JPH04156283A - Apparatus for controlling speed - Google Patents

Abstract

PURPOSE:To keep a rotation of a rotating body constant by calculating the deviation of an excitation, for the rotating body having FG's of low excitation accuracy, every one of the FG's, and by storing the calculated deviation, and further, by correcting the deviation through a software. CONSTITUTION:Based on values inputted to a sequential selector 2f for FG's, data are calculated by a computing element 2b, a memory for correction values and a correction starting switch 2d, and then, the data are converted into analog quantities by a D/A converter 2e, and thereby, a motor 4 is driven. When the motor is rotating, the outputs of the FG's are monitored. Every one of the n FG's, its speed deviation Tj from a reference speed value T0 is measured fixed times. Then, the average values of the measured Tj's are calculated for the n FG's, FG1-FGn respectively, and the average value is stored in a memory every one of the FG's. Then, on and after the completion of the foregoing computation, using the calculated average value alphaj as a correction value, whenever an error output is calculated for one of the FG1-FGn, by subtracting the correction value alpha of the FG from the error output of the FG, even when the error output is different from others every one or the FG's, it can be corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a speed control device for a rotating body having an FG multiplied signal.

2. Description of the Related Art An example of a conventional speed control device will be described below with reference to the drawings.

FIG. 7 is a block diagram showing a speed control device that controls motor drive, in which 1 is a shaper that converts the AC waveform generated in the motor windings into a rectangular FG waveform, 2 is a microprocessor, and 3 is a driver. , 4 is a motor (hereinafter referred to as a rotating body).

The operation of the conventional speed control device configured as described above will be explained below.

First, the rotating body 4 generates an FG multiplied signal by rotating and uses it as speed information.

In the waveform shaping step 1, the AC waveform is amplified into a rectangular wave. This is input to the microprocessor 2 as rotational speed information.

In the microprocessor 2, when the FG multiplied signal rising edge is input to the period detector 2a, the value of the internal counter (free run counter) at that time is latched. Next F
When a G-times signal is input, the latch value is updated.At that time, the period T is calculated by the 2b calculator from the difference between the value before and after the update, and the period T is calculated from the value T and the reference stored in the internal memory. The values T0 are compared and the difference ΔT=T−To is output as an error. Since this value is a digital quantity, it is converted into an analog quantity by the D/A converter 2c and sent to the motor driver 3 through a filter. The motor driver 3 drives the rotating body 4 in response to the error.

However, with this configuration, if the magnetization accuracy of the FG magnet is poor, the rotating body will rotate unevenly because it will be rotated so that the FG multiplied signal is 7 constant. FIG. 8 shows the waveform at that time. (A) is the FG specific output from the rotating body (
B) is the FG waveform shaped into a rectangular wave, and (C) is the error output ΔT.

On the other hand, as a correction method for magnetization unevenness that occurs alternately, if T1 and ΔT2 are always approximately constant to the error output, ΔTH is set as a correction value from the average of ΔTl and ΔT2,
A method of subtracting ΔTH from the error output of each FG to reduce ripples has also been considered.

Problem to be Solved by the Invention However, in the above correction, as shown in FIG. 8(D), the error output is ΔT+, ΔT2, . . . for each FC.
...If it differs greatly from ΔTo - the error of each FG cannot be corrected with one correction value ((A) is the motor's FG ratio output (
B) is the FG waveform shaped into a square wave, (C) is the FG
error when the error was approximately constant).

The present invention is a solution to the problem of FG unevenness in each FG, and also provides a speed control device for performing more precise rotation control.

Means for Solving the Problems In order to solve the above problems, the present invention monitors the output of the FG during rotation, and calculates the deviation ΔTj (where j=1 to n) of the speed from the reference value To for each of n FGs. A certain number of times (m times)
Measure. Then, the average value is calculated for each of FGI to FGn, and the value is stored for each FG. After the calculation is completed, the value α, (j=l~n)
Each time the error output is calculated for each of FGI to F G n, the correction value for that FG is reduced, so that even if the error output differs for each FG, it can be corrected. It is something.

According to the above configuration, the present invention calculates and stores the magnetization deviation for each FG and corrects it by software even if there is a magnetization deviation in each FG for a rotating body with poor magnetization accuracy of the FG. This allows the rotation of the rotating body to be constant.

EXAMPLE Hereinafter, a speed control device according to an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of one embodiment of the present invention. In the figure,
The FG output of the rotating body 4 is input to the waveform shaper 1. The output of the waveform shaper 1 is input to the microprocessor 2,
At the same time as the period is measured by the period detector 2a, it is input to the FG order selector 2f. Based on that value, 2b arithmetic unit, 2
The data calculated by the c correction memory and the 2d correction start SW are converted into analog data by the D/A converter 2e, and supplied to the motor driver 3 to drive the motor 4.

