JPS61104220A - Trouble-shooting circuit of body movement detecting apparatus - Google Patents

Abstract

PURPOSE:To accelevate calculation speed, by trouble shooting by connecting in a single address informations obtained by A/D conversions of a sine wave signals being generated by a body movement detecting apparatus and contrast table of trouble detecting informations. CONSTITUTION:A moving condition of a servo-motor 1 is converted to 2-phase pseudo sine wave signals A, B accompanied by a 90 deg. phase difference by an encoder 2. Outputs A, B of the encoder A, B are converted into digital magnitudes by No.1 and No.2 A/D convertors 3, 4. These signals is led into an address input of a ROM5. A contrast table of trouble detecting informations is stored in the ROM5 and trouble shooting is conducted by contrasting with address in puts. A phase identification apparatus 6 by identifying a direction of the signals A, B and issues pulses for counting to an up-down counter 8. A latch 7 is set when a trouble of the encoder is identified.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to an object movement detector that outputs two signals with mutually different phases proportional to the amount of movement of the object as the object moves. The present invention relates to a failure detection circuit for an object movement detector that detects when an encoder, linear encoder, etc. is used to measure the amount of rotation or movement and the encoder has failed and the correct amount of rotation or movement cannot be measured.

In servomechanisms using DC servo motors, such as robots, for example, the amount of rotation of the servo motor is measured using an incremental rotary encoder, and this is feedhacked to determine the target value of position or speed. Comparative closed-loop control is widely practiced. When performing position feedback control or speed feedback control using the output of such a rotary encoder, if this rotary encoder breaks down and the amount of rotation of the servo motor cannot be measured, there will be no feedback signal. , the servo motor will run out of control from the next moment.

This situation is due to the fact that in the servo mechanism that converts the rotation of the servo motor into linear movement and controls the linear movement, the linear deviation is measured using a linear encoder, and this is fed back and compared with the target position or speed. The same applies to closed loop control.

Therefore, in rotary encoders and linear encoders, when a failure occurs in the encoder, it is always required to immediately detect the failure and take measures to prevent the servo motor from running out of control.

[Conventional technology]

Conventional methods for detecting runaway of a servo motor include installing a limit switch that detects when the maximum movement range has been exceeded, or when the absolute value of the deviation between the commanded position and the position measured by the encoder has become abnormally large. Methods are being used to detect this. With these methods, for example, in the case of the former, the machine has run quite a long time from the time it first runs out of control until it is detected, and even if the machine itself can be prevented from being destroyed, it may not cause harm to people or prevent work, such as robots. This is not sufficient if there is a risk of damaging the object. In addition, in the latter case, the amount of rotation can be reduced when the servo motor runs out of control, but this is only effective in the case of runaway due to causes other than encoder failure, and when the encoder is malfunctioning, the amount of rotation of the servo motor can be reduced. However, the absolute value of the positional deviation remains zero and cannot be detected.

Another possible method is to determine a failure by detecting the amplitude or phase shift of the output signal from the encoder, but this method requires separate detection circuits for each, which has the disadvantage of complicating the circuit configuration.
d In order to solve these problems, the applicant for this patent previously filed a patent application filed in 1986-14.
No. 3502 ``Failure detection circuit for object movement detector'' was filed. As shown in FIG.
An encoder 10 that generates (-Vcosωt) is used, and its output is transmitted to a deviation measurement circuit 20 and a failure detection circuit 30 to detect the deviation state and failure.

In FIG. 4, numeral 10 is an encoder, and two contradictory and negative sine wave signals 51 with a phase difference of 90 degrees 51=V 5ina+
Generate tSS2=Vcosa+t. Reference numeral 20 surrounded by a dotted line is a deviation measuring circuit, which determines the direction of rotation in the case of a rotary encoder and the direction of movement in the case of a linear encoder, and supplies counting pulses to the next stage amplifier/down counter (not shown). .

A failure detection circuit 30 also surrounded by a dotted line is composed of a first multiplier 31, a second multiplier 32, a signal generator 33 and a failure discriminator 34 surrounded by a dotted line.

Of the above configurations, the encoder 10 and the deviation amount measuring circuit 11 are the same as known ones, so the configuration and operation of both will be briefly explained first.

