high standard of requirements for a mechanism that is parallel and descends in parallel. On the other hand, in the close contact type on-type / off-type mechanism shown in FIG. 16, the degree of demand for the button surface being tilted and pushed is low in principle.
The membrane type / pantograph type both have a membrane contact with a diameter of about 2 mm, the force transmitted from the fingertip is transmitted to the thin protrusion at the center of the rubber dome, and the diameter of about 5 mm surrounding the membrane contact is about 100 μm high. The rubber dome is tilted because it does not lead to conduction unless the spacer is accurately hit the center of the cylindrical hollow space where the spacer is broken and the film above it is bent accurately and symmetrically as a strong force per unit area. It is not allowed to bend. If the bottom area of the protrusion is increased, the same force is dispersed and the force for bending the film is insufficient, so it must be reduced, but it is affected by the inclination of the button causing the displacement from the center. In addition, when the opposing area of the upper and lower X / Y electrodes increases, the membrane contact must be about 2 mm in diameter even at the point where it is capacitively coupled and the differential waveform of the scan pulse is transmitted to reach the slice level. . In order to eliminate this, in the membrane type, the guide wall that pushes the button down is raised so as not to tilt, so the entire key mechanism cannot be lowered. On the other hand, in the pantograph format, since the pantograph keeps the parallelism, the tilt can be kept small without relying on the height of the guide wall, but the discomfort with a stroke of about 1 mm and the key next to the selected key. When touched, the adjacent key is not tilted, and the two keys are lowered at the same time. In the capacitance switch type, the maximum value of the variable capacitance becomes small when tilted, and the determination of the pulse height transmitted to the Y line is a critical system, which hinders the operation.
In the remote control that measures the conduction between the X / Y electrodes with rubber contacts, when the Z electrode is tilted and descends, the Z electrode and the X electrode are delayed while the Y electrode and the Z electrode are first conducted and maintained. Achieves the purpose by conducting. At this time, if the Z electrode of the conductive rubber is further compressed on the Y electrode side for many years, the deformation is gradually fixed, and there is a possibility that the Z electrode and the Y electrode become insufficient in height and become non-conductive. . If it is tilted and pushed, it will push harder by that amount, and trying to make both X / Y electrodes conductive will accelerate the deformation of the rubber contacts and the deterioration of the surface. There is also a need for a mechanism that can be pressed against the mechanism that keeps the parallelism.
In the case of parallel processing, let's consider the example of the form of close contact interface and the type of induction reversal that will be described later. In the contact interface type, the sheet that has been subjected to tensile stress until then is in a relaxed state at the moment when the ON type leads to contact and the pulse is generated. The seats are still in a relaxed state, and it can be said that the influence of the button tilt is difficult to reach at both points. In the induction-type top-and-bottom reversal format, the cylindrical magnet on the back side of the iron plate is said to be hardly affected by the tilt of the button in that it is always in close contact with the end of the stroke regardless of the stroke process.
If the eight fingertips that hit the PC keyboard are used as the home position with the ASDF-JKL: button, and the fingertip is extended from there to another button, the button cannot be pressed from directly above. Even in such a case, the protrusion must come down accurately to the small conductive contact. In principle, it is difficult to move the button down in parallel regardless of where the button of 19 mm * 19 mm is pressed, or to come down in parallel even if it is hit diagonally. This essentially switches the way of thinking and the only solution is to change to a method that does not hinder the detection operation even if it does not descend vertically.
In the contact interface type, as shown in FIG. 24, the rate of change in capacitance before and after contact between the G electrode and the X / Y electrode divided into two parts over the entire area under the button is related to the magnitude of the generated pulse. Therefore, the parallelism of the counter electrode is important, but since the close contact state is a relaxed state, the parallelism at the moment when the pulse is generated is maintained, and the required standard is relaxed. The main point is that in the on-type, the microdome that protrudes from the center of the sheet at the start point is symmetrical in that it has pushed up the elastic upper sheet. The upper sheet and the microdome have a relatively high static friction coefficient, and the upper sheet and the inclined bottom surface have a small coefficient of friction, so that the upper sheet is not pulled in either of the horizontal directions. Even if it is pulled, there is no difference in reaching a close contact state entirely.
On the other hand, in the close contact type off type, it is clear that the parallelism is maintained as a relaxed state at the start of separation, and that the protrusions are inclined and then asymmetric when separated. In the membrane contact, the facing area of the conductive contact having a diameter of about 2 mm is about 3 square mm, but in the contact interface type, 80 + 80 square mm under the 19 mm * 19 mm button can be allocated to the X electrode and the Y electrode. This is because the tolerance digits for lateral deviation are essentially different. The wide range of motion of the deformable film requires a small force to push down from the top, which is why the contact interface type becomes a light touch.
Instead of pushing down the center of the flexible membrane, when a multi-spiral spring is formed by punching a single metal GND sheet as shown in FIG. 16, the sensitivity to tilt becomes more critical. However, by changing the X / Y electrodes from the two-pole arrangement to the four-pole arrangement as shown in FIG.
Further, as shown in FIG. 24, Orthodox metal springs can be used in combination at the adhesion interface. Since the force to bend the large-diameter film area is weak, the metal spring should be as weak as possible to allow the sink when the fingertip is placed. Since a relatively large voltage dip is selected, it is allowed to be pushed down diagonally. Metal springs are also allowed to be pushed down and stored in the liquid crystal panel.
As shown in FIG. 25, at the membrane contact, a thin polyester film is bent at a depth of h = 100 μm over a cylindrical space having a radius a1 = 2.5 mm, and the upper and lower electrode layers are in contact with each other. Curvature and numerical value in the case of the adhesion interface type in which the vertical component of the tension received by the circular film portion and the stretched ratio are pushed down at the depth of h = 100 μm at the center of the area about 3 times the diameter a2 = 8 mm. The difference can be seen in comparison.
The distortion rate extended in the radial direction is as shown in Formula 5, and the average value in the entire radial direction is reduced to about 1/10 by changing the radius of the circular area from a1 = 2.5 mm to a2 = 8 mm. . As the radial distribution of the proportion of elongation, the central portion extends more than the peripheral portion.
Next, even if the circular area grows at the same rate, the force to push down the stretched rope will make a difference as shown in Equation 6 if the principle of decreasing the length of the rope becomes longer. The vertical force required to push down is reduced to 30% by changing the circular area from 2.5 mm to 8 mm.
In practice, it does not stretch evenly, so the 30% ratio is even smaller. This indicates that in the contact interface type, the force when the G electrode is brought into close contact with and separated from the G electrode is negligibly small compared to the force with which the fingertip is TAPed.
As the distortion rate in the sense that the possibility that the aluminum vapor deposition layer or the printed conductive layer on the surface of the G electrode repeats expansion and contraction 10 million times to generate cracks, the reliability is ensured by the large area area. In both membrane contacts and contact interface type on-type, the cavity is a sealed space, and air compression stress must be taken into account, but this is the direction in which the air layer interferes with the upper and lower electrodes conducting at the membrane contact. In the on-type contact interface, it works in a direction that protects the contact and does not lead to conduction.
As shown in FIG. 25, there is a means for converting the tensile stress into a bending stress by using a spiral spring G electrode or for dispersing the tensile stress in the periphery. In the following description, even when typically illustrated with a spiral spring G electrode, this is not necessarily a condition, which means that there is no significant difference in operation in other forms of stress relaxation.

In the induction type shown in the upper part of FIG. 26, M coils are connected in cascade in each X line, and N coils are connected in cascade in each Y line, and one side of the X / Y line is connected to a common DC bias. The side is connected to the pin of the IC chip and is simply terminated with a pin ground capacitance C3 = ˜0.5 pF, and there is no bias resistance. In addition to this, floating capacitance to the ground is distributed on each X / Y line. The current induction principle of the coil is that a short-circuit current i is induced in the proximity of the cylindrical magnet, but when this is an open terminal, it is converted to a constant voltage source of i * jωL, and M coils are connected in cascade. However, the voltage is not attenuated as long as it is an open terminal. As shown in FIG. 27, since the cascade connection coil is an open terminal, current does not flow except through stray capacitance and termination capacitance, and the induced voltage of one generated coil is superimposed as it is. The voltage is transmitted to the pins of the IC chip and the voltage does not become 1 / M even if the number of coils M is large. For this reason, there is no problem with the voltage amplitude transmitted even with the 101 keyboard. Further, mutual induction between the X coil and the Y coil is not affected if the loop current flowing is small. On the other hand, L = 50 μH * M, in which M inductances are cascade-connected, is the sum of the inductance further multiplied by the bottom iron plate, the input capacitance C3 of the IC chip pin, and the equivalent distributed ground capacitance. LC peaking response in which the termination resistance R is infinite is obtained by terminating with a value equal to a load of about 4 pF in total. This resonance frequency is sufficiently high. As shown in FIG. 27, a positive pulse is generated during an attack and a negative pulse is generated during a release. The amount by which the magnitude of the power generation pulse reaching the pins of the IC chip varies somewhat depending on the position of the coil is not a problem size.
In the upper part of FIG. 26, an X coil is inserted into the X line and a Y coil is inserted into the Y line as a double spiral in parallel. Dividing into X / Y coils by the parallel processing induction type is called a separation coil type. On the other hand, the lower part of FIG. 26 shows a form in which the X line and the Y line are connected by one spiral coil at the intersection. This is called a differential coil format. The conditions are the same for both separate and differential types in that the spiral winding advances from the outer periphery to the inner periphery and requires a jumper or through-hole to exit. In the differential coil format, two-dimensional interleaving is performed in which the winding directions of coils at adjacent intersections in the matrix configuration are reversed. Corresponding to N * M intersections of N rows of X lines and M columns of Y lines, N * M comparators are required, but this is taken out as an open terminal voltage from one coil or short-circuited. There is a method of extracting the induced current by flowing it through a small resistance, and in both cases, the crosstalk between the intersections can be ignored. This point is clearer than the case of the X / Y separation type shown in the upper part. There are several methods for providing the bias in the differential coil type, but since all are simple, the explanation is omitted here. There is no superiority or inferiority in performance between the X / Y separation coil type and the differential coil type, but the X / Y separation coil type becomes a standard form in that a wire harness can be configured. All subsequent descriptions are represented by the X / Y separation coil format.

In the separated coil type inductive key matrix shown in the upper part of FIG. 26, when one button is pressed, the built-in magnet approaches two concentric spiral coils inserted in the X / Y line, and several mV to As shown in FIG. 28, when the X line and the Y line are read together, the polarity is reversed depending on how the coil is wound. The positive and negative pulses are input to the differential amplifier corresponding to the intersection address of the button. Although the magnitude of this power generation pulse is not much larger than the DC offset of the differential amplifier, the disturbance is canceled as a common mode and also canceled by interleaving. As a result, the disturbance is sufficiently suppressed, It is much larger than the thermal noise / shot noise in the differential amplifier. After the input DC offset is removed from the output of the amplifier by a diode clamp or the like, the address of the pressed button is specified by finding the maximum value.
On the X line / Y line, as shown in FIG. 28, the direction of the power generation pulse when one button is pressed and the direction of the power generation pulse when the adjacent button is pressed are reversed. The two-dimensional interleaved one is input to the differential amplifier. The number of pins of the IC chip is N + M, and the number of differential amplifiers is N * M, which is the same number as the conventional key matrix scan. Corresponding to the two-dimensional interleaving of the coil, the direction of magnetization of the magnet is also two-dimensionally interleaving, which reduces the magnetic force to the periphery and attracts iron clips and the like. Has been reduced.
The current consumption of the parallel processing inductive N * M differential amplifier and maximum value circuit is smaller than the current consumption of the key matrix scan circuit and N * M comparators. The interval to be cut becomes long, and since it is parallel analog processing, there is no response time delay.
The stroke of 1 mm to 2 mm and the speed of hitting the key are reflected as analog quantities in the magnitude of the power generation pulse and the interval between positive and negative pulses, which can be used to acquire comfort and certainty to press the key. Switching from the rubber contacts of the mold is possible without significant cost increase. This method is similar to the reed switch type shown in FIG. 2 (f) in terms of using magnets, but the reed switch type, which is expensive as an individual component, is eventually digital in terms of conducting contacts. Since it is on / off, analog information cannot be extracted, and above all, a key scan is required.
The inductive key matrix shown in the upper and lower sections of FIG. 26 can generally replace a PC keyboard, but can also be used to make the life of the remote control semi-permanent. The size and spacing of the rubber contact conduction type buttons can be replaced without changing the cost and up to tens of yen, and you can enjoy some additional functions.
Although it is possible to replace the capacitance switch type with an ATM keypad that requires higher reliability, the induction type is not affected by the charging of the fingertips or disturbance through the fingertips. It is more stable than the capacitance switch type, has a simple structure, and is low in cost.
As shown in FIG. 29, the induction type has a design in which the magnet pressed by the button hits the substrate and a design that penetrates the hole, the power generation waveform is different, and the lightness of the button is influenced. In the membrane contact, after the protrusion of the rubber dome reaches the film, further force and residence time are required to bend the film, but in the induction type when the cylindrical magnet comes into contact with the substrate, the power generation pulse is already There is a difference that it has occurred, and the process of pressing the fingertip against the bottom is no longer necessary. This difference becomes a critical turning point in key touch. In addition, the cylindrical magnet penetrating the hole makes it possible to give a composite meaning to how the button is pressed.
In the feed-through type using a double-sided board, if the front and back spiral coils are not connected in the same clockwise direction, but in a differential connection that turns clockwise and counterclockwise, the amount of change will change when the movable magnet is outside the board thickness. The induced currents cancel each other, but when they are between the thicknesses, the induced currents are added together, resulting in a steep response. Under a mechanism in which a fixed position where the fingertip is placed on the key is near this position, highly sensitive sensing without dead ends is performed. Due to the differential connection, the influence of disturbance is remarkably suppressed, and the wiring is twisted and the requirement for electromagnetic shielding is eased.
When the direct current clockwise equivalent circular current I on the outer periphery of the cylindrical magnet approaches a vertical distance d with a spiral coil having a larger diameter (or smaller diameter), the loop current i is left as a direct current in the coil. Although it flows around and tries to balance, i immediately returns to zero with a time constant τ obtained by dividing the inductance L of the coil by the sum of the copper loss and the radiation resistance. That is, the relationship of current induction between I and i, i = k * I, is equal to or higher than the cutoff frequency of 1 / 2πτ, and the proportionality constant k is referred to as a current induction coefficient. The notion of Faraday world voltage-to-voltage magnetic coupling coefficient k is not frequency-dependent, but what actually has a relation between cause and effect is drive current I versus induced current i. The open terminal voltage is only the induced current i multiplied by jωL in the frequency domain. The relationship between the vertical distance d between the cylindrical magnet and the induced current and the induction coefficient k is required but complicated, and here there is no relationship for simplicity, but a similar Gaussian distribution equation. Replace with and handle. FIG. 29 shows the relationship between the replaced d and k.
The truth of current induction that is simply confirmed by measurement is that if a curve is created by plotting the coupling coefficient k between the circular coil and the power generating coil, measured separately, against the distance d, the cylindrical magnet is equal to the distance d. If the speed is sufficiently high, the induced current when passing through the power generation coil matches the curve plotting the coupling coefficient. Observing this as an open terminal voltage agrees with the derivative of this curve (multiplied by jωL in the frequency domain). From this, it is clear that the result in the causal relationship is the induced current i, and that the change rate of the magnetic flux is an illusion. If the passage speed of the cylindrical magnet is subsequently slowed, it is measured that the induced current is smaller than this curve. This is because it is effective to return to zero with the time constant τ obtained by dividing the inductance L by the sum of the copper loss and the radiation resistance, and at that time, the peak is somewhat ahead. It turns out that Faraday's allegations are nowhere valid. In short, direct current tends to flow in the opposite direction of drive current with current induction, but it becomes zero immediately with the time constant τ, so the cylindrical magnet must be moved quickly, but if it is somewhat fast, it will be more than that The induced current does not increase even at speed. Since the open terminal voltage is multiplied by jωL, the voltage increases as it is moved faster. It seems that Faraday is 100% fictional in this way, but what Lenz said is a piece of legitimacy implying that induction that cancels direct current with direct current occurs.

