JPS60231203A - Automatic offset adjusting method of digital servo system - Google Patents

Abstract

PURPOSE:To improve the accuracy of adjustment by detecting that generated oscillation is the one generated when a positional command is ''0'' and a servo is in the steady state, counting up the frequency of the vibration and adjusting offset voltage so that counted frequency is maximized. CONSTITUTION:Deviation parallel data 6.10 and the information of the least signified bit 6.2 are obtained from deviation counters 6.1. The information of the least significant bit 6.2 is converted into a voltage by an F/V converter 6.3 and then converted into a digital value 6.5 by an A/D converter and the digital value 6.5 is supplied to a microcomputer 6.6. The microcomputer 6.6 reads the deviation parallel data 6.10 and the digital value 6.5 and automatically adjusts an offset digitanl signal 6.7 so that the vibration of the least significant bit of the deviation is made the maximum frequency to output the adjusted value as an analog offset voltage through a D/A converter 6.8.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention is a digital device that vibrates with a width equivalent to one bit of a position detector when the servo is locked (position input is constant and in a steady state). This paper relates to an automatic offset adjustment method for servo systems.

[Conventional technology]

Conventionally, offset adjustment in a digital servo system has been performed only when the deviation counter reaches a desired value. In this case, there is arbitrariness by one bit of the position detector, but this has not been considered in the past. Therefore, when the offset adjustment is set to a low margin, a phenomenon occurs in which the offset adjustment is deviated due to drift in the analog system.

Such a servo system is very common, and if a dead zone is provided where the error register is zero, no vibration will occur.

[Problem that the invention seeks to solve]

The present invention automatically adjusts the offset at the point where the offset adjustment margin is maximum (where one bit of vibration has the maximum frequency) in order to eliminate the shift in the eight-round offset due to drift, which is a drawback of the conventional technology. That's what I'm trying to do.

[Means for solving problems]

The present invention detects that the generated vibration is vibration when the position command is O and the servo is in a steady state, then counts the frequency of this vibration, and applies an offset voltage so that the counted frequency becomes the maximum. This will be explained in detail below.

FIG. 1 shows the configuration of a typical digital servo system. In the figure, (1, 1) is a command pulse, (1, 3) is a feedback pulse, and (1, 2) is a deviation counter that adds and subtracts these two pulses. The deviation is the digital output of a deviation counter (1, 4), and the D/A converter (1, 5) outputs an analog signal (1, 6) proportional to the position deviation as a speed command. (1, 7) are offset signals that are the sum of the input offsets of the D/A converters (1, 5) or the speed loop controllers (1, 9), and are equivalently added to the speed command. The speed loop controller (1, 9) serves as a motor drive circuit including a part of the power. (1,1
0) is a motor, (1, 11) is a TG (tachometer generator), and (1, 12) is a PG (pulse generator). (1, 13) is the speed feedbank signal, (1
, 3) are position feedbank signals. When calculating the speed signal from the pulse generator, TG (1, 11
) is unnecessary, and the velocity feedback signal (1, 13) disappears, becoming the velocity feedback signal (1, 13) from the pulse generator (1, 12) via the F/V converter.

The response waveform of the pulse generator uses two-phase pulses of A phase and B phase as shown in FIG. 2, but there is hysteresis at the edge of each pulse.

This hysteresis is the hysteresis of the comparator in the pulse generator circuit. Usually 1750~ for 1 pulse
It is about 1/100.

Speed system (speed loop controller (1, 9), motor (1, 10), TG (1, 11), PG (1, 12)
), and the response of velocity feedback (a system composed of l(1,13)) is approximated as a quadratic system, the model of one pulse vibration (0←→1 vibration) becomes as shown in Fig. 3.

(2) is an offset voltage (0<vo<1), and the speed command signal for a 1-bit deviation is normalized to l.

When the position is 0, v01; when the position is 1, the positional deviation is -1, and a speed command of -1+V0 is generated.

Depending on the value of Vo, the vibration waveform of one pulse has various shapes as shown in FIG. Figure 4 ta+ is V. -1,
Figure 4(b) shows the case of Vo #0.5 (7).
01 is V. This is the case of #0.

If vo is -1<VO<1, the vibration will be between 04+-1 (FIG. 5). When vo=0, Figure 4 (al
(7) When vo = o, s, as shown in Fig. 5(b), ■. When #-1, the result is as shown in FIG. 5(C).

