GB2242587A - Switching circuits - Google Patents

Abstract

In a power control circuit for reactive loads, particularly inductive loads such as loads 0 to 5, a high power transistor (T1) is used under control of a control circuit (IC2, R9 etc) to switch the transistor on and off. The transistor, in effect, appears as a voltage controlled variable resistor in series with the load and acts either as a main switching device or as a buffer to the main switching device whereby the switching is gradual. The circuit greatly reduces the amount of noise suppression required and may, in some cases, eliminate the need for noise suppression entirely. The circuit may be controlled by a microprocessor and included in appliances such as vending machines, dishwashers and washing machines. The transistor may be a MOSFET, bipolar transistor or a triac. In the above arrangement, the transistor T1 is controlled by the added sum of a ramp generator output and a feed-back voltage and the switched voltage is applied to a threshold device via a differentiator to produce the feedback voltage. <IMAGE>

Description

"POWER CONTROL CIRCUIT FOR REACTIVE LOADS" The invention is concerned with controlling the changes of power applied to a reactive load.

Problems of electrical interference arise when power is applied or removed. In the case of an inductive load the problems are severe on removing power whereas a capacitative load has more severe problems on application of power.

Although the specific embodiments of the invention herein described relate to inductive loads, the invention may be applied in similar manner to capacitative loads.

When reactive loads are switched on or off, noise is both radiated through space and conducted through the electrical circuits.

One common instance of this is when a motor or a solenoid is switched off: such loads are normally inductive in character.

Increasing legislation over the years has been aimed at reducing this noise because of its adverse effect on other pieces of electronic equipment nearby. Designs now incorporate noise suppression, or snubber, networks, sometimes at significant cost.

For maximum effectiveness these snubber networks are frequently mounted as close as possible to the circuit element with which they are associated. This is often awkward and expensive to implement.

Usually the snubber network comprises a capacitor and resistor in series (RC network), connected in parallel with the circuit to be suppressed. The values of these components depend on circuit details and for optimum performance each circuit must be considered individually.

Usually one of three switching elements is employed: 1. Relay contacts for use in AC or DC circuits.

2. Transistors, primarily for use in DC circuits.

3. Triacs for use in AC circuits.

In DC circuits a "flyback" diode connected across the inductive element is often sufficient suppression although in some instances this may cause unacceptable delays in the switching action and more exotic suppression schemes must be used. The switching speed of the diode may also prove to be important, making it an expensive item.

When relay contacts are used to switch AC it is generally adequate to connect an RC network across the load terminals. The contact rating of the relay should cater for both the load current and for the current through the RC network at switch on. The noise generated by contact bounce is not eliminated but is reduced. When triacs are used the RC network is usually connected across the main terminals of the triac. In some instances the current that flows through the RC network can cause operational problems since the load is now never entirely unpowered. An example of this might typically be a water inlet valve on a vending machine: The valve is held open, ever so slightly, enough to allow water to trickle through and cause drip-trays to fill and shut the machine down.

Triacs offer a number of advantages over relays. The advantages are low price, small size and no contact bounce so they are inherently electrically quieter.

An object of the present invention is to greatly reduce the amount of suppression needed, and even in certain cases to eliminate the need for suppression, by introducing a feature which makes the switching action gradual, allowing the current in the inductive circuit to decay in a controlled manner. Such a scheme, in accordance with the present invention, allows many circuits to operate without any extra snubbing components. Although the present invention is intended primarily for use in AC circuits, it also finds application in DC circuits when switching response time is important or multiple loads need to be switched.

In its broadest aspect, the present invention relates to a power circuit for reactive loads comprising a power control element appearing, in effect, as a conductance in series with the load, controlled in such a manner as to reduce radiated or conducted interference, the power circuit including a control circuit to vary the conductance of the control element.

The control input for the circuit is preferably fed to a function generator the output of which is fed to a summing device whose output is fed to the power control element, the output of said element being fed to a threshold detection device whose output provides a feedback signal fed to the summing device.