An inductive and orthogonally connected spiral coil that forms an orthogonal matrix can be realized without waste on a printed circuit board of 2 layers / 4 layers. As shown in FIG. 30, the odd-numbered layer is a double spiral that turns clockwise from the outer periphery to the inner periphery, and moves to the even layer through the through hole. This is a right-handed double spiral. The X line and the Y line are a series connection of coils, and the clockwise rotation and the counterclockwise rotation are alternately reversed. However, since the direction of magnetization of the cylindrical magnet is also reversed, one X in which the coils are connected in series. A positive pulse is superimposed when the line is advanced to the right, and a negative pulse is superimposed when the line is advanced to the left. When going up the Y line, it becomes a negative pulse. Taking the X / Y difference is the same as the AND logic that specifies the address. As shown in FIG. 28, when the response of the stroke of the magnet at the intersection of the X line of the nth row and the Y line of the mth column is differentially input to the amplifier of the IC chip, twice that number of pulses are generated. Output, the remainder of the N * M outputs is 1x or zero. Comparing this, it can be seen that the button at the intersection of the nth row and the mth column has been pressed. The double spiral winding allows a simple through-hole connection instead of a jumper, and the number of layers is minimized. The number of windings is equivalently doubled, and the DC offset can be relatively reduced.
If the cylindrical magnet is regularly two-dimensionally interleaved as shown in FIG. 28, a magnetic attractive force that keeps the same plane as the net shown in FIG. The magnetic spins of adjacent atoms / molecules are two-dimensionally interleaved with the direction reversed, and automatic alignment occurs as a single layer in the horizontal plane that attracts, and the vertical direction stabilizes by forming a single layer with the same direction. Every time it goes, it turns out that it is thin, and it can maintain a stable state like Saturn's ring where the conclusion is postponed. When one induction-type cylindrical magnet is pushed down, the neighboring cylindrical magnet is also pushed down slightly, and as a result, a force that tries to maintain one plane works. Help them understand the mechanism of thinning the circle.
In the inductive type, it is preferable that the cylindrical magnet descends vertically toward the coil, but it is not required to be strict enough to provide a special guide to prevent tilting when the key is pressed. When the movable magnet is attached to one sheet and pushed down with the key above it, there will be no side shift, so it will work without problems even if the key is pushed from an angle, and even if you touch the next key, There is no problem because the key with the big action is selected. Even if you put your fingertips and the key sinks, you will be sensitive to changes from there, so there is no problem.
If you think again about a rounded sheet-like keyboard that has a need in the market but has not been realized in a practical form, if you open it up, it will be that the sheet itself will be flat to maintain its shape extending in the XYZ direction. It is done with the power to try and the repulsive force of the upper and lower layers. The latter can be realized by the repulsive force of two upper and lower magnet pairs. When the upper and lower sheets are two-dimensionally interleaved, the upper and lower layers are filled with the former, trying to become a single plane. The key attached to the upper surface creates a comfortable stroke of about 3 mm with magnetic repulsion. This meets the performance criteria of a sufficiently new standard keyboard, and the manufacturing cost is sufficiently low. This shape is shown in the lower part of FIG.

In order to construct the induction type X / Y coil with a single-sided printed circuit board, there is a jumper problem as in the case of a remote control board, so it is not always practical, and a double-sided board without a jumper is useful. However, a double-sided board that spreads over the entire area of the PC keyboard requires a certain amount of material costs, which can increase the cost compared to the case where a conductive keyboard currently forms a printed pattern on a flexible polyester film substrate. The capacitance switch uses a rigid board by pushing it. In the case of the induction type, it is better to print the single layer double spiral X / Y coil and wiring on the bottom iron plate and print the jumper on the additional layer. It can be said that it is advantageous. In particular, for a notebook PC keyboard that is desired to be lightened even by 1 mg, iron plate printing that also serves as an electromagnetic shield as shown in FIG. 32 is useful.
An alternative to this iron plate printed circuit is a method of laying the wire harness shown in FIG. 32 on the iron plate, and winding N + M thin copper wires, hardening them by self-fusion molding, and adding a bias wire Connect to IC chip with N + M + 1 pole connector. In any case, it will only weigh the button / magnet pair and copper wire, and will continue to operate even after spilling coffee, and can be removed and washed with water. The manufacturing cost of using a wire harness is not necessarily high. This wire harness composed of N + M copper wires is not suitable for the differential coil type in which the X line and the Y line must be separately connected with the N * M coils.

Up to now, there is a choice that has only a slight difference in key touch of mechanical contact / membrane type / pantograph type / capacitance switch type of PC keyboard and a standard that does not have an option of electrode touch of conductive rubber contact of remote control button. Was given. When a new close contact interface type is selected, the key touch is improved to be lighter than the previous choices, as much as it does not need to be conducted, but the possibility of a key touch completely different from these is inductive type Is given. The first is magnetic levitation shown in FIG. Magnetic levitation using the magnetic repulsive force of two magnets is the same as the first half of pressing down the rubber dome in that the upward drag increases as the button is pressed down. However, in magnetic levitation, there is a remarkable sense of stability when it is received and bounced back by soft landing, and there is no conducting contact that can be damaged by collision.
The second method is the same as shown in FIG. 33, and the configuration is reversed upside down, the electromagnetic shield iron plate is turned up, and a thin cylindrical magnet is adsorbed on the back side. In this method, the magnet is pulled away through the hole, and the operation is performed at the moment when the pressure of the fingertip meat exceeds the attraction force between the magnet and the iron plate, so the determination is made at the fastest speed. The more detailed process is that the magnetic adsorption between the iron plate and the cylindrical magnet is strong, so even if the TAP operation that depresses the operator's finger starts, the magnet does not move and the compression stress of the fingertip meat cannot overcome the adsorption force, but the force At the moment when the cylindrical magnet is slightly separated from the iron plate by reversing, the adsorption force between the iron plate and the magnet rapidly decays, but the fingertips do not move, only the pressed meat is restored, and the magnet moves away faster. At that time, the discrimination is already over, and the finger is thrown downwards following it. Since the separation speed of the cylindrical magnets at this time is large, power generation is performed as soon as they are separated. The key actually stops when the back side of the raised key is supported on the surface of the iron plate, but before that, the descent of the finger stops, so it does not actually stop and returns to the parabola. The collision sound that occurs when the magnet is again attracted to the iron plate is received as a confirmation of operation, or it is an option to mute with a cushioning material. The clarity of the process in which the magnets that have been attracted away from the iron plate lead to better button breaks. Not only does the relationship between the magnet and the iron plate turn upside down, but the human head is also turned upside down, and this mysterious key touch that you can't usually see is hard to come by until you actually see it.

The induction type generates a positive pulse when the permanent magnet approaches the X / Y coil, and generates a negative pulse when the permanent magnet is released. Speaking of key matrix scan key press / release input, in parallel processing, one or both of positive pulses and negative pulses generated by one reciprocating stroke is replaced by slicing. In key matrix scan, if you keep pressing the character button, the same character will be input continuously, so you have to be careful to press it once, This problem is the essence of key scanning and cannot be solved. On the other hand, in the parallel processing, even if the character button is kept pressed, only one character is normally inputted.
Since the shift key of the key scan type PC keyboard keeps pressing, the IC chip keeps detecting the scan pulse transmitted continuously. In the meantime, other keys on the same line cannot be used, so the procedure becomes complicated, and the number of pins required for the IC chip is more than N * M. On the other hand, in the induction / contact interface type PC keyboard which is a parallel processing, as shown in FIG. 34, the shift key is depressed by detecting the positive pulse when the shift key is depressed and seeing that the negative pulse does not continue. When the determination as a positive pulse continues, the negative pulse with the shift key released is sliced to recognize the end. This is not accompanied by a significant change, and the shift key is not separately wired. Other keys that are pressed while being pressed down and released will generate a pulse with no damage. In principle, in parallel processing, it is possible to perform separation determination that simultaneously grasps all N * M key states.
CapsLock and NumLock of the PC keyboard are toggle switches, and in the parallel processing, a toggle operation is configured in which the key is turned on by one reciprocation of the positive pulse and the negative pulse which are released by pressing the key and turned off by the next one reciprocation. In the volume increase / decrease button of the TV remote control, if the time of one reciprocation of the positive pulse and the negative pulse is short, it means increase / decrease of the volume of one step, and if it is long, it means increase / decrease of the continuous step. I can do things. In parallel processing, the number of pins required for the IC chip remains N * M.
The inductive type has the possibility of reading the speed of pressing a button and the depth of the stroke as analog information. The conventional conduction type key matrix scan has no room for it because only the scan pulse is transmitted. It is impossible to embed an analog quantity with a capacitance switch type.

As shown in FIG. 35, when a cylindrical magnet magnetized in the height direction is passed through a coil with the central axis aligned at a constant speed in the Z-axis direction, a pulse current flows through the short-circuited coil, and the coil Is opened, a zero-cross open terminal voltage pulse in a form obtained by differentiating the short-circuit pulse current (ie, multiplied by jωL) is generated. Neither the short-circuit induced current nor its open-circuit voltage is symmetrical on both sides of Z = 0, that is, t = 0 on the time axis, and the peak comes to the front by the time constant obtained by dividing the inductance L by the copper loss and the radiation resistance. Although this is a response change, it is described here as being symmetric for simplicity. If the direction of constant velocity motion is reversed and the horizontal axis is time coordinate instead of Z coordinate, the same waveform will be generated regardless of whether it moves from top to bottom or from bottom to top. The sign is reversed depending on whether is on the upper side or the lower side. Also, when the cylindrical magnet collides without going to the opposite side from the coil surface, for the sake of simplicity, it will first be understood as only one side of the current / voltage waveform that is assumed to be symmetrical.
If you can calculate how the short-circuit induced current and the open-circuit induced voltage appear in the design of parallel processing inductive keyboards and any other devices that use electromagnetic induction, the accuracy will be increased. There is no need to repeat the mistakes, and it is possible to achieve an optimization design from the beginning. Based on the idea that the induced voltage in the electromagnetic induction theory is proportional to the time change of the interlinkage magnetic flux (multiplying jω), a calculation result consistent with the reality cannot be obtained. It can be said that the cause-and-effect relationship was not correctly recognized. The cause is the current I, and the result is also the induced current i, and the two of the same dimension are connected by an unknown number k of a current induction relation with a flat frequency. Even if the relationship between the two is reversed and the cause is i and the result is I, the same coefficient k is established on the normalized basis. When a cylindrical magnet having an equivalent circumferential current of I is brought close to the distance d from the coil, the ambient current I suddenly appears in a step-like manner (position change or time change), and the current i is reversed in the coil. It tries to flow suddenly in steps. Normally, a stepped current does not flow suddenly through the coil, and the current of the lamp changes. However, when k is 1, the cylindrical magnet suddenly appears in a stepped shape, and a reverse stepped current flows in the coil. The sudden flow cancels out, and the induced current i can be stepped. In actuality, the lamp changes by k being less than 1, but this problem is out of the essence, so that the current can be changed in steps.
The reason why Faraday thought that the induced voltage was proportional to the frequency of magnetic flux change is that the result is not i for cause I, but induced voltage = jωL * i, but the original result i has frequency characteristics. I = k * I. However, the current induction coefficient k is attenuated as shown in FIG. 35 at a low frequency of ˜tens of Hz, and the direct current response becomes zero. This is because the inductance L of the coil is attenuated by a time constant τ obtained by dividing the inductance L of the coil by the sum of the copper loss and the radiation resistance, and is irrelevant to the Faraday shape of jωL * i = induced voltage. Based on such a correct relationship, how the i on the step changes when the vertical distance between the cylindrical magnet and the coil approaches can be easily obtained by sequential calculation using the approximate expression in FIG. Instead of adding a new step-like change to the value of the current remaining after the step change is attenuated by the time constant τ, an approximation that replaces the remaining current anew (that is, the induced current can change in a step-like manner) (Assuming not deviated from the essence). This is simply numerically calculated, resulting in the pulse response shown in FIG.
If FFT (Fast Fourier Transform) can be performed on the PC, a short-circuit induced current multiplied by jωτ / (1 + jωτ) is calculated in the frequency domain, and if the terminal is opened, a voltage multiplied by jωL is obtained. It was finally possible to calculate the non-computable amount of change in magnetic flux that had been confusing for 186 years since the electromagnetic induction law appeared. The Faraday has also been involved in the misalignment of the Q = CV button on the adhesion interface.
In the parallel type key matrix, the spiral coils may be used by being short-circuited, but the open terminal voltage may be used by being cascade-connected on the same X / Y line. In that case, if the waveform of the short circuit induction current is converted to the frequency domain, multiplied by jωL, and returned to the time domain again, the waveform of the open terminal voltage is obtained. In this way, it can be said that in parallel processing, it is not necessary to repeat trial and error since the optimum shape is known from the beginning for both the contact interface type and the induction type. Tokamak coils are no exception.
Here, it can be seen that the exact calculation method of the contact interface type and the exact calculation method of the induction type are the same in terms of the time constant τ, but this fills one pitfall in electromagnetics. . In other words, I thought that Faraday is proportional to the frequency or proportional to the rate of change of the magnetic flux. It is understood that the story was just not noticed. When a current response with a flat frequency is multiplied by jωL or 1 / jωC, a relative relationship with the propagation speed of light appears.
Whether the PC keyboard is externally connected to a desktop PC or a built-in type for a notebook PC, an iron plate is generally placed on the bottom for electromagnetic shielding and grounded. When the induction type is applied to a PC keyboard, it cannot be considered without the influence of an iron plate that lays on the bottom or covers the top. When the coil is laid on the iron plate, the short-circuit current induced by bringing the cylindrical magnet aligned with the central axis closer is larger than when the iron plate is not laid. The induced open terminal voltage is increased by the presence of the iron plate in the inductance L of jωL to be multiplied by this short circuit induction current, so that the increase in the open terminal voltage is larger than the increase in the short circuit current. Thus, a thin iron plate as an electromagnetic shield works effectively for multiplication of the detection voltage as an inductive basic principle. The coil and its wiring may be in close contact with the iron plate, creating a new form of copper wiring (POF) on the iron plate.
Similarly, as shown in FIG. 35, when the difference between the current responses generated by providing the coils on both sides of the double-sided PCB substrate having the thickness d is obtained, an S-shaped curve is obtained, and the linearity range of the zero cross response with the width of d become. This is used for acceleration sensors.
As shown in FIG. 36, the cylindrical magnet is pushed down from the upper side toward the inner side where the spiral coil is formed on the front side of the iron plate laid on the bottom against the drag given by means such as a rubber dome, and the position and time are related. It is on-type to generate a positive pulse that peaks at the lowest point when parabolic movement is made closer. On the other hand, when a cylindrical magnet is attracted inside the spiral coil formed on the back side of the iron plate, if it is pushed down through a small hole opened in the center, a negative pulse appears when leaving, and when parabolic motion is adsorbed again It is off-type that positive pulses appear symmetrically. The off-type is in the form of copper wiring (PUF) under the iron plate. In the off type, the conventional drag force of the rubber dome and the function of the hook spring are unnecessary because the attracting force of the cylindrical magnet plays its role.
Since neodymium magnets rust in the air, they are usually coated with nickel, but this can also be done with plastic. The plastic coating reduces the impact of the cylindrical magnet adsorbing on the iron plate and plays a role in adjusting the button breakage.

In the induction type, the current induction is the principle of the relationship between the cylindrical magnet and the power generation coil without the iron plate. In the cylindrical magnet, as the sum of the oriented electron spin currents, a clockwise ambient current I flows on the side surface, and no other internal current or end surface current flows. Since the magnitude thereof does not change with time, when a direct current I suddenly appears at the contact position from a separated position, a counterclockwise direct current i opposite to the direct current i suddenly flows through the coil, The change attenuates to zero with a time constant τ, which is a value obtained by dividing the inductance L of the loop coil by the sum of the copper loss of the coil and the radiation resistance, and thus becomes a pulse current instead of a direct current. The open terminal voltage is a form obtained by further differentiating this, and specifically, a pulse voltage of i * jωL is generated. If there is an iron plate, the inductance L increases, and the pulse voltage is multiplied for the same induced current i.
As shown in FIG. 36, when there is an iron plate that also serves as an electromagnetic shield immediately below the coil, when a cylindrical magnet having a clockwise equivalent circumferential current suddenly appears at the close contact position, the electron spin of the iron plate is magnetized and oriented. As a sum, the same clockwise magnetization current is induced in the iron plate. The clockwise magnetization current that appears as a step-like change that responds even as a DC component with an infinite time constant τ induces a counterclockwise current to the coil as a step-like change, and this indirect induced current is also the same copper Attenuates with a time constant τ on the line. The current induced directly in the coil by the sudden approach of the cylindrical magnet is counterclockwise, and the current indirectly induced via the magnetizing current of the iron plate is also counterclockwise.
On the other hand, if a cylindrical magnet with a clockwise circumferential current suddenly appears at the close contact position, a counterclockwise induced current (eddy current) flows through the iron plate and subtracts the clockwise magnetizing current flowing through the iron plate. Since the magnetizing current is excellent, including the fact that the time constant τ becomes small due to the large copper loss, the induced current can be ignored in the case of an iron plate.
Since the magnetizing current flowing in the iron plate is clockwise, which is the same as the clockwise current of the magnet, the magnet is attracted to the iron plate by the Lorentz force. One induced current is counterclockwise, so the coil receives a slight repulsive force from the magnet. Solving this and calling the copper plate diamagnetic is not an accurate view, but in the iron plate, the magnetizing current is much higher than the induced current, and in the copper plate, the copper atoms are not spin-oriented and the magnetizing current is zero. All that is equal is that only the reverse induced current is left behind.
Even when there is no guide for lowering the cylindrical magnet that is attracted to the iron plate in the form of inductive top-and-bottom reversal, it comes down relatively straight. Even if it falls diagonally, there is no significant effect on the power generation pulse. It is characterized by wide tolerance, such as selecting a key that has changed more greatly even if it touches an adjacent key. In order to relieve the initial drag to be peeled off, as shown in FIG. 36, when a metal spring is added, it approaches the conventional feel.