Here, the configuration of the period detector 2a will be briefly described.

The microprocessor has an internal free-running counter called a reference counter (time base counter), and when the FG rising edge is input, the count value at that point is latched by hardware.This latched value is the FG
It is updated every time the rising edge of etch arrives.

Furthermore, the calculator 2b temporarily stores this latched value in memory, and calculates the period from the difference between the latch value updated by inputting the next FG rising etch and the value stored in the memory immediately before. The error value based on the deviation from the reference value is output as a digital value.

After this operation, the speed is controlled by updating and storing the updated latch value in the memory and repeating it.

Since the output of the arithmetic unit 2b is a digital quantity, it is converted to analog by the D/A converter 2e, and then sent through a filter to the motor driver 3, which drives the motor to accelerate or decelerate depending on the error.

FIG. 2 is a detailed explanatory diagram of the present invention. (A> is the speed information signal FG generated by the motor, and (B) is a waveform shaped signal FG into a rectangular wave, which is input to the period detector 2a and the FG order selector 2f.

Here, as shown in Figure 2 (B), there is only one square wave with a high level, but this is the one on which the phase detection signal (PG) of the rotating body is superimposed, and it is transmitted once per rotation. appear. Based on this as shown in (B), from FGI to F
Input addresses from O to 5 in order up to G6. This address is then recognized by the FG order selector 2f in FIG. 1, and correction is made for each FG.

(C) is an error output waveform for FGI to FG6, and errors of different values are output for the FG multiplied signal and the reference. If the FG accuracy was originally good,
The value should remain almost constant, except that an error that fluctuates up and down is output only when a disturbance occurs relative to the reference. The reason why a fixed error always occurs in each FG is considered to be due to an accuracy error rather than the influence of a disturbance, considering that the disturbance suppression characteristic does not extend to 1/2 of the FG frequency.

However, such an error generates unnecessary torque in the rotating body and must be eliminated. Therefore, the error is monitored for each FG, and the error amount that has occurred is subtracted from each FG as a fixed value, t, to correct it so that it is constant with respect to the reference.

FIG. 3 is a flowchart showing the correction method. 3
At step 01, it is determined whether the mode is correction mode. In other words, it monitors that the rotation of the rotating body has become constant, and confirms that the phase system has entered the control region after the speed system has withdrawn.

302 waits for the first FGI to arrive as a starting point for correction. In addition to being counted by the FCC (FG counter) in the correction process, the FG is reset immediately before the FGI in the ISW (head switch) process, so the count from 0 to 5 is always repeated. Monitor the FCC and FG
The first arrival of I is determined. Once the FGI is received, this judgment will not be made thereafter. 303 clears the memory used for correction from the time the power is turned on until the rotation becomes constant. 304 is a learning mode determination, and if learning is not completed, it moves to the learning processing process. If it is completed and the correction value is determined for each FG, 306 checks the number of FG from FGCI 307
The correction values calculated in , respectively, are subtracted.

ERI is the error itself calculated from the edge period detection of FG, and is FGH.

(j=1 to 6) is the error average value for each FG, and this becomes the correction value. This is the value stored in the average value memory 2c in FIG. As described above, 308 increments the FCC (FG counter) every time a correction calculation is performed and waits for the next FG address.

FIG. 4 is a flowchart of the learning process described above. Here, the average value of the deviation from the reference is calculated for each FG a predetermined number of times (in this case, 16 times) after the power is turned on and the rotation becomes constant.

401.402 searches the FG number and calculates the cumulative error FGMj=ΣΔTi for each FG.

(where i indicates FG addresses 1 to 6) is determined and stored in memory. This is the memory corresponding to 20 in FIG.

403 is to monitor that when calculating the deviation of each FG from the standard, the phase system of the rotating body is drawn in and the rotation is constant. If unevenness occurs, clear all the memory addresses used for correction, memory contents, etc., wait for the rotation to become constant, and after determining that the rotation has become constant at 301 in Figure 3, start the correction. It restarts the value operation from the beginning.

This protects the correction value from entering an abnormal value. Further, in step 404, the number of averages is incremented, and in step 405, the number of averages is counted.

Then, when the predetermined number of times (in this case 16 times) has been reached, 4
In step 06, the average value (where i is the FG address, i-1 to i-6) is determined, stored in memory, and used as a correction value in step 307 in FIG. Furthermore, after phase system offset processing is performed in step 407, since learning has been completed, a flag is set, and based on this flag, it is determined in 304 of FIG. 3 that this process will not pass (and this process will be completed).

FIG. 5 is a flowchart showing the process of phase system offset processing in step 407 shown in FIG. In 501, the average value for each FG is further added, and 50
2, the average value FGS is further calculated. Then, in 503, FGS is subtracted from the correction values FGHI to FGH6 of each FG, that is, the offset value is subtracted to calculate only the amount of deviation from the reference, and this is used as the correction value for each FG.