The offset value changes due to the drift of the input amplifier of the D/A converter or speed loop controller, but in the case of Figure 4 (al) and Figure 5 (al), the value of the offset changes by 1. If a drift exceeding 1 is added, Although the drift is extremely small, the vibrations of 1++2 and Qel and Fig. 4 (
C1 and FIG.
-1 vibration, resulting in a positional deviation of exactly one pulse. When the conditions are 4) and 5), 0.
This means that there is a margin for drift of 5, which means that the margin for preventing positional deviation is the largest. Furthermore, the vibration frequency is at its maximum at this time.

According to the above explanation, when the vibration frequency at the time of servo lock is the highest, the margin for preventing positional deviation due to drift is the largest.

FIG. 6 shows an embodiment of the automatic offset adjustment circuit. (6
, 1) is a deviation counter from which information of deviation parallel data (6, 10) and least significant bit (6, 2) is taken. (6, 3) is an F/V converter, which converts the frequency of the least significant bit of the deviation counter (6, 2) into voltage. (6, 4) is the A/D converter, and the deviation counter (6, 1
) is used to convert the vibration frequency of (6, 2) into a digital quantity. (6, 6) performs the functions described later in the microcomputer.

(6, 7) are offset digital signals, which are output as analog offset voltages by the D/A converter (6, 8).

Automatic offset adjustment can be performed using the following procedure.

■Fix the command value input to make the servo locked.

■Wait 2 to 3 seconds to reach a steady state.

■The deviation counter (6, 1) is converted to the deviation parallel data (6, 1).
, 10) into the microcomputer (6, 6). (2) Output the offset voltage so that the deviation becomes the desired value (normal deviation = 0).

0 Automatically adjust the vibration of the least significant bit of the deviation to the maximum frequency using the method described next.

FIG. 7 shows the procedure for automatically adjusting the least significant bit of the deviation.

First, the parameter N is set to 0. Set the offset value to +Δ(
Try changing only 〉0). Measure the frequency at this time,
If the offset value has increased from before the change, N=1 and the offset value is further changed by +Δ. If the frequency decreases when changing the offset, determine the value of N, and set N=1.
If so, the adjustment is complete. When N=O, set the offset to -Δ(
Δ〉0). At this time, if the frequency increases, N=1 and the offset is further changed by 1 Δ.

If the frequency decreases when the offset is changed by -Δ, the value of N is determined, and when N=1, the adjustment ends, and when N=0, the process returns to the beginning of the offset adjustment.

[Application modification example]

■Realize frequency counting as a digital circuit or microcomputer operation.

■Replacing functions performed by microcomputers with digital circuits.

〔Effect of the invention〕

As described above, according to the present invention, offset adjustment can be performed automatically, and the offset value is adjusted to the point with the maximum margin to prevent positional deviation due to drift. This makes it possible to avoid a one-pulse deviation of , and it is possible to realize a digital servo system with higher precision than the conventional one.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the configuration of the digital servo system for motor control, Fig. 2 is an explanatory diagram showing the response of the pulse generator, Fig. 3 is a block diagram showing a model of l-pulse vibration, Fig. 4 and FIG. 5 are waveform diagrams showing aspects of one-pulse vibration, FIG. 6 is a block diagram showing a configuration example of an automatic offset adjustment circuit according to the present invention, and FIG. 7 is a flowchart of offset adjustment. (1, 1): Command pulse (1, 2)! Deviation counter (1, 3): Feedback pulse (1, 4): Deviation signal (1, 5): D/A converter (1, 6): Speed command (1, 7): Offcent signal (1, 8) : Adder (1, 9) : Speed loop controller (1, 10)
: Motor (1, 11) : Tachometer generator (1, 12)
: Pulse generator (1, 13) : Speed feed pulse signal Patent applicant Yaskawa Electric Manufacturing Co., Ltd. Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 6 g7 Figure 1+-

Claims (1)

[Claims] 1. In a digital servo system where the position loop is digitally processed and the velocity loop is analog processed, it is detected that the vibrations occurring are vibrations when the position command is 0 and the servo is in a steady state. An automatic offset adjustment method for a digital servo system is characterized in that the frequency of this vibration is further counted and the offset voltage is adjusted so that the counted frequency is maximized.