Considering that the cylindrical magnet approaches the coil quickly and that a step-like change in which a clockwise equivalent circumferential current suddenly appears at the approached position occurs, the step current tends to flow counterclockwise in the coil. Is immediately attenuated by a time constant τ obtained by dividing the inductance L by the sum of copper loss and radiation resistance. The fact that Faraday's induced voltage is estimated to be proportional to the time derivative of the interlinkage magnetic flux is a phenomenological interpretation of the mechanism in which attenuation due to this time constant does not occur at higher frequencies. In reality, the short-circuit induced current that is the result does not exceed the magnitude of the drive current that is the cause, and the explanation of the time variation of the magnetic flux that the open-circuit voltage will increase with frequency is as an understanding of the causal relationship Not appropriate. The direct cause-and-effect relationship is the drive current versus the short-circuit current, and the open-circuit voltage that is induced is the short-circuit current multiplied by jωL based on the Theo / Thevenin theorem.
When a cylindrical magnet with a clockwise equivalent ambient current flows away from the coil, the counterclockwise current is transiently added to the coil because the counterclockwise current is suddenly added. It decays immediately with a time constant τ divided by the sum of copper loss and radiation resistance. As shown in FIG. 37, if the coil is on the iron plate at this time, the magnetizing current flows in the iron plate stepwise or disappears stepwise, so that the appearance / retraction of the cylindrical magnet directly induces the coil. Is added to The reverse eddy current induced by the appearance / retraction of the cylindrical magnet in the iron plate is much smaller than the magnetizing current and can be ignored.
On the other hand, considering the magnetic attraction acting between the cylindrical magnet and the iron plate, as shown in FIG. 37, the clockwise ambient current of the cylindrical magnet is the clockwise magnetization-induced current (orbital electron spin orientation on the iron plate). ), And the two are attracted by Lorentz force. The time constant τ is infinite because of the spin orientation. When this is in close contact, it becomes a strong adsorption force, and by inserting a spacer between them, the difference in attractive force between the close contact and the separation can be reduced.
If there is a small hole for connecting the button on the iron plate and the cylindrical magnet on the back side of the iron plate with an arm as in the case of upside down, the adsorption power will decrease, but it will simply be Explaining that it is because the area is reduced is not a correct understanding, and the application problem from there cannot be solved. The clockwise circumferential current of the cylindrical magnet on the iron plate attempts to induce a clockwise magnetization current in a circular area whose circumference is projected onto the iron plate. This magnetizing current loop can be thought of as broken down into a small set of clockwise loops inside, but this small loop cancels out adjacent arcs, so the outer loop of the set determines the Lorentz force. ing. However, if there is a small hole in the center, a small loop is missing, and as a result, a counterclockwise magnetization current flows through the inner edge of the hole, and this magnetization current and the clockwise ambient current of the cylindrical magnet Lorenz repulsive force is generated between them. In other words, the explanation that the area of the facing part of iron is small is wrong.
With respect to the iron plate with a hole through which the cylindrical magnet can pass, the circular magnet fitted therein tends to be in the same horizontal position as the iron plate in the vertical direction. This is because the clockwise ambient current of the cylindrical magnet induces a clockwise magnetization current around the inner periphery of the iron plate hole, and the vertical component of the force attracted by the Lorentz force balances with gravity. However, when a cylindrical magnet is placed above the iron ring, the cylindrical magnet moves laterally from right above, and the cylindrical magnet moves onto the ring portion and equilibrates. At this time, it shifts from the center by the repulsive force from the counterclockwise magnetizing current flowing in the inner peripheral edge of the hole and the attracting force of the clockwise magnetizing current of the new ring portion and moves right above the ring portion. Thus, when magnetic attraction is considered as a permanent magnet current and a magnetizing current, a correct understanding can be obtained, but a correct result can never be obtained from the idea of magnetic field lines.
In other words, magnetic field lines do not exist essentially, and the movement of the drive current causes an induced current due to the movement of free electrons and a magnetizing current due to the spin orientation of orbital electrons, and the Lorentz force between the currents works. It is. Also, the time change (= alternating current) of the drive current is not essentially a line of magnetic force, but the appearance and exit of the current constitutes a step-like change current, which causes the coil inductance L to change. The fact that the Faraday is proportional to the frequency has just arrived today because it decays immediately with the time constant τ divided by the sum of copper loss and radiation resistance.

When a rubber dome often used for a PC keyboard is inserted between a button and a conductive contact, there is a bending point as shown in Fig. 38, and when it is passed, the drag decreases once, and when it is further pressed, the drag increases again. There is. When the protrusion of the rubber dome reaches the surface of the film substrate, the drag increases rapidly, reaches the point of conduction that has been waiting, and finally stops there. The click sound is an impact sound that the projection reaches the polyester film, and is not necessarily a signal that the conduction has occurred, and the sound is not conducted at all. Then, the click sound becomes an unpleasant unpleasant sound. Since the keyboard buttons are pressed against the cover with a stopper, the drag is not starting from zero, but already has an initial drag. Even if the fingertip is placed on the button, the button will not sink if the stress applied to the fingertip meat is smaller than the initial drag. It can be said that the threshold of the bending point is in line with the spirit of bending the lower film with the momentum accelerated by the ladder being removed.
In the capacitance switch type, a conical metal spring is inserted in addition to the rubber dome, and the linear drag of F = kX hook law is generated, but it is still compressed by being pushed by the cover at the stopper position. The initial drag is applied at the starting point. At the beginning of the stroke or at the blind touch home position, the force applied to the fingertip and the stress that compresses the fingertip's meat was balanced with the drag before the inflection point of the rubber dome. Then, the drag increases until the bending point, and after the drag decreases, the drag increases again at the inflection point and hits the surface of the film substrate. As for the mechanism of conducting, the descending speed at the moment when the membrane contact is conducted is infinitely zero, the contact is not subject to a large impact, but the interface is refreshed at every contact, and the speed is never insufficient. The critical condition is that it will not cause poor conduction. The close contact interface type does not have such a critical condition, and the induction type requires speed rather than position and force.
The essence of the thing is that when the fingertip is simply put on the button, the button sinks only slightly by the barrier, and when trying to strike with intention, the button is quickly lowered to the position of close contact conduction. Can be replaced with, for example, a general mechanism that keeps the button in place and a mechanism that automatically pulls the button at the lowest point upward.
As shown in FIG. 38, when the rubber dome is crushed from above, the height is reduced to about half. However, since the protrusion hits the bottom, the behavior in a large deformation that is pushed deeper is not considered. On the other hand, if the diameter of the metal coil spring becomes smaller toward the upper part, when the metal coil spring is crushed, the metal coil spring becomes as low as one metal wire diameter. Since the shape at the time of quenching after molding the coil spring determines the relaxed state, the proportional coefficient k of F = kX of the hook spring and the initial drag can be set separately.
Looking again at the difference in drag profile, as shown in Fig. 39, after the rubber dome protrusion bent at the end of the stroke touched the film, the part with a diameter of about 5 mm was bent with strong pressure from the remaining momentum. Thus, it can be said that all of the rubber dome including the 50 to 60 g barrier was present since the purpose of conducting the membrane contact was fulfilled. Even a light key touch is just comfort within this precondition. It is not the core part after the protrusion touches the film, but it is the part of the end of the end that makes the key touch for the run-up process for that purpose. In parallel processing, even if only the inductive type is taken as an example, there is no gap between the key touch and the core part, the focus on what should be done is narrowed, and the degree of freedom increases.
First of all, in the magnetic levitation type, the starting point is that the key is lifted by magnetic levitation with a stopper, and it does not sink only by placing the fingertip. This is a light key touch, but the drag increases when the key is depressed, so storing the key is a greater force than the starting point. Using a rubber dome instead of magnetic levitation improves over the membrane contact, but is not a good feel.
When the movable magnet is suspended in the position of a large hole in the iron plate, the drag is monotonously reduced beyond a small threshold, and detection is performed immediately after the starting point. On the other hand, in the top-and-bottom reversal type, the drag decreases monotonously, but there is no maximum value of the drag after the initial threshold value for peeling the cylindrical magnet.
Achieve a lighter, smaller threshold, 2 to 3 mm more comfortable stroke with a keyboard that is as thin as possible on a notebook PC with a magnetically levitated type. Considering this, when the liquid crystal panel is first opened, the button floats several mm and stops at the stopper, so that a short uncomfortable stroke of 1 mm that can be said to be unpleasant can be avoided. The floating button sinks when the liquid crystal panel is closed, but if the repulsive force does not leave a pressure mark on the surface of the liquid crystal panel, a power generation pulse can be obtained with a slight sinking from the raised position. That's it. This is a much lighter touch operation than the current PC keyboard has a threshold of 50-60 g. This can be said to be one form of key touch.

As shown in FIG. 39, it can be said that the threshold when the button is pushed down is optimally using the drag force when the thin cylindrical magnet is attracted to the iron plate. The adsorption force is suitably about 20 to 20 g, which is much smaller than the drag threshold of the conventional PC keyboard is 50 to 60 g. When the fingertip is placed on the button, the cylindrical magnet is not separated. .
At the moment when the magnet adsorbed on the iron plate is peeled off and separated, the compressive stress of the finger meat is released immediately after being peeled off. Since the lowering of the adsorption force is faster, the peeling is accelerated. Even if this acceleration does not occur, the time that the fingertip is naturally pushed down by about 10 mm is about 70 mS by TAP the finger, so a power generation pulse is generated at the moment when it is peeled off, and then the button Whether it descends and goes up via the lowest point and picks up again does not affect the power generation pulse.
The process of peeling a magnet, which can be said to be the kinematic ultimate form, also requires a precise understanding of what profile the power generation pulse occurs immediately after peeling. Not only that, even if the role of the power generation pulse and its detection ends immediately after the thin cylindrical magnet is peeled off, the biological touch is sufficient in that it is a subsequent process rather than holding the key touch. Verification is essential.

Previously, an iron plate was placed on the bottom of the keyboard for the purpose of serving both as an electromagnetic shield and stable gravity, and the GND connection to the PC main body via the USB connection was moved to the front side, so that it became the center of each character key. A small cylindrical hole is made and a plastic-coated thin cylindrical magnet is attracted to the back of the hole, but the strength of the attractive force, which is the initial drag, is reversed by the arm of a thin round bar through which the cylindrical magnet passes. Although it is weak enough to prevent the magnet from being peeled off from the iron plate when you put your finger on it even though it is connected to the button on the side, if the operator depresses the button with will, it will be easily pushed down, The magnetic attraction force still exceeds the total weight of the cylindrical magnet and the button, so when the finger returns upward, the cylindrical magnet is again attracted to the iron plate. An inductive X / Y coil is formed on the back surface of the iron plate to which the cylindrical magnet is attracted by a printing process, and the magnet generates a negative pulse or a positive pulse at the moment of separation or contact with the iron plate. The core process that can be said that the cylindrical magnet is peeled off from the iron plate is a very short time of about 20 mS. The induction type in the form of top-and-bottom reversal that gives the ultimate key touch consists of the simplest mechanism with no redundant elements in the reciprocating stroke process.
When the operator is willing to hit the button, the fingertip descends to become the most natural shape of about 10mm TAP, but since it has been detected when it begins, there is no room for discussion. The process goes to the point where the subsequent process is to replace the natural 10 mm TAP operation. With this natural TAP, the fingertip can push the button without resistance, or lift it a little and hit it quickly to break the initial drag. Even if it stops, it can be said that it maintains the property that is similar to the isochronism of the pendulum by the property that the muscle returns by itself. It is unacceptable for a notebook PC to hit something and stop at 1mm, but it is uncomfortable. On desktop PCs, this is about 2 mm or so, and the discomfort of dead ends is alleviated. If the dead end is 5 mm, the repetition of hitting the key as much as it stagnates at the bottom will be delayed. Recently, it has been considered that a depth of about 3 mm is the operating point where fatigue does not accumulate most, but many PC keyboards still have a stroke shallower than 3 mm. In the current keyboard, the key in the home position of ASDF / JKL: has a motion component that goes to hit the key by a reaction with the fingertip slightly lifted and hits the other key by the reaction to lift it. In the top-and-bottom reversal format, which can be said to be in line with the operating principle of humans, the length of the arm connecting the button and magnet can be selected freely within the range where the attractive force of the cylindrical magnet and iron plate remains. The maximum stroke can be chosen, but the actual stroke will be shorter. The second pulse is generated at the moment when the button + magnet drives the finger muscle back and re-adsorbs to the iron plate, and this pulse is also available if necessary.
That is, as shown in FIG. 40, the top-and-bottom reversal format has been turned on by forcibly passing the scan pulse by going to the lowest point on the keyboard so far without damaging the most natural fingertip TAP mode. In contrast, the determination is already completed immediately after the start of the TAP operation, as if the human intention was prefetched. There is a virtual keyboard in the air, and in order for the fingertip to press the button with the TAP operation, a fantasy threshold is required, but it can be said that the shape is expressed in the form of upside down.
This format is also suitable for PC keyboards that hit buttons, but it can also be applied to TV remote control that presses buttons at the same time with a large initial drag and lightness after that, and the magnet is separated. It can be said that it is suitable for the operation buttons of home appliances because it feels good. In order to further express the certainty, it is possible to distinguish the wrong pulse by simply double-checking the negative pulse when pressed and the positive pulse when returning. is there.
Since the coil and wiring formed in the printing process are in close contact with the iron plate on the back side, the shielding effect against external radiation is further strengthened, and the electromagnetic interference has already been applied to the interference from below in the notebook PC. It can be said that the shielding effect is sufficient. If there is still a problem with obstruction from below, the problem can be eliminated by sticking aluminum foil or the like to the bottom. This aluminum foil bag also has the meaning of preventing liquid intrusion from reaching the notebook PC body. Previous keyboards required both a bottom iron plate and a top plastic frame, but in the top-and-bottom reversal format, the top iron plate can be used as a frame, as shown in FIG. In addition, a small hole is made in the center of the button, and a groove for preventing rotation is cut.

In the induction type, the movable magnet that generates power must be prevented from attracting iron pieces such as clips at the surface of the button. In the top-and-bottom type, there is an iron plate with a small hole above the magnet, so the clip does not stick. As shown in FIG. 41, in any case of the induction type, two-dimensional interleaving is performed in which the magnetization direction of adjacent cylindrical magnets is reversed, so that the magnetic force does not reach the place away from the button surface. It has become. In the magnetic levitation type inductive type, the magnetic force of the upper and lower magnets magnetized in the opposite directions can be canceled just at the surface position of the button while obtaining an appropriate repulsive force. However, when combining the magnetic force of the power generation magnet and the rubber dome, in the configuration in which the iron plate is laid on the bottom, the two-dimensional interleaving of the power generation magnet is the only means for reducing the adsorption of iron pieces such as clips.

Among the external noises, the most prominent is a spatial electric field provided by a power line of AC 100V to 240V having a commercial frequency of 50 Hz / 60 Hz. In most cases, it is the remaining leakage electric field in which the positive amplitude and the negative amplitude cancel each other as parallel lines in the single-phase three-wire. In this leakage electric field space, for example, the potential difference between two conductors separated by several tens of centimeters is about 100 mVpp. It is most common that there is a human body and a keyboard in this spatial electric field, and there is an iron plate that also serves as a weight at the bottom of the keyboard, and this iron plate is grounded to the housing of the PC body through a USB cable It is a typical form. When a piece of metal is attached to the fingertip for measurement, the potential difference between the iron plate on the bottom of the keyboard and the metal piece is about tens of mVpp, but when another part of the human body touches the iron plate, the potential difference Becomes a fraction. However, this is not a general situation. The general state is that the metal piece conductor is at a height of several millimeters above the iron plate at the bottom of the keyboard, and the metal piece is shaken when there is a fingertip several millimeters apart. A wiring antenna having a long key matrix is connected to the metal piece, and the tip is grounded by a terminating resistor R on the Y line or terminated by a driving resistor r on the X line. In this state, it is excellent that the disturbance is suppressed because the long wiring runs around the height of several millimeters above the shield iron plate. The notebook PC keyboard has the same structure, except that there is a radiation source below the lower shield iron plate. On the other hand, in the contact interface type, the G electrode shields almost the entire surface without adding the number of layers, and it has a shield iron plate on the lower side, making it a complete shield. There is no problem even if the voltage pulse generated when hitting is small.
In the induction type, the power generation pulse amplitude is smaller than the generated pulse amplitude of the contact interface type, but one side of the wiring connecting the coils is connected to the voltage source, and the coil inductance is low impedance, so the noise is sufficiently suppressed. ing. In addition, the bottom or top iron plate shields external noise. At the same time, the disturbance is canceled out by nearly perfect two-dimensional interleaving. It can be said that the S / N obtained by combining these values is a better value than a conventional method such as a conduction type.
FIG. 42 illustrates margins for disturbance / thermal noise in the four systems of conduction type key scan format / capacitance switch type key scan format / contact interface parallel processing format / inductive parallel processing format. FIG. 42 (a) is an illustration of a conduction system with a three-pole structure. When the interelectrode capacitance reaches a maximum just before conduction, the intermediate electrode in which the disturbance is in a floating state is easily shaken via the fingertip. However, since the conduction type and the bias resistance R of the Y line are selected to be low, the termination voltage is lowered, while a sufficient margin can be taken compared with the case where the scan pulse is transmitted as it is. The driving resistance and stray capacitance must constitute the LPF.
Fig. 42 (b) is an illustration of a capacitance switch type system with X⇒Z⇒Y. The conical metal spring is connected to the constant voltage drive X electrode in an alternating manner, so that variable coupling is possible. While the scan pulse differentiated by the relatively large bias resistance R of the capacitor C and the Y line is received, the disturbance becomes a ratio of the magnitude that shakes the entire Y line. For this reason, on the printed circuit board, the Y line that crawls is sandwiched between the ground portions on both sides, and a thin ground pattern is drawn in a circle between the Y electrode immediately below the conical metal spring and the X electrode surrounding it. The drive resistance and stray capacitance must function as an LPF.
FIG. 42 (c) shows a pair of X electrodes hanging on one X line in the lower layer and N Y electrodes hanging on one Y line, and the upper G electrode is brought into close contact with each other. Thus, negative pulses are independently generated on the X and Y lines. Since the X line and Y line in the lower layer are shielded by the entire GND layer of the upper layer, little external noise is picked up. The capacitance at the time of adhesion is a capacitance sandwiching a very thin dielectric film, and the capacitance increases by the relative dielectric constant, which is sufficient for the sum of the capacitances hung off. If it is a resistor that pulls up the X line / Y line, the thermal noise is about 6 μVrms considering the LPF band determined by the CR time constant. If the coil has a pull-up built into the film substrate, the thermal noise Is negligible, and if the constant current of the differential amplifier is 0.1 μA, the shot noise is about 10 μVrms, which is sufficiently smaller than the generated negative pulse. This form operates normally even if the thin dielectric film is damaged and leads to conduction. Floating capacitance must be accepted.
FIG. 42 (d) is an illustration of an inductive system having an open coil structure, which can be understood by comparing with the reproduction of the magnetic tape by the magnetic head. A positive and negative pulse of several mVpp that is much larger than the voltage reproduced by the magnetic head appears, but the shot noise of the differential amplifier that consumes 101 bias currents of 0.1 μA and consumes about 10 μA has a bandwidth of 10 kHz. Then, since it is about 10 μVrms, there is a necessary and sufficient noise margin. It can be said that the maximum thing that the coil picks up is the radio wave of AM radio broadcasting, but careful verification is necessary near the transmitting antenna, but it is necessary and sufficient because it has an interleave configuration that alternately determines the direction of the coil In the inductive type that is canceled out, the induction picked up by the wiring can be canceled by twisting the return path connected in cascade. The stray capacitance causes LC peaking.

