JPH02168113A - Detecting device of rotational angle - Google Patents

Abstract

PURPOSE:To attain a high resolving power and a high speed by setting the amount of correction so that pulse signals outputted from first and second R/D (resolver/digital) conversion means have a phase difference of 0.5 pulse approximately between them. CONSTITUTION:An output of a detecting element 31 is adjusted with respect to the unbalance of a magnetic circuit of the detecting element 31 and the nonuniformity in DC resistance and the like of a coil thereof by variable resistors VR1 to VR4 of an adjusting element 32 so that the height of a peak value of an output voltage to be adjusted, and the value thus adjusted is supplied to amplifiers 33 and 34. Next, outputs of the amplifiers 33 and 34 are given to R/D converters 35 and 36. Subsequently, the amounts of correction of output signals of the converters 35 and 36 are set by correctors 37 and 38 so that they have a phase difference of 0.5 pulse approximately between them. Moreover, output signals of the correctors 37 and 38 are converted into analog signals by D/A converters 39 and 40. Then, the lowermost-bit output signals a0 and b0 out of the output signals of (n) bits of the converters 35 and 36 are given to an EXOR gate 41 to determine exclusive OR, and an output c0 thereof is added to an n-bit signal outputted form the converter 35 and is given to a register 42.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a rotation angle detection device, and is used, for example, in a servo system for various electrical equipment, for detecting or giving feedback to the rotation angle or phase angle of a rotating part, such as a servo motor. The present invention relates to a rotation angle detection device used by being attached to a.

[Prior art] A variable rotor in which a rotor is provided with a plurality of salient poles, and changes in the reluctance of a magnetic circuit between the plurality of salient poles and a detection section of a stator arranged concentrically with respect to the rotor are extracted as an electrical signal. As a reluctance type rotation angle detection device, there is an inductor type resolver.

The inventors of the present application filed a patent application in 1983 for a rotation angle detection device with improved detection accuracy for such an inductor type resolver.
It was proposed in No.-101126.

FIG. 4 is a diagram showing the detection section of the above-mentioned proposed rotation angle detection device, and FIG. 5 is a diagram showing the overall configuration as well. First, referring to FIG. 4, salient poles 11 are provided on the outer periphery of the rotor 1, with the circumference equally divided. This rotor 1 is constructed by forming several electromagnetic steel plates and stacking them together. A stator 2 is concentrically connected to the rotor 1 via an air gap 4.
is provided. Stator 2 has salient poles 11 of rotor 1.
Large salient poles AI, Bl, CI, Di, A2°B2 facing
．． ...C4 and Di are provided. Furthermore, each of these large salient poles A1...Di is provided with a plurality of small salient poles 21 at the same pitch as the salient poles 11 of the rotor 1.

The large salient poles of adjacent stators 2 are arranged so that the small salient poles 21 provided thereon are shifted in phase by 1/4 pitch of the pitch of the salient poles 11 of the rotor 1. Therefore, when the stator 2 has 16 large salient poles, the symbols are changed clockwise to Al and Bl.

CI, Di, A2. If B2...C4, Di, large salient pole Al, A2. A3. All A4s are in phase. Similarly, large salient poles 81 to B4. C1 to C4 degrees and D1 to D4 are also in phase. Note that the stator 2 is also manufactured in the same way as the rotor 1 by stacking several molded electromagnetic steel plates.

Excitation coils are wound around the large salient poles AI, Bl...Di of the stator 2 like al, bl...di, respectively, and the coils a1 and C3, bl and b3... cl and C3, di and d3. C2 and C4, b2 and b4. C2 and C4, d2 and d
i, one pole of the same phase is connected in series, and each terminal is connected to terminals 1al and 1a2 . lbl
and 1'b2. lcl and 1c2. ldl and 1d2, 2ai
and 2a2°2bl and 2b2, 2cl and 2c2, 2dl and 2d2. These terminals are connected as shown in FIG. That is, the coils al, a3 and the coils cl, c3 are connected in series with the same polarity, and are excited by the high frequency AC power source e. Coils C2, C4 and C2, C
4 is also connected in series with the same polarity as the coils al, a3°cl, and C3, and is excited by the high frequency AC power source e. On the other hand, coils bl, b2. b3. b4. di, d2.
d3, di are the coils a1 to a4. It is connected with the opposite polarity to cl to C4. Therefore, the polarities of the large salient poles AI, BL, CI, DI, etc. of the stator 1 at the moment of formation are alternately different, such as N pole, S pole, N pole, S pole, etc. Become.

Terminals 1a2 and lcl, 2c2 and 2al, lbl and 1d2
, 2dl and 2b2 are connected to respective movable terminals of variable resistors VRI to VR4 provided in the adjustment section 32, and R/D (resolver/digital) via amplifiers 33 and 34. ) is applied to converter 35. The R/D converter 35 further divides the sin θ output and the cos θ output adjusted by the A adjustment section 32, and outputs a signal divided into 210 to 216 signals per pitch of salient poles of the rotor 1.

The variable resistors vR1 to VR4 of the adjustment section 32 are used to adjust the height of the peak value of the output voltage by adjusting the unbalance of the magnetic circuit of the detection section 31 and variations in the DC resistance and reluctance of the coil. belongs to.