Two-phase pseudo sine wave signal 51 (=V
The principle of the encoder lO that generates sinωt) and S2 (=V cosωt) is the same whether it is a rotary encoder or a linear encoder, so
The case of a rotary encoder will be explained below as an example.

A rotary encoder that generates two-phase pseudo sine wave signals with a phase difference of 90 degrees is called an incremental rotary encoder. As shown in FIG. 5 (al, mountain), the incremental rotary encoder includes a transparent glass-like code plate 12 that rotates around a shaft 11;
Fixed mask plate 13, light emitting element 14 and first and second
It has 15 and 16 light receiving elements (it also often has two light emitting elements). The code plate 12 is arranged on the circumference as shown in Fig. 6 (a).
), slits S + , S 2-5 Ir formed by a vapor deposition pattern at equal intervals on the surface
, 512-., and the mask plate 13 has first and second mask codes MCI and MC2 consisting of one to several slits at positions corresponding to the slits of the code plate 12. (Figure 6(a)
In actuality, the code plate 12 and the mask plate 13 overlap one another so that the respective slits of the two correspond to each other, but for convenience of explanation, the two are shown separated. ). The first mask code MCI corresponds to the code plate 12 as shown.
When the slit SL of the upper slit code SC matches the slit 32, the second mask code MC2 becomes as shown in the figure.
The slits Sll and SI2 are selected to be located at positions shifted to the right by exactly that width.

As a result, when the code plate 12 rotates, the first and second mask codes MC+, MC2 and slit code SC
The amount of light passing through the light emitting element 14 generates an output proportional to the product of the angular velocity of the code plate 12 and the number of slits. First light receiving element 1 that receives light from the first mask code MCI
The waveform outputted by 5 forms a triangular wave T+, as shown in FIG.
The triangular wave T2 output from the second light receiving element 16 that receives the light from C2 has a phase difference of 90 degrees. Each of these triangular waves T
When only each fundamental wave is extracted through a filter (not shown) of I and T2, a two-phase pseudo sine wave signal Ss (-V sinω
t) , Sl (-V cosωt) is obtained.

Note that since the angular velocity ω changes from DC to about 100 KHz·2π, a filter is not possible.

The pseudo-sine wave signal appears to be due to light diffraction or leakage, and the peak of the triangular waveform becomes rounded.

The actual output will be a distorted waveform somewhere between a triangular waveform and a sine waveform.

The deviation measuring circuit 20 includes a first and a second comparator 2.
1.22, phase discriminator 23 and zero volt reference voltage source S
It consists of Vo. First and second comparators 21
, 22 are the inputs Sl, S2 and the reference voltage source SV of zero volts.
Compare o and input S.

1. A 0-phase discriminator 23 that outputs rectangular waves R1 and R2 having the same phase relationship as Sl compares the phases of the rectangular waves R1 and R2, and determines that R2 is 90 degrees with respect to R1 (that is, Sl with respect to Sl). When there is a delay (for example, when the servo motor (not shown) rotates clockwise), a pulse for up-counting is generated at the U terminal, and when it is ahead by 90 degrees (rotating counterclockwise), a pulse for down-counting is generated at the D terminal. and supplies it to an up/down counter (not shown) in the next stage.

Next, explaining the failure detection circuit 30, the first multiplier 31 outputs one output 51 of the encoder 1o=Vsinω
V2sin2ωt is output by squaring t. The second multiplier 32 squares the other output 52-vcosωt of the encoder and outputs V2C0526)t.

The signal generator 33 includes an adder 35 and a square root circuit 36, but the square root circuit 36 may not be used. Adder 35 adds the outputs of first and second multipliers 31,32. Output S+, Sl from encoder lO
However, as shown in Fig. 3, the output during normal operation, that is, a two-phase pseudo sine wave signal S with an equal amplitude of V and a phase difference of 90 degrees.
When outputting a+t, the output of the adder 35 is v2
becomes constant. The square root circuit 36 finds the square root of the output v2 of the adder 35 and outputs an output signal TS (=V). It is possible to use v2 instead of V as the output signal TS, but in that case, the square root circuit 36 becomes unnecessary.