The induction-type pressing pulse power generation mechanism has the same structure as a voice coil of a speaker / microphone, and, as in the example of the PC keyboard shown in FIG. Speakers can be incorporated. Since the format of the TV broadcast has been digitized, it takes time to process the signal because it is digital, whether the remote control button was pressed correctly, whether the signal processing of the TV itself took time, data from the broadcast I can't tell if I'm downloading. If the remote control itself knows that the button has been pressed, you can choose a mode where the speaker built into the remote control can be used to make it sound fine. The drag of the button is added to the new lightness.
In the illustration, the position of the magnet and coil of a general repulsive magnetic circuit speaker is switched to form a moving magnet. However, if the weight of the cone paper and the air load are fixed, two cylindrical magnets are used. This is because the weight of the magnet is doubled, but the driving force is also doubled.
As shown in FIG. 44, the microphone / speaker is ideally formed by making a hole through which a cylindrical magnet passes through an iron plate, and using push-pull drive in which the audio signal current flows in the opposite direction to the coils provided on both sides. , Distortion is minimized. When the coil is provided only on one side of the iron plate, there is a zero cross point where the distortion is minimized where the magnet is submerged from the center of the iron plate. Distortion is reduced by using two thinner cylindrical magnets and making them repel each other and passing through a hole in an iron plate provided with a coil only on one side. This is the same diagram as the moving coil type speaker which is a repulsive magnetic circuit, but it is used in the moving magnet type. The speaker built into the board is a moving magnet with a fixed coil and a moving magnet of about 100 mg, which has a low distortion rate and high efficiency compared to conventional moving coils that vibrate diaphragms / edges made of tens of mg of paper. It is. If the weight of the neodymium magnet is about 100 mg, the weight of the entire moving part of a normal moving coil type speaker having a caliber of 16 cm is much lighter than that of about 15 g. Since it is a microphone / speaker built into the substrate, the cone paper is turned down. The iron plate can be thought of as an air-core coil replaced with a PCB substrate.
If you feel that the button has been operated with the sound or feel that the stroke has reached the board, or hear a click sound that switches on in the middle of the stroke, even if it brings comfort, the operation speed decreases. It becomes a factor to make. With an inductive PC keyboard, if necessary, you can select a mode that emits a fine electronic sound with a built-in speaker, and you can confirm that the button is being pressed with a fast continuous operation with a continuous sound flow.
The old idea that the location of the remote control that was hidden from whereabouts can be understood if a human responds with a voice is that if you place a call to a mobile phone terminal where you do not know where it is, you will vibrate even in manner mode Unlike what can be found, it has been considered a redundant cost that is unacceptable to all consumers. In the remote control using radio waves, bidirectional communication between the TV main unit and the remote control is built into the format, which allows two-way communication without any significant cost increase, and requires that the TV be operated to emit an electronic sound. You can make sure where you are. A remote control detector is easily available on the market, and if you place it on the TV remote control, when it is not known, whistling it will react and return light and sound, 500 yen You can buy it at: As a standard equipment that informs whereabouts when talking to a remote control that has been hidden somewhere for about 45 years since the remote control was used instead of an electronic tuner instead of a mechanical detent type TV tuner The idea of having this feature was just a joke because everyone thought it wasn't a matter of practicality and the cost of it was wasted. However, the story is different if the cost was -10 yen instead of 200 yen. A TV remote control usually has a remote control contact life shorter than that of the TV main body, but the purpose of the present invention is to make this a semi-permanent life without increasing the cost. If these two are combined and L & S (LOST & SOUND) and semi-permanent are realized by up to -10 yen, everyone wants to have this as a standard on TV. When you press the button lightly with the life of the remote control becoming permanent, a fine electronic sound will be useful for confirming that it has been pressed, the remote control will inform you of the location with the built-in speaker for it, and what will you answer if you call it? Even if a small question is taken, the tuning fork, which always resonates if it takes a long time, has been made for a long time, and the TV body only emits a single sound with the built-in speaker, realizing an increase in cost of about 10 yen.
That is, as shown in the lower part of FIG. 50, there is no cost for L & S, the TV speaker is necessary to watch the program, and the electronic sound of the remote control confirms that the button was pressed No preparation is required other than being useful to do. In order for the remote control to emit a single sound, it is necessary to oscillate positively, but this gives the remote control speaker the ability to hear the operation sound from the TV body. Specifically, if the parallel processing induction type is used to make the TV remote control semi-permanent, a cylindrical magnet and cone paper will be prepared. It will be an embedded speaker at a cost of several yen, and a resonant capacitor will be connected in parallel to the spiral coil to the IC chip. When connected and positive feedback is applied, a single sound that oscillates by self-excited oscillation is emitted. If the loop gain of the positive feedback is set to be close to 1, it becomes equal to the tuning fork, and the sound wave is thrown back in resonance with a single sound from the outside. Here, the cost of -10 yen is not incurred for L & S, and it is almost zero. There is no invention here, but if you have everything you need for a remote control and a TV, you can just say that there was a path to enable the L & S function from the beginning. .
On the other hand, if a three-dimensional acceleration sensor is embedded in the remote control board, it is possible to assign registered individual-specific movements to specific commands. This includes the power button.

In game machines, a vibrator is preferably incorporated, but other devices also have an advantage that a vibrator is incorporated in a control panel held by hand, or that an induction switch itself has a vibrator function. In the current product form, a surface-mounted vibrator unit is mounted on a substrate together with a surface-mounted speaker unit. This is a division of labor in the industry, but if an inductive keyboard / remote control / game console controller is already selected, the function of the speaker and vibrator should be added with a small additional material cost. And this makes it possible to design freely and organically.
As shown in FIG. 44, in the vibrator incorporated in the substrate, the coil is fixed on the substrate, the case that touches the hand is connected to the movable magnet by the arm, and the momentum of action = reaction is exchanged. In some cases, a part of the substrate is vibrated in the direction opposite to the magnet with respect to the magnet. The human body feels the vibration of the case, and the vibration of the board exposes the mounted components to vibration testing. When two cylindrical magnets are reciprocated in the opposite direction with respect to one drive coil to vibrate two places of the case, the intermediate substrate becomes a neutral point and vibration is minimized.
It can be said that conventional acceleration sensors typically appear in a change in capacitance by, for example, supplying a sine wave signal and measuring a voltage / current ratio as an impedance. In addition to this form, there was a method for measuring the voltage at which acceleration occurred in a piezo element, but it was rarely used in practice. On the other hand, in parallel processing, the induction type can conveniently measure the acceleration in the form of F = mα using the mass of the magnet, and the acceleration vector is detected from the X / Y / Z triaxial power generation pulses. The In order to efficiently perform this detection, a penetrating magnet type in which a hole is formed in the central portion of the spiral coil provided on both sides of the PCB double-sided board and a cylindrical magnet is passed is useful. For one coil, the power generation pulse changes in the same manner with respect to time when the cylindrical magnet zero-crosses from above to below and when it crosses from below to above. That is, it can be used first that the cylindrical magnet is at the upper side and goes downward, and the change at the lower side and going downward is reversed. When the coils on both sides are differentially connected, differential detection with improved sensitivity, zero cross position accuracy, and linearity can be performed. The current pulse integrating the acceleration is proportional to the speed. On the other hand, the quasi-static attitude detection that tells which direction the screen is facing with respect to the earth's gravity cannot be strictly done by the penetrating magnet, but the change when the portrait orientation is turned to the landscape orientation is the acceleration. If the pulse generated by the differential connection coil of the sensor coil is detected separately, it can be accurately determined if the change is not slow. This can be used for game consoles and smartphones as a matter of course, but even with a TV remote control, only the amplifier for acceleration sensor is used from the key scan that puts current through the detection amplifier of the entire N * M key matrix. In order to increase the interval of battery replacement when the remote controller is not used very much, the current is also passed to other amplifiers after detecting the acceleration when the operator holds the remote control in hand. Can save you the trouble of pressing the power button on the remote control.

The ultimate key touch of a PC keyboard is based on the premise that, as shown in FIG. 45, as shown in FIG. As a result, it was far from the ultimate form that should have been, and the boundaries that were reached through the trial of the capacitive switch type were not much different from the previous ones. However, only from the new viewpoint of parallel processing, the ultimate form can be seen, and at the same time, the realization method will be accompanied.
FIG. 46 illustrates a contact interface type and an induction type in which discrimination is completed at the start point of the stroke. An example of the induction type is a top-and-bottom reversal type. However, in the case of one contact interface type, a punched metal spring GND sheet will be described. A multi-spiral metal spring is formed by punching a single metal sheet, and the spring is in a relaxed state. If the spring is in a state where it protrudes upward, it can be prevented from sinking with the fingertip placed on the key, with an initial drag as an off-type that pushes down from where the spring is flush. In this form, the starting point is the state in which the metal spring is pressed against the X / Y electrode on the polyester film thereon and is in close contact therewith. When the key is struck in this state, the spring sinks by 1 to 2 mm and returns, but a positive pulse is generated by the G electrode separated at the base of the parabolic locus, and the determination is made as it is. The metal spring sheet is required to have a durability of 50 million times, and serves as an electromagnetic shield in place of the conventional iron plate. This form is the simplest structure among PC keyboards, along with the inductive top-and-bottom form, but its function embodies the ultimate key touch, the simplest form is the most elaborate and advanced It can be said that it is in line with the cosmic principle of playing a role. The top and bottom reversal type has a key response, but it takes more time to return, and the punching spring returns faster, but the response when you hit the key is weaker.
The role of the punched metal spring shown in FIG. 46 is to make the key come back after pushing down ~ 3mm after detecting the generated pulse to resemble the natural 10mm movement of shadow typing TAP. is there.
The combination of the conventional capacitance switch type rubber dome with a monotonically increasing positive differential drag of a conical metal spring and a negative differential drag that turns from increasing to decreasing Since it can be created even by ingenuity, the claim that the capacitance switch type metal spring brings comfort does not pass. The difference between the capacitance switch type and the membrane type / pantograph type, where the stroke position of the button is displaced parabolically with respect to time and detected at the lowest point, is detected before the lowest point is reached. It is a point that can be set to complete. However, there are cases where it is assumed that this is comfortable from the reflection of the result such as few typing mistakes, but the operator does not know at which depth in the process of pressing the button, There is only a psychology that assesses that consumers are comfortable or less fatigued that the threshold or drag is simply small and does not need to reach a bottom.
Regardless of the evaluation result of the capacitance switch type, some evaluation results have been reported that it is more comfortable regardless of the method to be detected before bottoming out at the lowest point of the button. . When this is chewed, it is understood that the three elements of lightness, no bottom feeling, and no bending / inflection point of the drag provide comfort / lightness. Considering the rubber dome and the membrane contact point as a starting point, a barrier of 50g to 60g is necessary as a run-up that must strike the small contact point to make it conductive, and at the same time the wrong key It can be said that the consumer has become accustomed to what was a barrier for not pushing up.
The rubber dome starts from the initial drag pushed down by the stopper, and there is a bending barrier of 50g to 60g in the middle lowered, but the button is zero if the initial drag overcomes the compressive stress of the fingertip meat Displacement, the middle barrier has no direct meaning. The indirect meaning of the barrier is to save power as a run-up period that bends the membrane film. Capacitance switch type / contact interface type / induction type does not require a run-up period, and it is only necessary to break the initial drag with the struck fingertip. The most important point is to return to a fixed position in the shortest time, and if it stays at the lowest point, it will be delayed by that much. The second most important thing is to push down close to 10 mm of the TAP operation, but this is physically impossible with a 19 mm square button interval, so it is mechanically limited to ˜3 mm, and becomes shallower as it becomes uncomfortable. TAP action does not result in tendonitis, but tendonitis is caused by being held down to ˜2 mm or by applying more force than TAP. The next important point is that the 10 mm TAP stroke becomes ~ 3 mm because it receives a drag on the way. It decelerates by pushing down and shortens the stroke, but it takes time, and it takes time to return and accelerate. It is a buffering process where the time is the same and only the stroke is shortened when combined, and when you try it out, it feels like you landed on a trampoline and jumped up again, and it feels comfortable rather than uncomfortable. This lacks a response whether or not the key is activated, but there is a sound that exceeds the initial drag when the key is struck, and since the detection ends in fact at that moment, it is a genuine operating sound. If there is no interval, it may be replaced with a faint electronic sound.
Although it is confirmed by actual trial that the initial barrier in this viewpoint may be 20 g or less, what was the reason for the 50 to 60 g inflection point barrier for the first time? It seems doubtful. The inductive top-and-bottom reversal type / inductive magnetic levitation / contact interface type off-type matches the weak initial drag close to the ideal.
The ultimate key touch of a PC keyboard is that there is no reason that the stroke of a notebook PC may be shallower than that of a desktop PC. It's just that you want to make your notebook PC thinner. It is desirable that the desktop PC and the notebook PC have the same key interval, that is, 19 mm, the same stroke around 2 mm, and a touch that does not change regardless of which key is used. For this purpose, for example, both of them are unified into an induction-type top-and-bottom reversal format, and when viewed as a notebook PC, when the liquid crystal panel is opened, the keys are lifted to the same height as the keyboard of the desktop PC by the magnet's adsorption force. That's all you need to do.
Even if the ultimate key touch can be achieved, it is unbearable to be 8 to 10 times more expensive than conventional keyboards, and it is uncomfortable following tendonitis. Inductive type top-and-bottom type keyboards are similar to or less than the manufacturing cost of conventional standard keyboards. It is uncomfortable that the keyboard breaks down earlier than the PC body. The life of a coherent interface or inductive keyboard is virtually semi-permanent.
The method of using two magnets and magnetically levitating the upper movable magnet with the bottom fixed magnet is different from the conventional keyboard sensation and the continuity-type top-and-bottom reversal form, but it rises strangely against the fingers. Soft button that comes back just before the close contact that never comes to the bottom, the detection does not occur by pressing, the sense of security when it becomes a habit by learning that only speed is meaningful when used, It can be said that the proportion of people who feel comfortable as a combination is not low. However, the form of reversal seems to be simpler and more rational.
The top-and-bottom reversal form, in which the threshold appears immediately after the key is depressed, is punched to the lowest point when you feel a slight resistance when you tap from the top of the key, but without the intention of TAP, just use your fingertip as a key It just doesn't move. With the punching out of TAP, the finger and the magnet, which originally descended about 10 mm, reciprocated with the bottom point of about 2 mm due to the impact, but the button and magnet reached the mechanical bottom point and returned, Since it is faster than the fingertip return, push the fingertip slightly.
The feeling that this process gives to the operator is completely different from that of the membrane type, and there is no sense of discomfort. When the tip of the heel, which is lowered by 10 mm with TAP, breaks the ultra-thin shoji paper every time, decelerates and falls by about 2 mm and flips without bottoming, it just holds a light resistance, and the pressed button + magnet sticks together. . The round-trip time at this time is faster than the membrane type, even if no force is applied. If you look at it slowly, you can't input it, so you just need to tap lightly. If you try to tap lightly and lightly from the top, you can see that it is being entered, and you can immediately realize that the button does not have to be down.