Preferably the control element is a transistor or an FET device.

According another aspect of the present invention, there is a switching circuit for reactive loads, such as motors and solenoid operated devices including relays, said circuit including at least one power control device and a control circuit, the control device in effect appearing as a resistance in series with the load and acting as a main control device or as a buffer for a main control device whereby the switching action is gradual, allowing the current in the reactive circuit to decay slowly, the power control device being used as a power modulator, i.e. as a linear device, during turn-off to gradually reduce power. If the circuit is an AC circuit the main control device is preferably a triac.

In the case of a capacitative load, the device is arranged to control the current rise at turn-on and to perform as a zero crossing switch at turn-off.

Preferably the control loop includes three main functional parts, a ramp generator, a power controller, and a feed back element, whereby the control loop performs different functions during switch-on and switch-off. For an inductive load, the control loop modulates the load supply during load switch-off and performs as a zero-crossing switch during switch-on. For a capacitative load the loop behaves as a zero-crossing switch during switch-off and modulates the load supply during switch-on.

The control device may be placed in series with the DC terminals of bridge rectifier and the load may be placed in series with one of the AC connections and one of the supply terminals, the other supply terminal being connected to the other bridge terminal. In a preferred arrangement, the control circuit is supplied from a floating power supply.

The switching circuit may include a diode bridge feeding an optoisolator through a resistor to provide zero-crossing pulses for synchronising the device switch-on and switch-off to the supply.

The switching circuit may be adapted to control a plurality of loads, individual loads being selected by triac drivers without the use of conventional snubbers.

A preferred control element is a Mosfet.

A Mosfet may be used but it is also possible to use a bipolar transistor. The switching circuit may, for instance, be applied to a vending machine the loads being solenoids and motors which operate different parts of the vending machine.

An embodiment of the present invention will now be particularly described with reference to the accompanying drawings, in which: Figure 1 is a circuit diagram of the relevant parts of a vending machine in which motors and relays are represented by loads 0 to 5, the machine being operated via a microprocessor; and Figures 2 to 7 are graphs illustrating on-off switching cycles for different loads.

Figure 1A is a block schematic diagram of a load power control circuit incorporating the present invention; Figure 1B shows a practical example of the principle of a power control circuit controlling a number of AC loads in accordance with the present invention; and Figures 2 to 8 show voltage and current waveforms for each load combination tested etc.

In Figure 1A a control input is fed into a function generator 10, the output of which is fed to a summing device 11 which sums the function generator input and a feedback signal coming into the device 11 on line 12.

The summed output is fed via line 13 to a power control element 14 whose output is supplied to a load 15 and to a differentiator 16.

The differentiated output is fed to a threshold detection device 17 and the output of this device provides the feedback signal on line 12.

The load power change from full to no power and vice versa is controlled by a control circuit in the form of a function generator 10, the changes being initiated by the control input.

Control element 14 may be a transistor or an FET device but is a linear control device rather than a triac for example. The differentiator looks at the rate of change of load conditions and will limit the rate of change of output voltage to the initial rate of change of supply voltage so as to avoid distortion and power loss.

Limiting the rate of rise of the output or load voltage reduces the maximum frequency of the radiated spectrum, whilst reducing the amplitude of the switching peak lowers the mean radiated power.

Because of the closed loop nature of the circuit embodying this invention, it is tolerant of a wide range of load variations.

The block schematic diagram components of Figure 1A will now be related to the corresponding parts of the practical circuit embodiment of Figure 1B.

Operational amplifier I C2/2 and associated components R11, R13, R14 and C5 form a ramp generator which, together with D2 and D9, comprise function generator 10. When the control input state is changed, the function generator 10 changes the drive to the power control element 14.

The components ICB, R9 are the parts of the control input circuit which, in Figure 1B, is shown as being under the control of a microcomputer system. Power Mosfet T1 acts as the power control element 14 (Figures 1A and 1B), while R16 and C6 form a pole at R16.C6 to compensate for the Miller capacitance in the Mosfet T1 which would otherwise cause high frequency instability.