The failure discriminator 34 includes first and second comparators 37 and 38 and an OR circuit 39. The first comparator 37 has
As shown in FIG. 7), a first reference voltage source SVs having a predetermined level lower value than the normal amplitude V of Sl and Sl is supplied, and when the level of the output signal TS is lower than this SVs, Generate output. The second comparator 38 includes a seventh
(in the figure), the second value is a certain level higher than ■.
A reference voltage source SV2 is supplied, and an output is generated when the level of the output signal TS becomes lower than this SV2. The OR circuit 39 outputs a fault signal FS when an output is supplied from at least one of the first and second comparators 37 and 38. The first and second reference voltages #SVI and SV2 are different from each other when the output signal TS has the amplitude ■ of the signals Sl and S2 and when the output signal TS has the amplitude
It is set to a different value depending on the case, and how to set it will be explained in the following explanation of the operation of the failure detection circuit 30 in a failure state or an abnormal state of the encoder 10.

The encoder 10 may stop outputting or generate abnormal output due to the following causes, and the normal servo operation may not be performed.

(2) When the gap between the code plate 12 and the mask plate 13 changes.

The gap between the code plate 12 and the mask plate 13 (see Fig. 5(a)) varies depending on the number of patterns, but is approximately 100 μm±.
It is about 20 μm. This gap is, for example, shaft 1
If it fluctuates due to the axial deviation of 1, the 2-phase signal S+, 3
Although the phase difference between the two signals hardly changes and is 90 degrees, the amplitude and offset amount of the signal change greatly. For example, if the gap is narrow, the amplitudes of Sl and S2 will be large, and if the gap is wide, the amplitudes will be small, and the degree of change in the amplitudes of Sl and s2 will also be different.The offset amount and amplitude ratio of the outputs S+ and S2 will change. Then, the outputs 31 and 32 of the encoder 1o
The Lissajous waveform changes as shown in FIG. 8(a) and (bl).Here, when Pa is normal, Pl is 0.
When an offset of 5 V occurs, P2 is applied to Sl and S2. When an offset of 0.5 V occurs, P3 applies 0 to S2.
．． When an amplitude change of 5V occurs, P4 shows a Lissajous waveform when an amplitude change of 2V occurs in S2. Pl and P
In the case of P3, the first and second comparators 37, 38 both generate an output, and in the case of P3, the output is output from the first comparator 37, P4.
In each case, the second comparator 38 generates an output, but in either case, the OR circuit 39 outputs the failure signal FS.

■ When the rotation centers of the shaft 11 and code plate 12 are misaligned.

The eccentricity of the shaft 11 and the code plate 12 is Δ, and the code plate 12
When the radius to the slit is R, the angular error for one rotation of the comet plate 12 appears in the form of a sine wave with an amplitude of ±Δ/R. At this time, the two-phase pseudo sine wave signal S
Since the optical paths of t and S2 are installed with mechanical positions shifted, the phase difference between the two signals changes from 90 degrees with one rotation of the code plate 12 as a period. However, the amplitude and offset amount of both signals hardly change, but the shaft 1
When the center of rotation of the code plate 12 shifts due to the bending of the shaft (the gap also changes due to the bending of the shaft), the amplitude also changes.

In this case, the output signal TS changes asymmetrically and complexly upward and downward centering on ■ due to the phase difference fluctuation of 5tSS2.
Since the voltage changes above and below the level of the second reference voltage sources SV+ and SV2, the presence or absence of a failure can be easily discovered.

If the phase difference between the two-phase pseudo sine wave signals St and S2 is 90 degrees, the rotation direction can be accurately determined, but if the phase difference between both signals approaches 0 degrees or 180 degrees, the rotation direction cannot be determined. Go wild.

(2) When a relative positional shift occurs between the code plate 12 and the mask #1t13.