Assuming that input is done by blind touch on keyboards in various languages around the world, the ASDF / JKL: key, which is the fixed position of 8 fingers, is pushed down at that position, otherwise Because the key is struck at the end of the movement to move the fingertip to the target key by mobilizing the wrist and arm, the key touch setting divided into two groups, for example, the threshold of the eight keys at the home position is the fingertip It is preferable to consider that the remaining keys are divided into thresholds of about 14 g and the keys other than the home position are pressed obliquely at a distance of about 20 g, which does not decrease.
Or conversely, even if the threshold value of the eight keys of the home position is set to about 14 g, even if the fingertip is lowered, it does not affect the judgment, the judgment is made directly on the speed at which the key is pushed down, and the remaining keys go to strike. That is, since the fingertip is pushed down by the reaction of the muscle on the opposite side of the muscle that lifts the fingertip up once, the idea that the threshold may be about 14 g is also valid.
Proceeding this further, placing the fingertip on the ASDF, moving only the middle finger from the D position to the E position and pushing the fingertip down, the ASF buttons will move a little bit back and forth. By taking in as analog information, for example, take the largest of the four button movements that exceeded the threshold of 20g / 14g, but by grasping the overall pattern, etc. Learning and optimizing individual-specific habits is an option and can be placed in a way that is compatible with future recognition system standardization.
If there is no problem in the parallel processing that the ASDF and JKL: home position keys are moved by the blind touch, the meaning of the threshold value is lost. If shadow typing is imagined, the operator is conscious of one finger, and it is meaningless how other fingers are involved. In the close contact interface type and induction type parallel processing, all of these situations are reflected, and the determination in accordance with the operator's intention can be given immediately. The characteristics of the rubber dome as shown in FIG. 38, the characteristics of the metal spring, the slicing before the key stops, etc. are all meaningless. In short, it is enough to have a slight sense of resistance and touch sound that can be seen as a key touch when considering shadow typing.
Still, if you can choose to have the least amount of play as in the form of upside down, you won't waste your momentum. Exceeding the threshold is determined by the compressive stress of the fingertip meat, and if the button does not need to be lowered slightly, the first requirement is that if the upward drag decreases monotonically, it will lead to comfort. The second requirement is that both the fingertips and the buttons advance independently to the lowest point, entangled so that the movements move back and forth when returning to a fixed position through a parabolic time and stroke trajectory. The shape with less is more comfortable and the character input at the maximum speed, and it can be said that there is less fatigue.
It can be said that the membrane type keyboard is much more comfortable than mechanical typewriters, but in occupations where you want to sit down in front of a PC, hit the keyboard for a long time, input as fast as possible, and reduce the rate of operation errors. Everyone thinks that current keyboards do not have sufficient performance, including the recent capacitive switch type. This is because the reality is different from the ideal imaginary key touch that performs shadow typing in the air. The key touch is light and the button is turned on at a shallow stroke where the button goes to the lowest point, and the button is pressed and functioned with an electronic sound that replaces the feeling of bending of the rubber dome when the button arrives. It can be confirmed as it is, there is no delay or uncertainty due to scanning, the existing input of the 101 keyboard can be expanded extensively by adding voice recognition, etc., and the individual characteristics of the operator are learned to reach the ultimate operability In order to achieve this, there is a new demand for generalization, and for that purpose, it does not cost much, but the top-and-bottom-reversed type almost satisfies these conditions, and the magnetic levitation-type inductive type In accordance with this, a large improvement can be expected even in the contact interface type.
The key touch that was most comfortable in the era of mechanical typewriters and the key touch that is most comfortable as a current PC keyboard product have different meanings. Even with the same PC keyboard, the key touch achieved within the key scan constraints of membrane type / pantograph type / capacitance switch type and the ultimate key touch that can be considered in parallel processing by removing that limitation also have different meanings. It is.
In any case, it is necessary to ask again what is common comfort for all. Up to now, it has been replaced by the sense of resistance that the stroke has hit or the stroke has become lighter beyond the inflection point to recognize whether or not the button has been pressed down. The biggest problem is the stress of making the strokes shallower with the pantograph because it is different from the natural human TAP of 10mm. In the conduction type, a certain amount of pressure needs to be applied to the contact point, but in the contact interface type and the induction type, this is not necessary, and the touch can be reduced as much as possible.
When opening the notebook PC display, the stored key will rise and become a more comfortable deep stroke. In the induction type of the top-and-bottom reversal type, the threshold of the eight buttons of ASDF and JKL: The threshold of other buttons is 14 g. When the display is closed, the button sinks and is stored. Therefore, if the repulsive force is half 10 g / 7 g, the total repulsive force of the 101 button is 731 g. Since this is about 5 kg in the current keyboard, the compression stress is about 1/7, and there is no problem with respect to the point that the key compression mark does not remain on the surface of the display. When the LCD panel surface is closed, the key sinks. When an impact is applied to the notebook PC itself, not only the LCD panel but also the lower structure plays the role of an ideal cushioning material to reduce the impact. become. If the total weight of the cylindrical magnet, the button, and the arm connecting it is 200 mg, the reciprocating time of a parabola with a height of 2 mm can be calculated as follows: 8 and about 8 mS, it can be seen that it can sufficiently follow the round-trip time 150 mS of the TAP operation of a human finger.
Again, by adding a slight electronic sound flow with a speaker embedded in the board, it can be confirmed that if the flow is not interrupted, character input will be performed without delay, so far the bottom touch and bottom You can replace the comfort of confirming with a single click to reach the point.
In parallel processing, analog quantities are handled, so it can be read and reflected in the establishment of nimbleness. The ultimate nimbleness is that the majority of people increase input speed, reduce typing mistakes, and do not get tired even if they hit a key for a long time, but on the other hand, if they need something suitable for each individual However, in order to realize this, it is required to have a learning function for tuning and storing the optimum operating point in the keyboard in parallel processing that is an analog amount.
It should be noted that the key touch requirement of the remote controller / game machine controller is different from the key touch of the PC keyboard. While holding the remote control with your right hand, use your thumb to press it down to the bottom with a shallow stroke of about 1mm. It will always be a strong reminder, not a light TAP. In addition, the remote control for TV does not keep in mind that the response of the buttons will deteriorate over time, and in digital broadcasting there is a delay in the response on the TV body side, so it is not known whether it was pressed, but it can be pressed hard anyway. I'm on my body. When the TV remote control is replaced with a semi-permanent model, the experience of not having to press the rubber button any longer will be established over several years for consumers. The speaker embedded in the remote control board makes it possible to make sure that the button has been pressed, avoiding a strong double press.
In game consoles, the absolute requirement is to eliminate the compliment that sometime the rubber contact switch becomes defective in continuity, and to give priority to being released from the delay in response to wait for the scan pulse and its uncertainty. . Rubber contacts, which have been easy so far but have no other alternative, are the most unsuitable methods for game console controllers. A suitable method would be an induction type in the form of upside down. Thereby, the embedding of the analog quantity is made available in an expansive manner as speed / stroke length / positive and negative pulse intervals. Further, a stroke longer than 1 to 2 mm is preferable, and the induction type easily realizes it. The current rubber contact is a heavy button, but it does not cause any trouble by making it a light inductive type. Embedding vibrators, speakers and microphones adds comfort.

In a conduction type that is not a potentiometer, the analog amount of stroke cannot be reflected other than the length of the conduction section, and is affected by chattering. In the capacitance switch type, the change in capacitance value suddenly increases when approaching, but the reciprocal impedance is linearly proportional to the distance d, so it seems to be reflected evenly in the analog amount. In reality, however, the PP value at which the scan pulse is transmitted as a CR differential waveform is irrelevant to the change in the capacitance value, and the drive resistance r is increased to produce a difference in size, but this also causes a large attenuation at the same time. As a result, the design becomes critical, and in the end, the embedding of the analog quantity is not useful. In short, the detection section of the capacitance switch type is just a little longer than the conduction type, and it is not an appropriate expression to suddenly call it analog detection. The essence of this difficulty is that the capacitive switch type uses scan pulses.
On the other hand, the generated pulses read in the contact interface type and inductive type reflect the stroke process without being rounded down as an analog quantity. By reading this appropriately, the input speed increases and the number of typing errors decreases. Can be used to make it comfortable and fatigue-free. Since it is parallel processing, there is no scan pulse and there is no problem of time accuracy from the beginning.
The operator can enter keystrokes with the consciousness of striking strongly, the consciousness of striking quickly, and the consciousness of pressing down deeply, but the consciousness of strongly pressing does not necessarily work effectively, and the speed may decrease, and the result is that the user presses deeply. Become. Therefore, it is usually required to strike at a fixed TAP speed, but the input pitch can be increased so that the interval is continuous without any gap. The strength with which you hit the key will be deeper, and the character input pitch will decrease as long as the button stops at the lowest point and stays for a while.
As shown in FIG. 47, in the contact interface type and the induction type, positive / negative pulses are generated one after another near the lowest point of the parabolic locus, and positive and negative pulses are generated separately at the base of the parabolic locus. You can choose either. In the former case, the generated pulse increases when the velocity when colliding with the bottom is large, and the generated pulse is constant but the amplitude is small when leaving the lowest point. In the latter case, the pulse generated at the base of the parabolic trajectory at the beginning of the stroke depends on the speed at which the key is struck, but the pulse generated when returning to the original position causes the key to interfere with the fingertip that returns naturally. If not, the size is constant. Of these, the pulse size depends on the position of the key and the individual difference, as one of the three except for the generated pulse when leaving the lowest point is selected.
While setting the optimal slice level, the measurement of the interval between two generated pulses is also added to the discrimination, but the method of setting the optimal slice level depending on the key position and individual differences individually with a dedicated program is It is not suitable for everyone, and it is appropriate to optimize and memorize it with the learning function.
The induction type is more easily reflected in the analog amount of the stroke than the contact interface type. In order to use the true analog quantity that can be obtained only in the close contact interface type and the induction type, it is necessary to secure a good S / N ratio, avoid the drag non-linearity and the collision process as much as possible, take away the opportunity of chattering, etc. It is valid. There is no problem with the thermal noise of the amplifier, so attention must be paid to the electromagnetic shield for suppressing disturbance noise.
As shown in FIG. 47, compared to the induction type that moves a cylindrical magnet having a diameter of several millimeters, the close contact interface type that captures a separation of about 100 μm from the close contact incorporates the bending of the film in one section of the stroke of about 3 mm. For this purpose, a rubber dome or a metal spring and a film whose bending is displaced in two steps are used. This is to prevent bottoming from becoming uncomfortable for human TAP operation. The difference from membrane contacts is that the drag of the film is extremely small.

As shown in FIG. 48, using the parallel processing coherent interface type / induction type keyboard that can handle analog quantities, as shown in FIG. 48, the slice level of the generated pulse at the time of attack and the generated pulse at the time of return are also set. It is an analog quantity that is usually obtained that the setting of the slice level at the time of participation in the determination is made programmable and the interval between two pulses is also measured. In the conventional conduction type / capacitance switch type, most of the analog information of the stroke was discarded, but only a distorted response was obtained. In the contact interface type / induction type, the response reflecting the entire stroke was obtained. Therefore, detailed data processing becomes possible with the difference of cloud mud.
Basically, when you place your four fingertips, excluding the left and right hand thumbs, at the home position and hit the button with any of the ten fingertips with your thumbs, The remaining four fingertips on the same hand are likely to have some generation / power generation pulse. Because it is clear that a light touch is the source of the speed of input, less pressing errors, and less fatigue, the more you pursue it, the lesser movement of the other fingers This is because there is a high possibility that the response appears in the response. For the individual operator, if the wrinkle is constant and fixed, it is almost decided how the selected finger is accompanied by the movement of the other four fingers, and it is really concluded that the finger was selected If the patterns of the remaining four fingers are also recognized, the determination is reliable. Such an idea is not possible in the key scan format, and can be achieved only by the close contact interface type / inductive type parallel processing that grasps all N * M buttons at the same time. As shown in FIG. 48, if the ASDF / JKL: key at the home position does not cross the same X line / Y line on the matrix, the separation becomes simpler.
When performing such parallel processing determination work, the role of the PC main body is only to store the 4 + 4 bit attack / return slice level. In other words, if it is sufficient to have an 808-bit memory, the cost and the occupation time of communication through the USB cable can be ignored. The calculation is performed exclusively with the IC chip of the keyboard, and the measurement of the interval between two pulses is processed as reference information only in the IC chip. If the cost of this IC chip is only 20 yen higher with the standard function of the learning function, The keyboard that can be purchased as standard for about 2000 yen can be bought for 20000-30000 yen just by saying that it is comfortable, and as a result, there is no change in the current situation. In other words, it is not necessary to press each key correctly, and if each operator hits the key with 101 combinations of left and right fingers while maintaining the individuality, it is much faster than before. It is nimble, can be used for character input, with no judgment errors, and less fatigue. In addition, anyone can become a keystroke specialist of his own style by fixing a unique combination of movements, without necessarily fitting strictly into the blind touch type. If the manufacturing cost of the adhesion interface type / induction type is reduced, the cost increase of 20 yen that gives the learning function is absorbed, and it can be purchased as a standard product for 1800 yen instead of 2000 yen of the popular product so far. Is semi-permanent. Rather than practicing blind touches, My Keyboard remembers your own way of doing things like the growth of a pet robot that transforms favorably.
It is presumed that the back space key was used to make an error due to insufficient judgment on the keyboard side, not the operator's mistake. Following this, when placing the four fingers of the left hand on the ASDF and hitting the E key, ASDF reacts slightly and E reacts later. If it is stored as a car drive recorder, the information contained in the response of the entire five fingers can be used to conclude that E was really pressed. The learning function continuously supports the transition in which the operator raises the skill of irregular input in such a procedure. It is truly an era of individuality, not individual differences. This also means that a personalized keyboard can be used only by operators who are stored and registered as patterns.
The entire drag, including the membrane / capacitance switch type threshold, is large, and it has a learning method for the contact interface type / inductive type that has been replaced. If the keyboard touch does not lightly bottom out, the gap will result in a sense of incongruity. In order to avoid this, a bias amount that fills the step of discomfort is given as an initial setting, and it is gradually removed, and after that, it is guided in the right direction by the learning function, and the performance of key input is mainly Improves sharpness on the keyboard side. This can be said that a wise horse is inevitably accustomed to a person. The problem with getting used to it is that typing speed will drop when you go back to the old keyboard, but there will be more typing mistakes. However, if the lifetime is semi-permanent, it will not go back to the old days.
In shadow typing, in which a virtual key is selected and hit, there is no individual difference, and anyone can reciprocate a stroke of about 10 mm in a time of about 150 mS. Since 10 mm is too deep, a reactive drag is required to reduce it to a few mm, but this drag should not lengthen the round trip time with isochronism. Individual differences are caused by the fact that a threshold value of 50 to 60 g is attached. If the threshold value is 20 g or less and the bottom does not reach the bottom, the original TAP operation is easily reflected as it is. Less. A person can tap the key lightly, but if the key is pressed hard or the consciousness of pressing down is activated, the speed will slow down. Therefore, the individual difference to be learned is the distribution of the magnitudes and intervals of positive and negative pulses generated / generated from each key as a result of continuous character input without any gaps. Finding out, choosing the slice level and converging the discrimination with the time window to the best fit for the individual. If you are conscious of pressing the key, shallow strokes are uncomfortable, and if you don't feel the bottom, you'll be worried about whether or not you pressed. If the key is detected simply by TAP, the stroke depth is irrelevant, and if the key is felt to sink, then the input is complete. Or if the electronic sound is not interrupted, correct input continues. The natural TAP lowers the fingertip by about 10 mm, but the key is to receive it and soften it to about 2 to 3 mm, and the conduction type does not work with this. The problem is whether the operator feels comfortable, the long lasting to be depressed deeply is painful and it is uncomfortable to hit at 1 mm trying to depress. The movement is TAP, and if it is felt that it is blocked by the key, the input is over. At the same time, if a soft sound is heard from the touch of the surface, it will be pleasant and you will not get tired even if you keep hitting for a long time. I can say that. However, it is uncomfortable that characters appearing on the screen come out late for digital processing.

Since the keyboard for PC 101 has a strictly determined role for buttons because of its versatility, the input operation speed in this sense is limited. The time required for input is reduced by a method such as shortening an input procedure by using a combination of character strings as a cue for a shortcut or registering a word. In addition to this, there is still a means for discriminating the cue of the voice uttered by the operator through a microphone incorporated in the PC keyboard, which does not necessarily impair the versatility of the keyboard itself. For example, there is a case where a keyboard input is started auxiliaryly while voice recognition input is performed mainly by exchanging with the PC main body, starting from a simple voice command such as erasing between punctuation marks or returning the cursor. It begins by placing the microphone on a PCB board with an inductive semi-permanent keyboard for future expansion. A microphone array is useful for overcoming noise from the operator's mouth and collecting sound, but this can be done without significantly increasing costs.