Operational amplifier IC2/1 (Figure 1), with R50, R51, C7 and R52 are a scaled differentiating amplifier, which serves to monitor the rate of change of voltage at the load controller terminal.

In order that a preset level of load voltage rate-of-change may occur, a threshold device 17, (dull and R5) is placed in the feedback path.

The resulting signal from the differentiator 16 is summed with the output of the function generator by the resistor R54 which acts as summing device 11.

Other functions fulfilled by the circuit of Figure 1B are as follows: Zero Crossing Signal A zero crossing (mains voltage) signal is generated by the diode bridge D12 to D15, the resistor R55, and the optoisolator IC1.

Diode Bridge Dio The power control element 14 is placed in the DC path of bridge rectifier D10, thus allowing control of a load or loads placed in the AC path.

Control of Individual Loads Load O to Load 5 may be individually switched , once the power source to the loads has been removed by the power control element.

The circuit shown in Figure 1B has been built and tested, siting both the switching circuits and the loads within a few inches of the microprocessor-based circuit controlling the switching.

The leads from the switching circuit to the loads were deliberately draped randomly across the microprocessor circuitry. The microprocessor controller, based upon an 80C31, utilised no special error trapping nor watchdog techniques and the whole system was powered from single phase UK mains in a light industrial environment.

Bipolar transistors with suitable voltage ratings for use in an EEC mains application are generally of the triple diffused type and cannot handle high voltage and high current simultaneously, tending to fail through secondary breakdown.

In accordance with the invention, a more suitable type of device is used, which is a high power Mosfet; such devices are now available from a wide range of manufacturers.

These devices neither suffer from secondary breakdown nor are they very sensitive to dV/dt. This latter characteristic makes them ideal buffers for triacs as now the advantages of triacs can be utilised without the dV/dt problems associated with them.

Provided the peak transient power dissipation of the device is not exceeded, a Mosfet is capable of dissipating a large amount of power for the relatively short time required to perform a slow switching operation.

The circuit operated for in excess of twelve million switching cycles without any operational or device failures.

The loads employed included typical motor and solenoid loads found in vending machines, photocopiers et al.

In the vending machine applications examined,. it was found that relays with a nominal 250V 8A rating and a stated 1,000,000 switching cycle MTBF (electrical) would, in fact, begin to operate unreliably after about 10,000 operations when used in conjunction with loads similar to, but generally lighter than, those described herein.

The switching scheme employed was as follows: Switch on Load O & Load 1.

Switch off Load 0, switch on Load 2 Switch off Load 1, switch on Load 3 Switch off Load 2, switch on Load 4 Switch off Load 3, switch on Load 5 Switch off Load 4, switch on Load O Switch off Load 5 and Load O Repeat sequence.

Details of the loads are as follows: Load No. Inom.(A) Power Description Factor ~~~~~~~~~ Factor ~~~~~~~~~~~~ 0 2.5 .34 ECM 4152-1B1 Motor (5 rpm Gearbox) 1 .05 .59 M & M M8 Water Valve (Solenoid) 2 .54 .34 ECM 5376 (20 rpm Gearbox) 3 .02 .96 Crouzet 82334.5 Motor ( 8 rpm Gearbox) 4 .95 .37 Crouzet SP1147 Motor (100 rpm Gearbox) 5 .02 .96 Crouzet 82334.5 Motor (8 rpm Gearbox) These electrical loads above were run without mechanical loading, aside from the associated gearbox; this probably represents the worst case power factor condition.

All the motors were shaded pole types. All loads are designed for EEC mains operation, approx. 10% duty cycle max. Power factor figures stated are only approximate.

Each load combination was on for 4 seconds at a time. The period when all loads were off was 24 seconds. This allowed all loads to run close to their maximum rated duty cycles of approx. 10%.