If the centers of the code patterns on the code plates 1 and 2 and the mask plate 13 are no longer concentric due to loose attachment of the mask plate 13 or wear of the bearing of the shaft 11, etc., 2.
The phase difference between the pseudo sine wave signals St and S2 deviates from 90 degrees. The phase shift at this time is a constant amount regardless of the rotation period of the code plate 12. However, the amplitude and offset amount of both signals do not change much. Figure 8(C) is
This shows the surge waveform when a phase shift occurs in the two-phase pseudo sine wave signals S+ and S2, and when Pa is normal,
P5 shows the surge waveform (Figure a) when S2 shifts by 30 degrees, P6 shows the waveform when S2 shifts by 60 degrees, and Pl shows the waveforms of the output signal TS (Figure b) when S2 shifts by 90 degrees. . Since the first and second comparators 37 and 38 generate outputs in all cases of ps, ps, and Pl, the OR circuit 39
outputs a fault signal -FS.

In addition to the above-mentioned causes, various cases can be considered, for example as shown in Fig. 8, but in any case, it appears as amplitude changes and phase shifts of the two-phase pseudo sine wave signals St and S2. Therefore, the presence or absence of a failure can be easily discovered by the failure detection circuit 30.

Furthermore, the values of the first and second reference voltage sources SV+ and SV2 are as follows:
It is statistically selected based on the abnormal state of the encoder that occurs due to these various causes and the waveform of the output signal TS at that time.

[Problem that the invention seeks to solve]

By the way, in the above failure detection circuit 30, q or A2
Since the calculation of +82 must be performed in real time while the servo motor is rotating at high speed, it is necessary to use a high-speed device that is impossible to perform when performing the above calculation with a digital arithmetic unit such as a microcomputer. shall be.

If a high speed digital multiplier could perform this calculation in real time, it would be very expensive; Also, analog computing units (
When an analog multiplier, square root calculator) is used, the device becomes larger and moreover expensive.

Furthermore, analog multipliers and square root operators have off-cent fluctuations, so it is necessary to adjust the offset of individual components using variable resistors. Therefore, not only does the amount of adjustment work increase, but the offset level fluctuates due to changes in the temperature of the offset, changes in the amount of offset over time, etc., and for example, the comparison level becomes unstable, causing problems in the reliability of the failure detection circuit 30 itself. was there.

[Means for solving problems]

In order to solve the above-mentioned problems and provide a more compact, low-cost, and highly reliable object movement detector, the failure detection circuit of the object movement detector of the present invention detects the amount of movement of the object as it moves. In a fault detection circuit for an object movement detector that outputs two sinusoidal signals having different phases according to the signal, an analog-to-digital converter for converting the sinusoidal signal into a digital quantity is provided, and the two-phase sinusoidal signal is The present invention is characterized in that a storage means is provided in which a correspondence table is recorded that uniquely associates failure detection information with information converted into digital quantities. ′ [Function] This allows processing to be performed digitally, which not only speeds up calculation speed (and facilitates real-time control), but also reduces fluctuations in offset levels, etc. because it can be processed by digital circuits. It is possible to perform accurate control without any problems.

〔Example〕

An embodiment of the present invention will be described based on FIGS. 1 and 2.

Figure 1 is a configuration diagram of an embodiment of the present invention, Figure 2 (al is a configuration diagram of its ROM (Read Only Memory), and figure 2 (bl is a locus when ROM addresses are selected as the motor rotates). .

In the figure, l is a motor whose movement state is controlled and monitored by a servo motor, 2 is an encoder corresponding to the encoder 10 in FIG. 3, and 3 is a first analog-to-digital converter (hereinafter referred to as 4 is a second A/D converter, 5 is a ROM, 6 is a phase discriminator which corresponds to the phase discriminator 23 in FIG. 3, and 7 is a latch and an encoder. 8 is an amplifier down counter to which the encoder failure detection signal FS is set when a failure occurs.

In the present invention, first, the two-phase output of the encoder 2 is
D-convert them into digital quantities, and then convert these signals into RO
By leading to the address input of M5, a failure determination of the encoder is made based on the correspondence table written in this ROM5.

The encoder 2 is, for example, an incremental rotary encoder, and is coupled to the rotating shaft of the motor 1, and constitutes a sensor that detects the rotation angle of the motor shaft. The rotary encoder 2 outputs a sinusoidal signal of several hundred to several thousand cycles per revolution of the motor shaft as two-phase signals A and B having a phase difference of 90 degrees. These signals A and B can be expressed by the following equations.