Remote control / game console controllers that make heavy use of conductive rubber buttons have been hit by button contact problems not far away. Capacitance switch types that use conical metal springs to achieve contactless PC keyboards are sufficient. The market is dominated by membrane contacts because it cannot perform.
In a game machine, it can be said that the contact failure problem that is currently held is almost completely solved by shifting from the conduction type to the capacitance switch type instead of the conduction type. The parallel processing eliminates the time delay due to the scan cycle and its uncertainty, and allows fine game operations. This is the core part. The advantage added to the core part is that an analog amount can be embedded in the control, and a speaker / microphone / vibrator / three-dimensional acceleration sensor can be embedded in a PCB board at a minimum cost.
At the core of the TV remote control transition from the contact type of the rubber contact to the contactless contact interface type, as shown in FIG. 49, the conductive rubber shorting bar is replaced with a polyester film provided with an aluminum layer of an electrostatic shield. This is a semi-permanent life, and the IC chip must be replaced with parallel processing, but the manufacturing cost remains the same, and it is performed in a very simple procedure. The so-called contact point of the remote controller may be so-called a metal needle descending from the metal surface, but the fixed contact point deteriorates immediately. Therefore, although surface-to-surface conduction is established, in practice, conduction is established at a large number of contact points, and the entire area does not contribute to conduction. However, the larger the counter electrode area, the greater the number of conductive contact points. On the other hand, if the parallelism is not taken, it is meaningless to increase the area. Therefore, it was expected that the rubber would be deformed and adhered when a large force was applied to the conductive rubber contact. This is disappointing and there is a rubber contact problem for today's remote control. If this type is not completely parallel, conduction between X-Z and conduction between Z-Y will not occur at the same time, but both will eventually conduct due to deformation of the rubber.
In the contact interface type in which the height of the generated pulse changes in accordance with the change in the distance between the lower X / Y electrode and the G electrode coming down in parallel, the two pulses need not be generated simultaneously. Considering that the approach speed at the end of the opposite X electrode is doubled to 2 V after landing at the speed V first on the end of the Y electrode due to the tilt of the electrode, the pulse sizes are uniform. It will be a thing.
When the conical metal spring used in the capacitance switch type is pressed from above, the periphery of the spring is parallel to the substrate, so that the parallelism is maintained even if the apex is offset and pressed. It is. One of the factors that cause it to collapse is that the metal spring is in a relaxed state, and the parallelism is lowered by the stress when it is compressed obliquely. In order to avoid this, it is desirable that the spiral electrode is flat, and in order to achieve this, an irregular shape in which the conical metal spring pulls down the rubber dome is extended. I need. The micro dome plays the role of center matching in the on-type and the off-type of the close contact interface.
As shown in FIG. 50, the TV remote control is switched from a contact rubber type to an inductive interface type or an induction type as shown in FIG. 50, but various problems are solved. In the case of the induction type, the remote control answers the location of LOST & SOUND (L & S). Can be incorporated into the PCB board, and can also be used as a confirmation sound when the button is pressed. This confirmation sound is generated by resonating the inductance of the embedded speaker winding with a resonant capacitor at several KHz, and placing the feedback gain of a positive feedback loop connected with an amplifier close to 1 so that the button is pressed. Attenuating vibration occurs triggered by the detection of falling, and it can be heard from the speaker. In other words, it is not necessary to synthesize waveforms with an IC chip. With L & S howling, if the resonance frequency with manufacturing variation is swept and memorized in advance by the TV main body, or if the location of the remote control is not known, it can be swept and searched while exciting the range of variation. When a sound wave of this frequency is radiated from the TV-side speaker, the remote control that has a positive feedback loop gain as close as possible to 1 resonates passively on the principle that the tuning fork resonates even with subtle sounds. The remote control that responded with the sound of that frequency and was pinched on the bottom of the sofa relying on it was rescued. Assuming that the confirmation sound of the remote control button is already used, this function can be standard equipment because the cost increase on the remote control side and the TV main body side is only a few yen or less.
As shown in FIG. 51, the game machine controller shifts from a conductive rubber contact to a close contact interface type or an induction type to solve various problems. In the case of the induction type, the speaker / vibrator / three-dimensional acceleration sensor can be simultaneously incorporated into the PCB substrate.
In the orthodox form of magnetic levitation shown on the left side of FIG. 51, even if the key is pushed from an oblique direction, it operates without any problem. If there is an iron plate below, the repulsive magnetic force will increase. You can press it diagonally, or select the key that moved greatly when you touched the next key. Even if you put your fingertip and sink, if it is still, it will respond to changes from there. There is no bottoming out, and the round trip time is short because it reaches down and does not stagnate. When this is turned to a PC, the keyboard including the storage function becomes thinner and at the same time a deep strok can be secured, so that the same format can be applied to the desktop and the notebook. Computers are often used for games, and there is no need to separate the game keyboard from the normal PC keyboard.
On the other hand, the core part of the PC keyboard is to change the life of the membrane contact from the current finite to semi-permanent, and in principle, the key touch is comfortable and the character input speed is increased. Taking a notebook PC as an example, as shown in FIG. 52, it is possible to shift from a pantograph type membrane contact type to a contact interface type.
Further, as shown in FIG. 53, it is possible to shift from a pantograph type membrane contact type to an induction type. A part of this transition is such that detection is already completed when the stroke starts by striking the key with a fingertip of the contact interface type or the induction type, which was conducted when the stroke bottomed out. When an iron plate is laid on the induction type, the number of turns can be reduced because the voltage pulse of the coil increases.
In order to voluntarily advance these replacements necessary for society, it is necessary to wait for the replacement period and reduce the manufacturing cost, so the manufacturer must switch the method. The close contact interface type and the induction type of parallel processing satisfy this condition. On the other hand, as the choice on the consumer side, performance and reliability are overwhelmingly improved, so the market competition principle of not buying products from manufacturers that do not supply it needs to work. Next, standardization is required, and it is necessary to have no trouble in accepting a new form with excellent performance and reliability. Finally, the most important requirement is that your TV remote control and PC keyboard are still functioning well, and the characters printed on the buttons are not faint, so you can buy new products without the remote control or keyboard. It is desirable that the concept of My Remote / My Keyboard be established, and it is desirable that most of the parts can be reused when discarding the 3rd generation Remote / Keyboard inherited from Grandpa or Grandma. This is the heart of replacement.

If it is indispensable to place an iron plate on the bottom of the PC keyboard in terms of electromagnetic shielding, it will be used to reduce the overall weight, reduce the thickness, reduce manufacturing costs, and use the iron plate as part of the case. It is required to use it to improve the functionality of the keyboard / keyboard. FIG. 54 shows the main points. By way of illustration, the inductive top-and-bottom reversal format can be realized with a much thinner keyboard because only the thickness of the button remains on the iron plate when the notebook PC liquid crystal panel is closed. However, as the number of buttons is equal to the number of buttons, the attracting magnet is pushed down by the stroke, so that space must be left open. As a general-purpose PC keyboard, the plastic case is abolished in the form of upside down, and as shown in FIG. 63, the product is designed with only one iron plate, and further cost reduction is achieved.

If it is indispensable to place an iron plate on the bottom of the PC keyboard in terms of electromagnetic shielding, it will be used to reduce the overall weight, reduce the thickness, reduce manufacturing costs, and use the iron plate as part of the case. It is required to use it to improve the functionality of the keyboard / keyboard. FIG. 54 shows the main points. By way of illustration, the inductive top-and-bottom reversal format can be realized with a much thinner keyboard because only the thickness of the button remains on the iron plate when the notebook PC liquid crystal panel is closed. However, since the attracting magnet is pushed down the iron plate by the number of buttons, the space must be opened. As a general-purpose PC keyboard, the plastic case is abolished in the form of upside down, and as shown in FIG. 63, the product is designed with only one iron plate, and further cost reduction is achieved.

Shows an overall view of a two-contact conduction type key matrix scan. Is a conduction type / capacitance switch type / contact interface type / inductive type. Indicates the transmission waveform of the key matrix scan. Indicates a capacitance switch type / scan type. Indicates the change width and transmission waveform of the capacitance of the capacitance switch. Is a means for reducing current consumption by non-scanning the conduction type with a multi-value system. Is a magnetic coupling type scan type. Indicates the stroke, threshold value and operating point of the fingertip. Is a quantization unit of nerve transmission pulse and muscle reaction. Is an interleave of L-type / R-type carbon atoms and a benzene ring. Is a defect in the bonding of the 32-hedral fullerene due to the L-type / R-type carbon atom. Shows the background of the elastic / flexible structure and insulator of natural rubber. Is the establishment of conductivity and poor conductivity due to the carbon black filler. Is the essence of the difference between a conduction type switch and a variable capacitance switch. Indicates a coherent interface type key matrix for parallel processing. Indicates an on-type / off-type of a contact interface type remote controller / PC keyboard. Indicates a capacitance switch type and a contact interface type on-type / off-type. Indicates capacitance switch type waveform transmission. Is a self-generated pulse of close contact interface type. Indicates a contact interface type off-type self-generated pulse. Is the mechanism of waveform generation based on the adhesion interface principle. Is a close contact interface type in which X / Y electrodes face each other. Is the tolerance for the tilt of the button. Shows the relaxation of the parallelism maintenance requirement in the contact interface type and the metal spring. Is the spiral structure of the contact interface electrode. FIG. 3 is an overall view of inductive self-sustaining power generation key matrix parallel processing. Is a subordinate connection of inductive coils. Is an inductive PCB pattern coil and interleaved structure. Is a through type, a dead end type and a differential connection of a movable magnet. Is a PCB pattern of a double spiral coil array. Is a single plane maintenance mechanism with a magnetic attractive balance array. Is POF / PUF by iron plate printing. Is a new key touch by magnetic levitation and upside down. Are shift keys and volume up / down buttons. Is an induction type power generation waveform, multiplication by iron plate and differential coil. Are inductive off-type and on-type using iron plate. Is the interaction of a cylindrical magnet / coil / iron plate. Is the bending characteristics of rubber domes and metal springs. Is a conversion from a rubber dome to an inductive magnetic force. Is a top-and-bottom reversal form as a complete form of comfortable key touch. Is the problem of clip sticking by permanent magnets. Is the margin for each type of disturbance / noise. Is a board-embedded microphone and speaker. Is a built-in microphone / speaker / vibrator / acceleration sensor. Is the ultimate key touch on a PC keyboard. Is a contact interface type / induction type that is discriminated at the start point of the stroke. Is a reading of the analog quantity reflected in the generated / generated pulse. Is useful for keyboards that are personalized by learning functions. Is the core part of TV remote control / PC keyboard and achievement of goals. Is a transition from a conductive rubber contact of a TV remote control to a contact interface type / induction type. Is a transition from a rubber contact game machine controller to a contact interface type / induction type. Is a transition from a pantograph type to a contact interface type of notebook PC. Is a transition to an inductive keyboard for notebook PCs. Is thin / lightweight by POF / PUF (iron plate top / bottom wiring). Is an embodiment of a remote control. Is an embodiment of a PC keyboard. Is a drag profile of a PC keyboard. Is an embodiment of a game machine. Is an operation button of a general household appliance. Is the solid reliability of the ATM operation buttons. Is a non-contact remote controller that eliminates poor continuity and its requirements. Is the transition from the conductive rubber contact point of the game machine controller to the induction type. Is the transition to the contact interface type / induction type top-and-bottom reversal type of general-purpose keyboards. Is a transition to the form of close contact / inductive top / bottom reversal of notebook PCs.

The embodiment in which the present invention is implemented as a TV remote control with a semi-permanent lifetime without an increase in manufacturing cost, as shown in FIG. 55, does not use a key type scan for a conductive rubber contact using a current uniform key scan. This is a parallel processing method, and it can be replaced with 9 methods in total, 6 types as a close contact interface type using spontaneous pulses and 3 types as induction types using power generation pulses. These are: crimped GND sheet, membrane GND sheet on type and off type, spiral metal spring GND sheet on type and off type, metal reversal spring GND sheet, proximity magnet, penetrating magnet, top-and-bottom reverse type.
Each characteristic is shown in Table 2.

Remote control for TVs, etc., which is conductive and uses conductive rubber contacts, is a drag force that pushes the button to short-circuit the exposed X / Y electrodes on the printed circuit board with the conductive rubber floating Z electrode. And the spring property that returns to the original position is realized by the integrally molded rubber sheet. The rubber sheet prevents liquid and the like from entering the contact portion from above. The Z electrode of conductive rubber is destined to become poor in conduction not far, but no means for solving this has been found so far.
In the simplest case, in the current conduction type, one in which conductive rubber is attached under the integrally molded rubber button is replaced with one full shield ground sheet. The touch of the button remains the same, and the durability is semi-permanent because it is a close contact interface that does not require continuity.
The following is a membrane GND sheet, which is a close contact interface type for parallel processing, and when the X / Y electrode and the G electrode are in close contact with each other using two polyester films printed with a wiring pattern like a membrane type PC keyboard. There are an on-type that detects a generated pulse and an off-type that detects a generated pulse when separated. This also has semi-permanent durability and no cost increase. In addition, the method shown in FIG. 22 in which the X / Y electrodes are allocated to the upper and lower sheets of the PC keyboard and the GND sheet is not used can be used.
Furthermore, the spiral metal spring GND sheet replaces the spring property of the conventional one-piece rubber button by using the shape of a spiral coil spring formed by punching a thin metal sheet and extending it upwards, and at the same time, replaces the spring property of the conventional integrally molded rubber button. It also functions as an interface, and there are an on-type that detects a generated pulse when the punching spring comes into close contact with the G electrode, and an off-type that detects a generated pulse when it is separated. The drag of the rubber button that bends and the drag of the metal spring that increases monotonously when pressed down are different, but both can be used as remote control buttons. Neither the on-type nor the off-type requires the spring property of the integrally molded rubber button or the rubber dome.
In contrast, reversing springs that are molded by punching out a single metal sheet are generally used to make the bistable click operation of a tap switch. And a close-contact GND electrode.
On the other hand, a parallel processing induction type using a power generation pulse when a magnet is brought close to a coil first has a proximity magnet. This is because the PCB board of the current remote control is replaced with a polyester printing film and a double spiral coil is formed, and a small cylindrical magnet is attached instead of the conductive rubber contact at the tip of the integrally molded rubber button of the current remote control. It is a thing.
The penetrating magnet is formed by forming a double spiral coil on both sides of a double-sided PCB board, connected so as to be a differential output, and providing a hole through which a cylindrical magnet passes in the center of the coil on the board. The rubber button is a way to keep the shape as it is. This differential power generation uses the fact that a symmetrical voltage waveform appears on both sides of the passage of the cylindrical magnet, the amplitude increases, and the forward and return movements can be used separately to increase and decrease the volume, for example. New functions such as control with the movement of one button can be performed. This form functions as a three-dimensional acceleration sensor, so it has an additional function of not only pressing a button but also moving it in the air as a registered motion command to the TV body, You can add a shortcut function that turns on the power switch of the main body with just a hand movement.
The last form of top-and-bottom reversal is that a double spiral coil is formed by a printed circuit on the back side of a thin iron plate, and a small cylindrical magnet adsorbed from the back side is passed through the hole in the iron plate opened in the center of the button. It is a button that can generate electricity by being pushed down from the front side, and it is a sharp button that can produce a relatively large initial drag even with a very small magnet.
Compared to the key scan method, the parallel processing method has a smaller generated pulse and requires an electromagnetic shield. However, in all of the nine methods shown in FIG. 55, this electromagnetic shield has already been performed. Each method can be mass-produced at the same manufacturing cost as the current conductive rubber contact remote control, and the touch of the button is slightly different, but it is almost the same as the conventional one, but it is only semi-permanent. And the interval for battery replacement becomes longer.
The IC chip uses the same or less number of pins as the conventional key scan type IC, and is designed separately for two types, the induction type and the close contact interface type. It becomes a simple form and current consumption is reduced. A TV maker who is a customer of an IC maker can design a remote controller by selecting one of nine types using these two types of ICs.
The contact interface type that detects spontaneous pulses cannot express that the button is being held down, so use the double click of the volume button to continuously increase or decrease the volume with the TV remote control, or return The detection of the pulse generated at the time can be in the form of a long press to stop the continuous increase / decrease.
In order to reduce the current consumption when not in use to the limit, the TV remote control that uses parallel processing turns off all the current except for the comparator that connects the X line and the Y line across the power button, and the power button is pressed. It is possible to take a method of supplying power to the entire IC chip after detecting the event.

The embodiment of the present invention which is realized as a PC keyboard with a favorable and light key touch, which has never been achieved, has a certain manufacturing cost reduction, a semi-permanent lifetime, and is almost in the form of a membrane as shown in FIG. Although it is occupied by a notebook PC, a parallel processing method that uses key scan that uses a key scan that has been fixed in a uniform form that only has a pantograph mechanism added to reduce the thickness. It can be replaced with any of seven methods in total, three as the contact interface type using the spontaneous pulse and four as the induction type using the power generation pulse. These are the on-type and off-type of the membrane GND sheet, the on-type of the spiral metal spring GND sheet, the proximity magnet, the penetrating magnet, the magnetic levitation, and the top-and-bottom reversal type.
Membrane GND sheet is the first example of the parallel processing contact interface type. When the X / Y electrode and the G electrode are in close contact with each other using two polyester films printed with a wiring pattern like a membrane PC keyboard. There is an on-type that detects the generated pulse and an off-type that detects the generated pulse when separated, and a rubber dome is used as a run-up. This is also semi-permanent and there is no cost increase.
Furthermore, the spiral metal spring GND sheet is formed by punching out one thin metal sheet to form a spiral coil spring. The spiral metal spring has a neutral relaxation point and does not require the help of a rubber dome. Configure on-type.