This is because, in the present system that mixes the one-phase speed system and supplies it to the driver, if this processing is not performed, the problem will occur that the error will vary greatly after correction, causing rotational unevenness.

Its operation will be explained with reference to FIG. To is an internal reference value of the speed system. α1 to αG are average values of deviation amounts for each FG. This system calculates errors by mixing the phase system and velocity system, so as shown in Figure 8, the error that would normally occur around T0 is α due to the balance with the phase system.

An error may occur based on the deviated T'o. At this time, if α1 to α6 are used as they are as correction values, the interrogation error with the correction will drop to around To, causing a large change in C. The only necessary correction value is the amount of deviation from the reference value (, the correction value in FGO is T゛.Target average -l J α1-α (however, if calculated as α.-□) Good.This will reduce the error at the measurement point to T.
−T, −(α1−α.).

As a result, the error variation caused by the correction is only the deviation from the velocity system standard, and continuity before and after correction is maintained while considering the phase system.

Effects of the Invention As described above, the present invention provides for a rotating body having an FG multiplied signal, a period detecting means for detecting the period of the FG, a memory means for storing the detected value T of the period detecting means, and a period detecting means for storing the detected value T of the period detecting means.
a- An error calculating means for calculating an error output from a reference value To, and a driving means for supplying driving power to the rotating body based on the error output, and the error calculating means includes: n generated FG times signals and a period detection value T stored in the memory means; (j=1・
. . n) and the reference value To, calculate each deviation amount ΔT, and further store the deviation amount and the value in the FG multiplied signal as a correction reference value α1; , according to the correction reference value α, the device is characterized by comprising a correction means for correcting the error output calculated by the arithmetic unit for each of the n FG multiplied signals, so that the FG multiplied signal magnetization error is adjusted for each FG. Even if the speed error output is different from F, using the correction reference value,
By making corrections for each G, a very large effect is achieved in that more accurate control can be realized. In addition, if irregular rotation occurs due to an out-of-range amount during correction value calculation, the calculation can be stopped and the calculation can be restarted after the rotation has reached a certain level, thereby preventing errors in correction value calculation.Furthermore, when calculating the correction standard, □ Even when the reference has an offset due to the relationship with the phase system, the new correction reference value α for each FG of correction reference value α1
By having an arithmetic unit that calculates 'J (=α, -α.), uneven rotation can be prevented before and after correction.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the speed control device of the present invention, and FIG. 2 is a waveform diagram explaining the operation of the device.
3, 4 and 5 are flow diagrams explaining the operation of the same device, FIG. 6 is a waveform diagram explaining the operation of the same device, and FIG. 7 is a block diagram showing a conventional speed control device.
FIG. 8 is a signal waveform diagram of the same device. l...FG waveform shaper, 2...microprocessor, 2a...period detector, 2b...
...Calculator, 2c...Correction memory, 2d
...Correction start SW, 2e...D/'A
Converter, 2f...FGJ [Character selector, 3...
...Morta driver, 4...Motor. Name of agent: Patent attorney Haruaki Koebi and two others

Claims (3)

[Claims] (1) Period detection means for detecting the period of an AC signal (hereinafter referred to as FG signal) having speed information of n rotating bodies per revolution, and memory means for storing the detection value T of the period detection means , an error calculation means for calculating an error output from the detected value T and a reference value T_o, and a drive means for supplying driving power to the rotating body based on the error output, and in the error calculation means, the rotating body The period detection value T_j (j=1) stored in the memory means every n FG signals generated during one rotation of the
・・・・・・n) and each deviation amount ΔT from the reference value T_o
an arithmetic unit that calculates _j, and further calculates the m-time average value of the deviation amount for each FG signal ▲There are mathematical formulas, chemical formulas, tables, etc. ▼; and a memory means that stores the value as a correction reference value α_j. and a correction means for correcting the error output calculated by the arithmetic unit for each of the n FG signals using the correction reference value α_j. (2) A means for detecting that the rotation of the rotating body is no longer constant due to disturbance etc. during correction reference value calculation and stopping the calculation, and a means for determining that the rotation has become constant and storing it from that point. The speed control device according to claim 1, further comprising an arithmetic unit that clears the contents and performs the calculation again. (3) When calculating the correction standard value, calculate the average value α_o=▲of the correction standard value α_j for each n FG (numerical formula, chemical formula, table, etc.), and calculate the correction standard value α_j and this average value@ The difference from α@_o (α′_j−@α@o_) is calculated as the new correction reference value α_j(=α_j−@α@) for each n FG.
2. The speed control device according to claim 1, further comprising an arithmetic unit that calculates the speed as _o).