B=Vsina+t Here, ω is proportional to the rotational speed of the motor, and the sign is + for forward rotation and - for reverse rotation.

Next, these signals A and B are converted by the first A/D converter 3 and the second A/D converter 4 into digital quantities A' and B' in a range including -V to +V. Now, if these are 6-bit analog-to-digital converters, the range of analog quantities that completely includes -V to +V is (000000)2.
Stones can be made to correspond to six levels of digital quantities from ~(111111)2. And these 6 bits A', B
' data is combined into 12-bit data, and the ROM
In this case, ROM5 is 2.
The encoder failure detection signal FS, which is 1-bit output information, corresponds to -4096 addresses.

Details of the ROM 5 will be explained in more detail with reference to FIG. 2(a). A failure detection signal correspondence chart is written in ROM5, and the column of this correspondence chart shows the value of A' (00000
It is divided into 64 parts from 0)2 to (111111)2,
Also, the row of the corresponding chart shows the value of B' again (000000
)z to (111111)2 is divided into 64 parts. The values “O” and “1” shown in this chart indicate the 1-bit failure detection information FSO value that is the output of the ROM, and the values inside the two concentric circles centered on the origin are all “
0” is entered, and the outside of the two concentric circles (Fig. 2 (a)
) are all filled with "1". Here, "0" indicates that the encoder 2 is operating normally, and "1" indicates that an abnormality has occurred in the encoder 2. Note that the concentric circles in Fig. 2(a) correspond to S in Fig. 7(b).
Although it corresponds to V+ and SV2, it is determined by a method described in detail later. By the way, this correspondence diagram is a matrix, and while the area occupied by one address is a tiny square, the boundaries of the encoder failure detection information values are arcs, and the failure detection information of the addresses on the boundaries are arcs. A correspondence table can be created by setting the value to the one with the larger area.

At a hidden instant, the rotation of motor 1
.

Only one of the 4096 addresses in the ROM 5 is selected corresponding to the corner, and one value of the failure detection information corresponding to this is selected. Therefore, if this failure detection information FS is "1", it will be held in the lunch 7 and an alarm signal will be output to notify the encoder failure.

Here, as shown earlier, the signals A and B are A: V
sin (6) t + riB: V sinω
t, the address of the ROM 5 is selected counterclockwise in the case of forward rotation and clockwise in the case of reverse rotation on the circular locus shown in Fig. 2 (bl).Encoder 2 operates normally. As shown in Figure 2), this circular trajectory passes only where the value of the correspondence table in ROM5 is "0", so the output FS of that ROM5 outputs only the value of "0". do.

Incidentally, the phase discriminator 6 and up/down counter 8 in FIG. 1 are means for measuring the amount of rotation of the motor, and since they are well known, they will be briefly explained. The phase discriminator 6 in FIG. 1 corresponds to the phase discriminator 23 in FIG.
, Q correspond to R1 and R2 in FIG. 4, and signals up and dN correspond to U and D in FIG. In addition, the functions of the comparators 21 and 22 in FIG. 4 are the functions of the first A/D converter 3 in FIG.
, the second A/D converter 4 and the ROM 5. Therefore, according to another correspondence table that overlaps with the correspondence table in Figure 2 (al), when A≧0, P=“1” When A<0, P=“0゛When BaO, Q=’1” B< When it is 0, it should be associated with Q=“0”.Moreover, when the signal A or B crosses Ov, the phase discriminator 6 selects u according to its rotation direction.
A p pulse or a dN pulse is generated, and the counter 8 counts this.

Therefore, when encoder 2 is operating normally,
At the point where the circular trajectory (in Figure 2) crosses the A=0 or B=0 line, that is, at the black circle mark in the same figure, an up pulse is generated when the trajectory is counterclockwise, and d when it is clockwise.
N pulses are generated, and the rotation angle of the motor l can be measured faithfully.