After 46000 cycles the circuit was modified to incorporate circuitry to limit dV/dt during the switch-off period to little more than would normally occur at the zero-crossing of the mains sinusoid. This additional circuitry, shown within chain-dotted lines, was incorporated to further reduce noise generated by the poorer power factor loads.

After a further 50400 cycles the control program was modified to reduce the cycle time from 48 seconds to 4.0 seconds. Since the components are only under any real stress during the switching periods the tests are still valid. In this extreme condition the Mosfet case temperature was still only 50 degrees Kelvin above ambient.

Voltage and current waveforms for each load combination tested during the switch-off and switch-on periods are given in Figure 2 to Figure 7.

Figure 8 shows Vgs and Vds for T1, and the output voltage of the differentiator.

CIRCUIT DESCRIPTION The switching circuit embodying the present invention, shown in Figure 1, comprises two main sections, separated from one another by an opto-isolation barrier.

One section contains the power control elements, the load power modulation circuits and an isolated low voltage supply.

The second section is a control interface which allows a microprocessor system to initiate load changes.

The main power control element is a high voltage Mosfet, T1, inserted in series with the DC terminals 1 and 2 of a bridge rectifier D10. The load devices (loads O to 5) are placed in series with one of the AC connections 3 and one of the supply terminals N, the other supply terminal L being connected to the other bridge AC terminal 2.

The control circuit is supplied from a floating power supply that is referenced to the Mosfet source connection. A +/- 12V DC (nom.) supply is generated along with +/- 5V DC reference supplies. A diode bridge D12, D13, D14, D15 feeding an opto-isolator IC1 through a resistor R55 provides zero-crossing pulses for synchronising the device switch-on and switch-off to the supply.

The control circuit essentially modulates the load supply with a ramp, created by ramp generator RG, during load switch-off and performs a conventional zero-crossing switch during load switch-on.

The objective is to control the rate of change of current by limiting the rate of change of voltage to that of the supply sinusoid at zero-crossing. The control loop, therefore, performs different functions at switch-on and switch-off.

Dividing the control loop into three main functions gives the ramp generator RG based around IC2/2, the high voltage Mosfet TI acting as a power control element and a feedback element or differentiates based around IC2/1.

Individual loads are selected by conventional triac drivers T7, T8 and T12 but no snubbers are used.

In the quiescent state (IC3 off) the output voltage of IC2/2 is limited to approx. -0.6V by D9 and T1 is switched off. The nominal voltage across the triacs (T5, T6, T9, T10, T11, T13) is determined by the resistor network comprising R50-R51-R49 and is approx. .27 of the supply voltage. Further, the impedance of this "source" is such that even if one of the triacs should conduct no damage will occur.

Hence, provided the Mosfet, T1, has a sufficient voltage and dV/dt rating to withstand the full supply voltage range low cost and relatively poorly rated triacs may be used.

When some combination of loads is to be switched on the appropriate triacs are switched on first by activating the associated optotriac (IC6 etc.) via the outputs of the controller #CH0 - iCH5.

The P SWITCH output is activated at the next supply zero crossing event pulling the junction of R9 and R11 to -5V.

R13 and R14 cause the output of IC2/2 to swing to +10V instantly.

A more positive output swing, which would occur as C5 starts to charge up, is prevented by D9 and C9 charges slowly to about 10V in 200mS.

C6 effectively swamps the Miller capacitance of T1. R16 limits the current drawn from IC2/2 and the gate of T1 charges to 10V in about 800pS, when the supply is at approx 30% of its peak value. The Mosfet is now switched hard on and the loads are activated. During this initial 800pus, T1 starts to conduct once the gate reaches about 3V, within about 200pus, and is effectively fully conducting within approx. 500pS. The load current increases slowly during this period and the device remains effectively in saturation.

As T1 switches on, a sharp negative transition can occur at the junction of R50-R51, the input to the feedback differentiator IC2/1.

The resulting positive swing on its output is blocked from the integrator by Dull.