A parallel processing induction type using a power generation pulse when a magnet is brought close to a coil first has a proximity magnet. A double spiral coil is formed on a single polyester printing film, a jumper layer is added, and the electromagnetic shielding is performed on an iron plate placed on the bottom. This method requires a rubber dome.
The penetrating magnet is formed by forming a double spiral coil on both sides of a double-sided PCB board, connected so as to be a differential output, and providing a hole through which a cylindrical magnet passes in the center of the coil on the board. The rubber button is a way to keep the shape as it is. In this differential power generation, a symmetrical voltage waveform appears on both sides through which the cylindrical magnet passes, the amplitude increases, and the movement of the key in which the direction of the pulse is reversed on the way back and forth is given meaning. This method also requires a rubber dome.
In the magnetic levitation type, a fixed magnet for magnetic levitation adsorbed on the bottom iron plate is used in the form of a proximity magnet. Instead, the rubber dome is removed, and the iron plate strengthens and stabilizes its magnetic force. Magnetic levitation using two cylindrical magnets does not require a rubber dome or a metal stamped coil spring. The cylindrical magnet fixed on the lower side may be smaller because the upper iron plate is magnetized to increase the upper repulsive magnetic field. In addition, the distribution can be selected so that the magnetic field combined with the iron plate cancels the magnetic field of the floating magnet on the surface of the button. Since the repulsive force of the levitating magnet suddenly increases toward the close contact, the threshold value at the upper end is set to the limit that supports the fingertip, and even if the fingertip falls, it does not lead to power generation, and so Electricity is generated when the operator hits the keys with consciousness from the lowered position. In this way, magnetic levitation provides a new aspect of comfort.
The top-and-bottom reversal type is a double spiral coil formed by a printed circuit on the back side of a thin iron plate, and a small cylindrical magnet adsorbed from the back side is pushed down from the front side through a hole in the iron plate opened in the center of the button. It generates electricity, and even a very small magnet has a relatively large initial drag and produces a sharp touch. The form of upside down does not require a rubber dome. Even a very small neodymium magnet requires a great deal of force to peel off anything stuck to an iron plate. A thin and easy-to-break neodymium magnet with a plastic coating or other anticorrosive treatment requires a certain diameter to maintain parallelism when the button is pressed down. The suction force is alleviated by the diameter of the hole through which the arm connected to the button on the opposite side passes. At the starting point where the button is struck, the magnet is separated at the moment when the compressive stress of the fingertip meat exceeds the initial attractive force of the cylindrical magnet and the iron plate, and a pulse is generated. This has the fastest reaction time and the fewest parts. However, it will be a different dimension key touch. The bottom of the button hits the lower iron plate, and the drag quickly decreases from the previous separation, and helps to release the finger when returning. Even if you know that this key touch will be the most comfortable, you need to go through a certain stage to switch from the membrane type / pantograph type / capacitance switch type that people around the world have become accustomed to. Let's go.
The idea of selecting each form of the keyboard technique of parallel processing in a separate form suitable for each desktop PC / notebook PC is simply the desire to reduce the weight and weight in the notebook PC, Originally, it is desirable that the key touch is the same when the laptop PC is operated on the go and the desktop PC is used in earnest after returning home. Therefore, if it is the lightest from the beginning, it will be pushed down and stored when the notebook PC's LCD panel is closed, durability is semi-permanent, current consumption is minimal, and manufacturing cost is the lowest. You can do it. The present invention systematizes peripheral technologies from this viewpoint. Fortunately, the buttons are both about 19 mm in size, and there is no need to reduce the button spacing just because it is a notebook PC.
The amount used for 100 buttons with neodymium magnets is not an amount that will affect the amount of resources or increase the price of raw materials even if the world keyboard is replaced. The magnet is recovered from the controller / keyboard that can be fully used after the product body has reached the end of its life. A semi-permanent life is a royal road to save resources.

Which form of parallel processing that replaces key scan is used depends on the design philosophy in every aspect in order to provide customers with ultimate key touch for maximum satisfaction and to spread the better to the world. Must be verified. As a clue, FIG. 57 shows a profile of the drag force against the key depression of the seven methods. In order to pursue this deeply, we foresee again the understanding of the TAP action of human fingers based on the biological principle, the role of the heel that drag and threshold originally had, and what is wear? It is premised to recognize without any problems.
There is a fundamental difference between the conventional PC keyboard of the key scan format and the upcoming parallel PC keyboard. It can be said that conventional PC keyboards are turned on and off depending on the position where the fingertip is pushed down. On the other hand, in the parallel processing PC keyboard, the generated / power generation pulse appears depending on the speed, not the position where the button is pushed down. Because of this difference, a mistake in pressing in the conventional method mainly occurs because the position of the button changes, so it can be said that the speed type is prevailing from the beginning, and the fundamental difference is not filled.
Considering the characteristics of the living body and considering the combination of the ASDF / JKL: home position key and the other keys, it is preferable that the fingertip placed on the home position key is pushed down without any preliminary movement. To that end, the current initial drag of 50 to 60 g is about 5 g, and the slice reaches the lowest point as it is. Even if you do not intend to press a key and press it down toward the lowest point, it will not malfunction unless it is detected by speed. This is a characteristic of the on-type contact interface. On the other hand, when one of the fingertips in the home position extends and hits another key, it strikes downward by the reaction of the muscles raised up, so it is not pushed down further, but at the moment it is hit Ideally, it should be sliced, and the ultimate drag may be zero. The detection is already completed even if it is further pushed down from there. This is a characteristic of the off-type of the adhesion interface. This combination is thought to give the eye-catching properties beyond all the combinations described so far. However, this combination shown in the lower part of FIG. 48 poses a big problem in that only eight keys are on-type and the remaining keys are off-type. In other words, in the ON type, all other key electrodes have the minimum capacity, but only one of them changes to the maximum capacity, so it is discriminated without any problem, but in this ON / OFF mixture, each X / Y line While most keys have the maximum capacity, one key changes from the minimum capacity to the maximum capacity, so that the voltage change is absorbed and the generated pulse amplitude is greatly reduced.
On the other hand, if the home position key is off-type and the remaining keys are on-type, this change will be absorbed by the capacity hanging on the X / Y line and reduced. It will not be. But this is not what you want. As a countermeasure, the key electrode of the home position is relatively large. There is a method of making the remaining key electrodes relatively small. Or you can reduce the absorption by collecting the keys of the home position on the same line.

While notebook PCs had to accept the various harmful effects of using the standard pantograph-type keyboard and making the keyboard thinner and thinner, desktop PC keyboards have sought to be more sophisticated and comfortable. There was demand. However, it remains highly questionable whether the understanding of what comfort is shared with the world's industry in a rational manner. In other words, the membrane contact was not left in the range of optimization based on trial and error experience, which was constrained by the fact that the membrane contact was conducted at the lowest point of the stroke. It can be said that the keyboard design was focused on providing motivation for consumers to buy with some sort of differentiation, and the belief that the ultimate would not change for 100 years was lacking.
As a standard keyboard, it can be said that when the ultimate key touch is clarified, it must be a common comfort for everyone. Otherwise it will not become a standard. It can be said that the drag force that increases following the initial threshold value does not need to start decreasing after the inflection point, but may decrease monotonously or increase gradually following the initial drag force. If it increases slowly, detection is completed immediately after the initial threshold is exceeded, and then the depth of the lowest point at which the key is pushed down becomes deeper if the drag decreases monotonically, and becomes shallower if it increases slowly. . Even if it becomes shallow, it is not uncomfortable if the fingertip does not feel a noticeable dead end. If such an ideal is realized, for example, in the form of a guided top-and-bottom reversal, it can be said that it can be in the form of a standard keyboard.

If there is a standard USB-connected keyboard on the desktop PC, it is clear that the same specifications can be applied to the notebook PC keyboard having the same arrangement of the same 19 mm square keys. However, at the expense of the pantograph mechanism for reducing the thickness, the stroke is only about 1 mm, causing discomfort. If the notebook PC button is pushed down on the LCD panel and stored, the same format as the desktop PC can be selected. From the viewpoint of parallel processing, membrane GND sheet on type and off type, spiral In the five methods of the metal spring GND sheet, the proximity magnet, and the penetrating magnet, the pressure when the button is stored is much smaller than that of the conventional membrane type, so that the fear of pressure marks is reduced. Among them, when the cylindrical magnet comes to the lower end, the component of the seesaw operation is added by the magnetic attraction between the magnetic shield and the iron plate of the electromagnetic shield laid under the film. Shows that it acts as a packing cushion, although the drag during storage increases. In the top-and-bottom reversal format, the drag decreases as the pressed distance increases.
In retrospect, when you enter the screen on the LCD panel, the screen is hard and the fingertips are squeezed, but that leaves no traces other than hand-grip. If you do the same thing with the keyboard buttons, fortunately the buttons will sink in the upside down format, eliminating the discomfort of pressing a hard screen. At the same time, the drag after exceeding the threshold also decreases. Therefore, if the key touch in the form of upside down is universally accepted by the people of the world, it is no longer necessary to separate the notebook PC and desktop PC keyboards. If you become accustomed to the key touch in the form of upside down, it can be said that the record of the number of characters input per unit time will be clearly rewritten. The number of parts is small, the manufacturing process is the simplest, and even if a failure occurs, the button and the magnet can be easily replaced.
Another requirement is that even if the external keyboard of the desktop keyboard is submerged, it is standard, so it is sufficient to replace only the keyboard, and there is little damage. On the other hand, if you accidentally spill coffee on the keyboard of a notebook PC, it would normally be unusable even if you did not enter the main body, so it was a social loss to replace the notebook PC itself. Is big. For this reason, it is desirable to have a natural mechanism that is resistant to submergence and can remove only the keyboard of the notebook PC from above if possible, and can be washed and recovered.
If these two requirements are satisfied, it can be said that there is no difference between the requirements of the notebook PC and the requirements of the desktop PC. In addition, the current consumption of the keyboard portion of the notebook PC driven by the battery should be as small as possible. The power consumption of the close contact interface type / induction type PC keyboard is reduced to the limit compared to the past. For example, if it is 0.05 mW, the scale of wireless power transmission used for charging the battery is usually 3 to 5 W. For example, in the ISM band of 13.56 MHz, the PC keyboard is passively read like a radio ticket without using a battery or a transmitter. Just get things done. If the IC chip is provided with this reading function in conjunction with the wireless movement of a mouse that does not move at the same time as a normal keyboard, it will greatly reduce the cost of cables and encoders / decoders for USB connection. In particular, in the keyboard of a notebook PC, the wiring between the keyboard and the touch pad becomes an obstacle at the time of washing, and the airtightness on the main body side is impaired. However, on the other hand, since radio waves leak, security measures are required.

The game console controller is more durable than the attempt to change the rubber contact conduction type to the capacitance switch type by re-recognizing the core part of the contact beyond the distinction of capacitance change / conduction as the adhesion interface. Can be raised. However, the game console controller with a small number of buttons jumps to a form of penetrating levitation that combines a penetrating magnet using a double-sided PCB board and a magnetic levitation using two magnets as shown in FIG. You can enjoy the benefits.
The power generation pulse when the power generation magnet goes past the middle position of the upper and lower differential connection coils due to the retraction of the button when the fingertip is hung has immediacy without the time delay of the key scan, and the time resolution is Extremely expensive. It can be discriminated whether the generator magnet has reciprocated on the upper side and whether it has reciprocated past the hole or has been stationary for a while. In addition, it is possible to relatively distinguish between reciprocating fast and reciprocating slowly. Since the pulse voltage generated in the forward direction and the pulse voltage generated in the return direction are differentially amplified, the polarities are reversed and discrimination is possible. Even if the button is not pressed, the structure of the three-dimensional acceleration sensor allows the movement to be detected to some extent even if the controller is moved. By passing a current from the IC chip side to the differential coil, vibration for each button can be transmitted to the fingertip. In that sense, it will be possible to build a foundation for building the ultimate game console controller.

General household appliances such as microwave ovens, washing machines, dryers, refrigerators, etc. are usually rubber contact continuity types except for some membrane switches. Since it is operated, the deterioration of the contacts is accelerated. If the operation panel mechanism has the same thickness and area, the manufacturing cost does not increase, it has a practically semi-permanent lifetime, and it can withstand the ingress of water and smoke. It is desirable to switch to the appropriate means. In this case, since the matrix size of N rows / M columns varies, several types of IC chips must be prepared. For uniform replacement of the contact structure, the off type of the contact interface is recommended, but the induction type top-and-bottom reversal type can also be selected because of the good switch off. In home appliances, buttons are often pushed through a cosmetic cover on the surface, so it is necessary to consider the breakage and durability of the buttons including them.
Previously, the mechanical switch before the appearance of the key matrix scan was clean and there were not many failures, but an unpleasantly integrated push button panel was pressed to reduce manufacturing costs It has been a long time since it was replaced by a shape. The problem is that no matter where the manufacturer's home appliances are bought, it is a uniform rubber contact, so there was no choice but to buy it with a preparedness for failure. Because the operation panel is built directly into the equipment, it is necessary to pay close attention to electromagnetic interference in parallel processing, and in some cases it is also necessary to verify lightning surges. FIG. 59 shows a basic structure required for an operation button of a general household appliance.

ATM or other similar transactions are progressing on screen input, but full dissatisfaction, delay in response, ramping and uncertainties in input determination will be necessary separately from the present invention. . What should be clarified here is not a screen input but a button such as 10 keys. It is rare that a rubber contact that is prone to failure is used for this button, and more reliable means are used, among which a capacitive switch type is selected. However, it cannot be said that the manufacturing cost is high and the feel of the button is excellent as an ATM operation. By organizing the requirements for ATM buttons, it is highly reliable but not compromising other requirements, tolerance for not knowing who to press, comfort of operation accepted by everyone, code stolen There should be no robust mechanism, such as not.
The gap is filled so that there is no particular sense of incongruity when operating the operation panel of the ATM, regardless of whether the touch of the key of the remote controller for TV, which is completely different for consumers, is touched by about 1 mm or the touch of the key on the PC keyboard. Things are desirable. On top of that, it is required to be commercialized with a unified standard. As shown in Fig. 60, for example, when the top-and-bottom reversal format is used as a standardization method, the key operation procedure is as follows: press the button ⇒ confirmation sound without delay in response ⇒ double check of two round trip pulses ⇒ It is a form to input to the end without looking at the screen. The button shape should be a hard button that is not rubber / The button will hit in about 2 mm / There will be a strong threshold / Unified key arrangement and its pitch / You can enter even if you press it continuously or open it / Press hard Habits are useless for parallel processing, and the keys that are obsolete / improperly pressed are not the <Back> button, but one character is deleted by double-clicking. Nowhere else in the world you can try the 10-key buttons on ATMs, you can't find anything that is satisfactory from these perspectives. The ease of use of the buttons is determined by design constraints, not user ease of use.

In most cases, the remote control for TV, which suffers from poor continuity before the TV main unit breaks down, is an excuse for contact wear and not strictly failure. The situation that TV that is a part of life does not work is serious. As shown in FIG. 61, the structure is replaced with the simplest form of inductive top-and-bottom reversal, which gives a feel of a sharp individual switch, and adds an embedded speaker with almost no cost increase. The remote control can have an L & S function that informs whereabouts. By switching to the parallel processing format, the part of the current that was always consumed by the key matrix scan is also eliminated, and the battery exhaustion interval when the remote control is not used can be brought close to the battery exhaustion due to natural discharge.
However, the button of the integrally molded rubber sheet is preserved as it is, and the modification of the manufacturing process so far, which has been mostly consigned to outsourced manufacturers, is minimized, and the Z electrode of the conductive rubber directly under the button is made of polyester. If it replaces with the GND electrode of the whole surface shield based on a film, it will become an adhesion interface on type. It seems that this is more reasonable. The manufacturing cost is the same as before, and the durability is semi-permanent. However, it can be said that there is a difference in how consumers push the buttons. Until now, it was mainly the thumb, not the consciousness of pressing quickly, but just pressing strongly. It can be said that this reflects the experience that even if a contact failure is about to occur, it can be communicated by pressing it strongly. It can be said that the contact point will not reach unless it is pushed hard because the drag is increased so that it is not pushed by mistake. It does not have the property that the drag is reduced when the rubber is bent and pushed down.
When this is switched to a tight interface, the drag will be so small that consumers will immediately swallow it to work even if pressed lightly. If the button is pressed lightly, the stroke will naturally be faster and the generated pulse will be larger. If the button is pressed down when it is not selected, the speed will be slow, so setting the appropriate slice level will not cause a malfunction.
The remote control for TV uses the parallel type, and by turning off all bias currents other than the comparators on the IC chip connected to the X line and Y line across the power button, wasteful current consumption is almost zero. In the case of a close contact interface, in addition to the buttons shown in the table, the weight inside the case shows the potential change that appears when the G electrode film on the X / Y electrode is pushed by the acceleration at which the remote control is lifted. By detecting, it is possible to select and set a mode capable of automatically starting the TV main body and capable of operating procedure shortcuts.

A game console controller that can be said to be destined to become a contact failure not far away, has continued to use conductive rubber contacts without any other conceivable solution, or the motion detection controller is a contactless contact type Instead of the above, all problems can be solved together by switching to the induction type as shown in FIG. First, the life of the controller becomes practically semi-permanent by adopting the induction type, and there is no equivalent of wear. Secondly, it becomes possible to embed analog quantities in the operation of all buttons. Thirdly, the buttons of the game machine, which are not necessarily arranged on the plane, are integrated with the wire harness, and the PCB board is used to arrange the buttons. You can avoid it. Fourth, a vibrator / speaker / microphone / accelerometer can be installed for a minimum of manufacturing cost and used for multiple purposes.

As a general-purpose keyboard for a desktop PC, as shown in FIG. 63, a conductive membrane type having a threshold value of 50 to 60 g and a stroke of about 2 mm has been used in the past. This can be replaced with a contact interface on-type that uses the repulsive force of a microdome with a threshold of about 30 g and a stroke of about 2 mm, or a guided top-and-bottom reversal type with a threshold of about 20 g and a stroke of about 2 mm. This replacement gave a different dimension of key touch and did not bring much progress, but simply ended the development competition for differentiation, promoted standardization again, dramatically improved durability, and simplified the manufacturing process. Cost can be reduced. Moreover, the character input speed can be improved with less fatigue, and those with a learning function can be personalized to reduce input mistakes.
Among them, one iron sheet is pressed to integrally mold the top plate and legs, and a through hole is provided in the top plate according to the key arrangement to form a top-and-bottom reverse type. A printed circuit consisting of a double spiral coil and wiring is constructed, and by simply adding an IC chip and a USB cable connected to this, it becomes a simple and robust keyboard. By selecting the three elements of the hole size, cylindrical magnet thickness, and height, the stroke and drag profiles can be selected relatively freely for each key. Even if there is a malfunction, only the key part can be easily replaced, and if it gets dirty, it can be washed.
Traditionally, roll-up keyboards have been thought to be impossible to implement in a true sense because of their shorter strokes. If you roll using two sheets first, the outer sheet must stretch. Use the material of the sheet that allows it, as shown in Fig. 63, make the upper and lower sheets face each other when flattened, secure a comfortable 2-3mm stroke with magnetic repulsion, and tightly adhere when rolled up If the double spirals are formed on both sides of the film and are connected in a differential manner, disturbances are suppressed so that various conditions are met and a roll-up keyboard that does not sacrifice any characteristics. It ’s done.