In other words, if the circular locus shown in Figure 2 (bl) does not pass through each of the four quadrants once during one period of the encoder two-phase sinusoidal signal, the amplifier down counter 8 will detect the correct rotation angle of the motor 1. Don't show (naru)

Here, the configuration of the phase discriminator will be briefly explained with reference to the ninth drawing. As shown in FIG. 9 (al), this phase discriminator has a rectangular signal output section composed of four flip-flops FFl-FF4, 21-wire inverters INI, IN2, and four AND circuits a1 to a4. (As shown in bl, there is an up pulse output section consisting of four AND circuits a5 to a8 and an OR circuit ORI, and (as shown in C1, there is an up pulse output section consisting of four AND circuits a9 to a12 and an OR circuit OR2.
It is composed of an N pulse output section and the like. Figure 9 (al
As shown in the figure, when signals P and Q are input, inverter I
Hundred, stone and AND circuit al output from NL, IN2
The rising and falling signals of P and Q output from -84 are shown in (b) and (c) of the up pulse output section and d
Each AND circuit a5 to a8, a9 to a1 of the N pulse output section
2 in the illustrated state, an up pulse and a dN pulse are obtained from OR circuits ORI and OR2, respectively. The relationship between the signals P and Q and these up pulses and dN pulses is as shown in FIGS. 9(d) and (e).

(al to (f) in Fig. 3 are all ROM in Fig. 2 when the up/down counter 8 cannot count correctly.
It shows the trajectory by which addresses are selected.

Here, Fig. 3 (al) shows the case where there is loosening of the mask plate mounting screw and offset fluctuation, and Fig. 3 (b) shows the case where the gap between the code plate and the mask plate widens due to loosening of the bearing in the thrust direction. (C1 indicates when there is a dropout due to dirt or scratches on the mask board or code board, (d) indicates when the B phase output line is disconnected, and (e) indicates that all wires are disconnected or disconnected due to disconnection of the connector.) (f) shows a case where the - power line is disconnected. In these cases, as is clear from Fig. 2, it is possible to detect a failure in the encoder 2, and the ROM5 Therefore, there are times when the encoder failure detection signal FS becomes ``11''.
If so, the latch 7 shown in FIG.
” and thereafter holds “1” until a reset input is received, and outputs an alarm signal indicating encoder failure.

Especially when performing so-called servo control in which the measured value of up/down counter 8 is fed back to motor 1, if there is an error in this measured value, motor l may run out of control. When the output of latch 7 is "L", servo control should be immediately interrupted.

Further, FIG. 3(g) shows an eccentric state due to loosening in the radial direction of the bearing supporting the code plate of the encoder. A,
When B is detected differentially, offset deviations are less likely to occur and failure conditions in such cases are difficult to detect, but the present invention can detect even in such cases.

Figure 3 (hl indicates a state in which the gap between the mask plate and the code plate has widened due to loosening of the bearing in the thrust direction.In this case, if A and B are detected differentially, the offset deviation will occur. Although it was difficult to detect faults due to the difficulty in detecting them, it is now possible to detect them even in such cases.

FIG. 3(i) shows a state in which the gap between the mask plate and the code plate is narrowed due to loosening of the bearing in the thrust direction. Note that this trajectory is located at @1 outside the large circle in Figure 2.
’ region.

Even in such a case, when A and B are detected differentially, offset deviations will not occur, and faults that were difficult to detect can now be detected.

In Fig. 3 (IE) ~ [υ], if the cause of the change in the encoder two-phase output is the loosening of the bearing that supports the code plate of the encoder as described above, or the loosening of the screws that attach the mask plate, then , operation may have to be stopped before the loosening progresses further and the slit pattern is damaged. In addition, there are cases where a loose motor bearing is pushing the encoder's rotating shaft, and in this case, it is important to stop the rotor and stator of the motor before they come into contact. In other words, once loosening begins, the loosening progresses rapidly, so even if it is required to stop operation before a fatal failure occurs, early detection is possible using the sensitivity of the encoder. .

For this reason, "0" and "11" in the correspondence table in Figure 2
Although it is desired to increase the detection sensitivity by reducing the difference in the radius of the two concentric circles that form the boundary between the It is necessary to allow for the reduction in amplitude due to the frequency characteristics of the photodetector at the time of the test, and fluctuations in the offset of the signal due to the ambient temperature, etc. Therefore, the radius of the concentric circles in Figure 2, which represents the failure detection sensitivity, is determined as a compromise between the two. It will be done.