When the combination of loads is to be switched off, the action is entirely different.

The #SWITCH output is deactivated at the next supply zero crossing.

The junction of R9-Rll immediately swings towards +2.5V forcing the output of IC2/2 to swing negatively to approx. +4.5V. This negative swing is limited by the action of D8. The Mosfet is still in a moderately conducting state. Integrator action then takes over and the output continues to swing to its quiescent low of -0.6V. As it does so, the voltage across T1 increases. The rate of rise of voltage is monitored by IC2/1, whose output modulates the speed at which the integrator operates, limiting the rate of change of voltage across T1 to something like that of the supply at its zero crossing.

Within 30mS Ti is switched hard off and the drive to the triacs is removed. Any current now flowing in the loads is purely residual, stemming from phase differences between them. As the current through any particular triac drops to zero that triac switches off and remains off. When there are only two triacs still conducting the current in both must drop to zero at the same time, and they switch off as well.

During this period the supply impedance is once again about 200K so even if there is a sufficiently high dV/dt to cause a triac to switch on erroneously, no damage will result.

Because the period during which T1 dissipates high power is relatively short only modest heatsinking is required in most applications. All triacs and the Mosfet were mounted on a common heatsink, only the Mosfet being electrically isolated from it. The invention has been illustrated in relation to a vending machine, but it may equally be applied to any apparatus which uses motors and/or solenoids. For instance, it may be used in domestic appliances such as dishwashers and washing machines. The switching circuit may also be embodied in the form of an integrated circuit.

Claims (15)

1. A power circuit for reactive loads comprising a power control element appearing, in effect, as a conductance in series with a load, controlled in such a manner as to reduce radiated or conducted interference, the power circuit including a control circuit to vary the conductance of the control element.

2. A power circuit according to claim 1 and in which the control input for the circuit is fed to a function generator, acting as said control circuit, and the output of the function generator is fed to a summing device whose output is fed to the power control element, the output of said element being fed to a threshold detector device whose output provides a feedback signal which is fed to the summing device.

3. A switching circuit for reactive loads, said circuit including at least one power control device and a control circuit, the control device in effect appearing as a resistance in series with the load and acting as a main control device, or as a buffer for a main control device, whereby switching action is gradual, allowing current in the reactive load to decay slowly, the power control device being used as a power modulator, i.e. as a linear device, during turn-off to gradually reduce power.

4. A switching circuit according to any of claims 1 to 3 when applied to an inductive load.

5. A switching circuit according to any of claims 1 to 3 when applied to a capacitative load, the power control device being arranged to control the current rise at turn-on and to perform as a zero-crossing switch at turn-off.

6. A switching circuit according to any of claims 1 to 5 and in which the circuit is an AC circuit and the main switching device is a triac.

7. A switching circuit according to any preceding claim and in which the power control device is placed in series with the DC terminals of bridge rectifier and the load is placed in series with one of the AC connections and one of the supply terminals, the other supply terminal being connected to the other bridge terminal.

8. A switching circuit according to any preceding claim and in which the control circuit is supplied from a floating power supply.

9. A switching circuit according to any preceding claim including a diode bridge feeding an opto-isolator through a resistor to provide zero-crossing pulses for synchronising the device switch-on and switch-off to the supply.

10. A switching circuit according to any preceding claim adapted to control a plurality of loads, individual loads being selected by triac drivers without the use of conventional snubbers.

11. A switching circuit according to any preceding claim and in which the power control device is a Mosfet.

12. A switching circuit according to any of claims 1 to 10 in which the circuit is a low voltage AC circuit and the power control device is either a Mosfet or a bipolar transistor.

13. A vending machine incorporating a switching circuit according to any preceding claim, the loads being solenoids and motors which operate different parts of the vending machine.

14. A switching circuit substantially as hereinbefore particularly described and as illustrated in the accompanying drawings.

15. A switching circuit according to any of claims 1 to 10 embodied in an integrated circuit.