As shown in FIG. 64, a keyboard of a notebook PC has a conduction type pantograph format that has a threshold value of 50 g or more in most cases, and has a slightly uncomfortable feeling by shortening the stroke to about 1 mm for thinning. I came. If the threshold is around 30g and the stroke is about 2mm, the thickness will not change, but the adhesion interface will be on-type using a microdome, or the threshold will be 20g or less, and the button will pop up when the liquid crystal panel is opened. With a stroke of about 2 to 3 mm, the thickness when the button sinks and stored is thinner than in the case of the pantograph type, and by making it an off type of induction type top and bottom reversal type that gives a lighter key touch, Even if you spill coffee on the keyboard, it doesn't become malfunctioning. If it repeats, you can remove the keyboard from the top and wash it with water. You can put an end to unacceptable business practices you didn't get.
Works even if the key is pressed from an angle. Even if you touch the adjacent key, the key that performed the big action is selected. Even if you put your fingertip and sink, it will respond to changes from there.

The on / off switch is originally an individual switch, and is achieved by a mechanism that uses the presence / absence of conduction by contacting / separating the metal electrode, but measures such as applying gold plating to prevent contact failure. Is taken. However, remote control buttons and game console controller buttons do not use individual mechanical switches for the purpose of being operated with a light touch or to reduce the cost of a large number of buttons. If not, a key matrix scan format is used, and a combination of an integrally formed rubber sheet and a conductive rubber electrode, which gives the button springiness and also functions as a protective film, is used. The key matrix of this conductive rubber contact is destined to eventually become defective, but the fate of poor contact remains imposed on the consumer without any choice.
It can be said that all of these conductive rubber contact key matrix scans must be replaced by close interfaces or inductive parallel processing. In addition, it is desirable that all the membrane type / pantograph type / capacitance switch type PC keyboards, etc. that do not use conductive rubber contacts should be replaced with a close contact interface or inductive parallel processing. These applications are listed below.
(1) Control buttons in the interior of automobiles where the usage environment is harsh (2) Control buttons and remote control buttons on the main body side of video / audio equipment (3) PC keyboard (4) to pursue all comfort, semi-permanent life and low manufacturing cost ) Operation buttons for portable devices that are compact, comfortable and require low cost (5) Operation buttons that do not require cost to improve the reliability of cooking utensils / refrigerators / washing machines (6) Control buttons overused in game consoles Ultimate durability (7) Elevator and other operation buttons that require higher reliability (8) Wall-mounted control panel for controlling air conditioners

DESCRIPTION OF SYMBOLS 1 Conductive contact 2 Conductive rubber contact 3 Membrane switch 4 Scan period 5 Conical metal spring 6 Drive resistance 7 Three-value comparator 8 Short ring 9 Pressing stroke 10 Nerve transmission pulse train 11 Optical anisotropic step 12 L type and R type atom 13 benzene ring bond 14 contact conduction chain 15 polar coordinate display 16 GND sheet electrode 17 microdome 18 scan pulse 19 wiring stray capacitance 20 spontaneous pulse 21 CR discharge waveform 22 step displacement 23 rubber dome 24 separation distance 25 cylindrical cavity 26 induction Mold intersection 27 Power coil 28 Movable magnet 29 PCB through hole 30 Through hole 31 Saturn's ring layer balance 32 Printed wiring 33 Fixed repulsion magnet 34 Proximity capacitance leakage 35 Magnetized spin current 36 Iron plate eddy current 37 Negative differential Drag 38 Bending conduction of film 9 horizontal holding 40 suction compressive stress 41 iron clip fingertips meat 42 return the twisted wire 43 pattern coil 44 diaphragm 45 fingertip magnetic attraction TAP
46 Loose state of metal spring 47 Parabola trajectory of stroke versus time 48 Programmable slice level 49 Parallel processing pulse 50 Comb-shaped cross electrode 51 Multiple spiral punch spring G electrode 52 Durable cushion 53 Double spiral winding 54 Connecting arm 55 Projection 56 Button 57 Polyester film substrate 58 Detection point

Claims (8)

Conduct a key scan by connecting the intersections of the X line of N rows and the Y line of M columns with a conductive contact or variable capacitance in a conductive or capacitive switch type key matrix scan using scan pulses. However, based on the new concept of a close contact interface where the parallel and flat X and Y electrodes approach / separate but do not lead to conduction, the parallel generation of self-sustained pulses without key scanning is eliminated. An active key matrix that can identify a key pressed from a combination of N * M by processing, and is connected to a bias resistor from a power source and charged, and a parallel plate electrode having a constant potential difference equal to the power source voltage is Because the electric field strength between the electrodes is proportional to the charge distribution density on the electrode plate, the electric field strength is proportional to the change in the distance between the electrodes in a short time when positive and negative charges do not enter and exit. When the distance between the surfaces is 100 μm away from the close contact state where the dielectric insulating layer of about 10 μm is sandwiched between the electrodes, the potential difference which is the product of the distance changes transiently and does not lead to conduction between the electrodes, the capacitance becomes the insulating layer Although it changes about 20 times or more including the relative dielectric constant of the above, it does not pay attention to this capacitance change ratio, and the transient generated pulse is generated by the transiently changed potential difference returning to the power supply voltage with the CR time constant. In the first separation electrode type in which there are two methods for parallel processing of a matrix composed of N rows of X lines and M columns of Y lines, each X line is used. The X electrode and Y electrode connected to the intersection of the Y line and Y line are placed in parallel, and the distance facing the common GND electrode is changed, but N + M X / Y lines are connected to the GND potential through the terminating resistor in the IC chip. Biased to high voltage When the distance between the electrodes at one intersection of the matrix changes from separation to contact in the on type, a negative pulse is generated. When the distance changes from contact to separation in the off type, a positive pulse is generated and the X line and Y line are generated. Proceeding in parallel through each of them, the combination of N * M of the comparator of the IC chip is judged as the address of the pressed button if there are two pulses, or the second differential electrode type Then, the GND electrode is not used, and the X electrode connected to the X line and the Y electrode connected to the Y line are directly opposed to each other. If the Y-line is biased at a lower voltage, a negative pulse is generated differentially on the X line and a positive pulse is differentially generated on the Y line when the distance changes from separation to contact in the on type. When the distance is changed from close contact to separation in the off type, a positive pulse is generated differentially on the X line, and a negative pulse is differentially generated on the Y line. In this combination, the one with the largest difference is determined as the address of the pressed key. The adhesion interface is contactless because it is in close contact with the dielectric insulating layer, so that it has no contact, has a semi-permanent life, and has no time delay in response. In the differential electrode type, the conventional membrane-type PC keyboard has no jumper by connecting the upper sheet on which the X line is drawn and the lower sheet on which the Y line is drawn to make the intersection perpendicular to each other. It is as simple as configuring a key matrix with only two simple sheets, improving reliability and key touch, and the speed and depth at which the detected generated pulse was pressed. The analog information is reflected in the magnitude of the positive and negative pulses and the time interval, making accurate judgments and realizing comfortable key touches. In parallel processing, the state of all keys can be read simultaneously. It can be used to provide the learning function used. In the case of a TV remote control, all the bias currents other than the comparator that receives the generated pulses from one X line and one Y line across the power button are turned off on the IC chip by utilizing the parallel processing. By setting the state, the standby current can be made almost zero. In addition, the weight stored under the top cover of the remote control in parallel with the power button above is the same as pressing the power button when the remote electrode is brought into close contact / separation due to the increase in gravity when the remote control is lifted. Then, turn on the TV itself. In addition, by adopting a parallel interface coherent interface type, the power consumption of the PC keyboard is reduced to about 50 μW, and the IC chip has a circuit that can read the result of key input from the PC main body side by radio frequency such as ISM band. For the purpose of realizing inductive matrix parallel processing without scanning, which replaces the conventional key matrix scanning using the conduction switch and the capacitance switch, which has problems in reliability, etc., N rows of X lines and M columns of Y Spiral coils are arranged at each intersection of the line, and a short-circuit induction pulse current generated by a change in which a small thin permanent magnet such as a neodymium magnet is closely attached to / separated from the coil, or an open terminal pulse multiplied by jωL by the inductance L of the coil A key that was pressed while observing all N * M key states at the same time by inputting the converted voltage into one amplifier on the previous IC chip that traced the X and Y lines. It is determined that the other key movements are extracted as reference information. The first of the two methods is the serial connection to the X line. The spiral coil inserted and the spiral coil inserted in series with the Y line are formed concentrically, and two power generation pulses due to the close contact and separation of the magnets are transmitted to the IC chip in parallel with the X line and the Y line, respectively. Or a separate coil type inductive type that will be the differential input of the second, or as a second method, one spiral coil connects the X line and the Y line at the intersection of the matrix, and the generated short circuit induction pulse current By making the open terminal voltage multiplied by jωL into the differential coil type inductive type that uses the differential input of the previous comparator following the X line and Y line, more accurate determination and analog information A parallel processing inductive key matrix that enables reading. By reading the open terminal voltage in the former separated coil type, the influence of N / M coils being cascade-connected can be eliminated, but in the latter differential coil type, the input terminal of the differential amplifier is not connected. Whether the short-circuit induced pulse current is examined by short-circuiting with a small resistance or the open-terminal pulse voltage that is not short-circuited is examined, both methods have independence that can be distinguished from information from other intersections. In parallel processing for examining active power generation pulses, there is no delay in detection delays and uncertainties that wait for scan pulses to come around with a repetition period of 10 mS or less in conventional key matrix scans. Time resolution can be improved. Also, by stopping scanning, the life of the remote control battery can be extended first. Subsequently, in the case of a TV remote controller, all the bias currents other than the comparator receiving the generated pulses from one X line and one Y line across the power button are all utilized on the IC chip by utilizing parallel processing. By setting the off state, the standby current can be made almost zero. In addition, the weight housed under the upper cover of the remote control in parallel with the power button above causes the lower magnet to closely contact / separate due to the increase in gravity when the remote control is lifted, and has the same role as pressing the power button. The TV can be turned on. In addition, by adopting parallel processing inductive type, the power consumption of the PC keyboard is reduced to about 50 μW, and the IC chip has a circuit that can read the result of key input from the PC main body side by radio frequency such as ISM band.
In an inductive parallel processing system in which the movement of a magnet gives a power generation pulse to a coil to constitute a key matrix, a multilayer substrate of 2 layers / 4 layers /. Make the coil into the same concentric right-handed double spiral from the outer periphery to the inner periphery in the odd layer, then move to the even layer through the through hole, and this time the concentric right-handed double-sided spiral returning from the inner periphery to the outer periphery Therefore, it is not necessary to provide an additional layer for forming a jumper, and N * M pulses obtained as a result of amplifying the difference between the positive and negative power generation pulses of the X and Y lines on the IC chip. By taking out one of the amplitudes indicating the maximum value and detecting the address of the pressed button, the number of turns of the coil is doubled so that the detected voltage is apparently doubled. Is differential That the DC offset of the width device is made to be 1/2. Also, using a thick double-sided board, the movable magnet is outside the thickness of the board by making a differential connection that reaches the center with a clockwise spiral on the front side and moves to the outer side with a counterclockwise spiral on the back side Inductive current cancels at times, but when it is between the thicknesses of the substrate, a steep differential current becomes a highly sensitive motion detector, and disturbance is canceled out by differential connection and suppressed. To reduce the burden on the electromagnetic shield. In the parallel scan matrix scan, the induction type performs two-dimensional interleaving in which the direction of the winding at the intersection of the X line and the Y line is alternately reversed clockwise and counterclockwise to repeat the spiral coil. In addition to twisting the connection wiring alternately, prevent external magnetic induction from accumulating on the same line. In order to make the positive and negative directions of the power generation pulses of the spiral coil of the same X line or the same Y line uniform, the direction of magnetization of the permanent magnet is inevitably reversed alternately in the XY directions. It must be interleaved, but the leakage magnetic field of the arranged magnets cancels out so that iron pieces such as clips are not attracted by the button surface. In the parallel processing induction type key matrix, a magnet magnetized in the repulsive direction is placed and fixed below the movable magnet, the movable magnet is magnetically levitated, and the weight of the movable part and the magnetic repulsive force are balanced. In order to give the initial drag from the state that was pressed, the starting point is that the button was held down and held by the stopper, and only the magnetic repulsive force of the two magnets was given as the drag, and the four keys placed at the home position of the blind touch The starting point is that the pressure stress of the fingertip's flesh is below this initial drag and the key is not moving, and the rest of the keys are zero balanced and the starting point is the starting point. Inducing positive and negative power generation pulses to be detected on the IC chip. Since the initial drag is small, the slight movement of the 10 fingertips is reflected in the power generation pulse as an analog quantity, and since it is a parallel process, the related keys can be monitored at the same time. Can be done. At the same time, in order to dispel the discomfort of the dead end, which is the price of the click sound of the bottom, even if the button is pressed more strongly, it will not reach the bottom due to a sudden increase in repulsive force. This click sound, which has been a signal to conduct, can be replaced by a fine electronic sound of a speaker built in the PCB substrate. In addition, the magnetic fields of the upper and lower repulsive magnets cancel each other at the keyboard surface position, further reducing the possibility of attracting iron pieces and the like. In the inductive key matrix of parallel processing described above, the iron plate placed at the bottom of the keyboard is moved to the top to serve as an electromagnetic shield and stabilizing weight, and a small hole corresponding to the number of buttons is opened. The initial drag is that the cylindrical magnet is adsorbed on the back side, the button is placed on the front side, and the magnet and the button are connected by the arm through the hole in the iron plate. A keyboard that detects the generation pulse generated in the spiral coil printed on the back side of the iron plate in the stroke to be peeled off, by transmitting it to both the X line and the Y line in parallel. The key factor of comfortable key touch is that the magnetic attraction force decreases with the distance and returns with magnetic attraction, and the re-adsorption is an important factor, and the drag change profile is the hole diameter, cylindrical magnet thickness and its diameter. By using what is determined by three elements, it is possible to perform different optimization for each key. In addition, prompt detection that a power generation pulse is generated at the moment of separation is performed, which is one of rapid character input. Using this, it consists only of a steel plate or silicon steel plate with the top plate and legs integrated, such as mechanically pressing a single iron sheet, and plastic covers are basically not used for mechanical parts. A printed circuit consisting of a spiral coil and wiring surrounding the hole is attached on the back side of the iron plate or silicon steel plate with a button on the upper side of the arm through which a small hole is made in the top plate, and a thin cylindrical magnet on the lower side. Configure and connect to a small PCB main board that only has an IC chip for parallel processing, and connect it to the PC main body with a USB cable from this board. . In order to avoid electromagnetic interference from the outside with a normal PC keyboard, it inherits that a thin iron plate for electromagnetic shielding is placed on the bottom, and insulates the iron plate in the induction type N * M key matrix By forming a layer coating and forming N + M subordinate connections between the coil and wiring in the printing process, etc., or placing a pre-molded and solidified wire harness consisting of N + M copper windings in close contact with the iron plate In the case of magnetic levitation, the lower terminal magnet is placed in close contact with the iron plate so that the magnetic repulsion is achieved. A keyboard that maintains the role of an electromagnetic shield while increasing the power up to twice as much as when there is no iron plate. In order to reverse the life of the rubber contact of the key matrix scan type remote control which is shorter than the TV main body on average so far and make it semi-permanent, the drag of the button becomes extremely light if it is replaced with the parallel processing contact interface type or induction type Based on the situation where a combination of spiral coil and cylindrical magnet with printed pattern or wire harness, which is a basic component in induction type, is prepared, a speaker with a diaphragm and parallel resonance capacity added to that combination A microphone is formed and connected to an IC chip to form a positive feedback loop. If the loop gain is set close to 1, triggering with a generated pulse when the remote control button is pressed will start a damped vibration, and a minute level It becomes the operation confirmation sound. So far there is no such thing as additional cost. On the other hand, the TV main body is provided with a speaker for watching a program. These states are an essential part of a TV receiver, but the remote control plays the same role as a tuning fork so that L & S (LOST & SOUND) sonar search can be performed without any further cost. It is possible to respond to a question from a distant place by causing a resonance of a small level of sound to take a long time by placing the positive feedback loop gain closer to 1 during standby. . Prior to this, the TV main body must have a single tone generating circuit that sweeps the frequency range before and after the resonance frequency, but there is no cost. In this sonar search, other devices that may be hidden somewhere are also provided with this tuning fork function, and an L & S search can be performed from the TV receiver. In the continuity type key matrix scan, the analog information of the stroke process is discarded and cannot be reflected, and even if it is replaced with a capacitance switch type, the reflection is distorted, so it is not useful. By adopting the inductive type, the analog information of the stroke process is transmitted to the IC chip without distortion as the positive and negative pulses whose amplitude changes respectively, and the time interval between them, setting the slice level of the positive and negative pulses, and the pulse As a window for interval measurement, it can be used for optimizing light key touch / reducing mistakes / input speed. The first of the methods is to have a learning function that performs this optimization according to the individual operator, and the second is that information from keys other than the intersection is not transmitted at all in the key matrix scan, or the transmitted information is not transmitted. However, in parallel processing, the information of all N * M intersections can be separated and read at the same time, and in the blind touch method, one key pressed with a finger is used. In addition to the movement, the certainty and distinctiveness of the discrimination are killed by identifying the pattern of whether or not it matches the 101 kinds of stored information of the combination of monitoring the entire movement of the remaining 9 fingers. A keyboard that allows you to achieve the two objectives of the keyboard at the same time.