Further, although two concentric circles are set in FIG. 2, for example, only the smaller concentric circle may be used, and the inside of the circle may be set as "1" and the outside as "O".

If only the smaller concentric circle is used, as shown in Figure 3 (1), abnormalities due to the narrowing of the gap between the code plate and the mask plate cannot be detected, so the effect is less than when using two concentric circles. You can say it's small.

However, when using a flexible coupling between the motor shaft and the encoder shaft, no undue force is applied to the encoder shaft, and abnormalities such as narrowing of the gap are less likely to occur, so consider using only the smaller concentric circle. It will be done.

In addition, in the above explanation, the number of bits of encoder failure detection information is 1 bit, and two types of output, normal and abnormal, are explained. However, if the output of the ROM is set to multiple bits, for example, 2 bits, By distinguishing between when the person is inside the concentric circle and when the person is outside the larger concentric circle, it may be possible to distinguish the type of abnormal posture to some extent. Also,
Concentric circles are placed further inside the small concentric circles in Figure 2 4J
It may be classified into cases where operation is continued with only a warning and cases where operation is stopped. Similarly, further concentric circles may be provided outside the large concentric circle shown in FIG.

Of course, when the output of the ROM is made into multiple bits, the above P, Q
can also be included.

Also, in Fig. 2, for ease of explanation, the signals A and B are 0 (
The explanation is about the case where the center can swing positive or negative like ±V (V), but of course it is also possible to not center around 0 (VJ), and in that case, the first A/D converter and the second A/D converter
It is sufficient to simply shift the conversion range of the analog input signal of the A/D converter.

As described above, the present invention not only prevents the servo motor from running out of control due to encoder failure, but also detects abnormalities in the mechanical system at an early stage and stops operation before a fatal failure occurs. It is possible to detect things quickly and with high reliability using a simple circuit configuration.

〔Effect of the invention〕

As described above, in the present invention, the pulse generation system for the counting operation of the amplifier down counter 8 shown in FIG. 1 and the encoder failure detection system are separated in the ROM 5,
Moreover, since both processes are digital processing, the conventional case shown in FIG. 3 has much higher reliability than the case where the encoder failure detection system is constructed from an analog calculation circuit.

In addition, in the present invention, since encoder failure determination is performed only using ROM, the encoder failure detection system is much faster than the conventional case where the encoder failure detection system calculates % or A2 + 82 using a digital calculator and determines the value. And it can be constructed at low cost.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram of an embodiment of the present invention, and FIG. 2 (al is RO
The configuration diagram of M, Figure 2 (b) shows that RO changes as the motor rotates.
The trajectory when the addresses of M are selected, FIG. 3 is an example of what is considered to be a failure according to the present invention, FIG. 4 is a conventional example, FIG. 5 is a block diagram of the encoder, and FIGS. 6 and 7 are FIG. 8 is an explanatory diagram of the operation when a failure occurs, and FIG. 9 is an explanatory diagram of the configuration and operation of the phase discriminator. In the figure, ■ is a motor, 2 is an encoder, 3 is a first analog-digital converter, 4 is a second analog-digital converter, 5 is a ROM, 6 is a phase discriminator, 7 is a latch, and 8 is an up/down - Indicates a counter.

Claims (1)

[Claims] 1. In a fault detection circuit for an object movement detector that outputs two sinusoidal signals with different phases depending on the amount of movement of the object as the object moves, the sinusoidal signal is converted into a digital quantity. An analog-to-digital converter for converting the two-phase sinusoidal signals into digital quantities is provided, and a storage means is provided in which a correspondence table is recorded that uniquely associates failure detection information with information obtained by converting each of the two-phase sinusoidal signals into digital quantities. Failure detection circuit for an object movement detector. 2. As the correspondence table, arrange the memory addresses in a matrix, set one or more virtual concentric circles that overlap this matrix, and set the values of the failure detection information using the concentric circles as the boundaries of the failure detection information. 2. A failure detection circuit for an object movement detector according to claim 1, wherein the failure detection circuit records information.