[Problems to be Solved by the Invention] In the rotation angle detection device proposed above, the maximum rotation speed is limited by the frequency characteristics of the R/D converter 35, so it is difficult to achieve high resolution. . For example, D.D.
The frequency characteristic of the R/D converter manufactured by Company C is 800Hz (1
0 bit output). Therefore, in the above-mentioned proposed multi-pole resolver, the number of salient poles of the rotor 1 is 1.
In the case of 50, the maximum rotation speed that can be followed when one revolution is divided into 150X2"-153600 is 5.33rpS
(320 rpm), and with an even higher resolution 12-bit R/D converter, it is 1/4 of that, 1°33 rps.
(80 rpm).

However, for example, a direct drive motor that directly drives the arm of a robot may be required to divide into one million or more. At this time, the maximum rotation speed is 5 Or
pm or less, which often becomes a bottleneck in terms of speed.

Therefore, the main object of the present invention is to provide a rotation angle detection device that can achieve high resolution and high speed.

[Means for Solving the Problem] The invention according to the first claim provides a rotor having a circumference divided into n equal parts (n is an integer of 3 or more) and provided with a salient pole at each equally divided point, and a rotor that is concentric with the rotor. 4m (m is an integer of 1 or more) poles are provided on the stator arranged in The phase of the small salient pole is 1 of the salient pole pitch of the rotor.
It is an inductor type resolver arranged with a /4 pitch shift, and has an output error correction function, and has two output signals with a phase difference of 90 degrees in electrical angle obtained from the coil wound around the poles of the stator. first and second resolver/digital conversion means for converting into a digital signal; and a correction amount setting means for setting a correction amount to the resolver/digital conversion means.

The invention according to claim 2 includes a rotor and a stator provided with 3 m (m is an integer of 1 or more) poles, and the convex poles of the stator are provided with small salient poles facing the salient poles of the rotor, This is a synchronizing device in which the phase of the small salient poles of the stator is shifted by 1/3 of the rotor salient pole pitch in the order of the stator pole arrangement, and has an output error correction function, and is wound around the stator poles. first and second synchro/digital conversion means for converting two output signals having a phase difference of 120 degrees in electrical angle obtained from the coils into digital signals; and output from each of the conversion means. The apparatus includes correction amount setting means for setting correction amounts in the first and second synchro/digital conversion means so that the pulse signals have a phase difference of approximately 0.5 pulses.

Third Claim: 9 This invention provides exclusive control of the least significant bits of the respective n-bit output signals of the resolver/digital conversion means of tt41 and tt12 or the first and second synchro/digital conversion means. By adding the signal obtained by taking the logical sum to the lowest order of one of the two n-bit output signals,
It is configured to include logic circuit means for outputting an n+1-bit output signal.

[Operation] In the invention according to the first and second claims, the pulse signals output from each of the first and second resolver/digital conversion means or the first and second synchro/digital conversion means are approximately The correction amount is set to have a phase difference of 0.5 pulses. In the invention according to claim 3, among the n-bit output signals having a phase difference of 0.5 pulses, the signal obtained by taking the exclusive OR of the respective least significant bits is By adding it to the lowest order of one of the two output signals, the resolution can be increased to 2 as an n+1 bit output signal.
This is equivalent to doubling the frequency characteristics of a resolver/digital converter or a synchro/digital converter.

[Embodiment of the Invention] FIG. 1 is a concrete block diagram of a detection section and a control section of an embodiment of the invention.

In FIG. 1, the detection section 31 and the adjustment section 32 are connected to the fifth
This is the same as the conventional example shown in the figure, and the outputs of the amplifiers 33 and 34 are connected to two R/D converters 35 provided in parallel.
．． given to 36. A corrector 37.38 and a D/A converter 39.4 correspond to each of the R/D converters 35.36.
0 is set. The correctors 37 and 38 are R/D converters 35
．． It is for correcting errors in the signals outputted from each of the 36, and is constructed of, for example, a ROM. The D/A converters 39 and 40 convert the output signals of the correctors 37 and 38 into analog signals. Among the n-bit output signals of the R/D converters 35 and 36, the least significant bit output signal a. ,b.

is applied to the EXOR gate 41 to obtain an exclusive OR, and its output C8 is added to the n-bit signal output from the R/D converter 35 and applied to the register 42.

FIG. 2 is a diagram for explaining how to determine the amount of correction set by the corrector included in one embodiment of the present invention,
FIG. 3 is a timing diagram for explaining the operation of one embodiment of the present invention.

Next, with reference to FIGS. 1 to 4, a specific operation of an embodiment of the present invention will be described.

First, R/D converters 35 and 36 have sinθ output and COS
The operation until the θ output is given will be explained. The large salient pole A1 provided on the stator 2 of the detection unit 31 and the plurality of small salient poles 21 provided on the large salient pole A1 are connected to the salient pole 11 of the rotor 1, which is coaxially arranged with the stator 2 and rotatably arranged.
are opposed to each other via the air gap 4, and furthermore, if their phases are the same, then the large salient pole A2. A3.
The small salient poles of A4 are also in phase with the salient poles of rotor 1.

On the other hand, the large salient poles CI, C2, C3°C4 of the stator 2 are the large salient poles AI, A2. A3. The phase difference with A4 is 180° (
(electrical angle), the large salient poles C1, C2,
The plurality of small salient poles provided on C3 and C4 do not overlap the salient poles of the rotor 1 at all. In such a phase relationship, when viewed from the magnetic circuit composed of the rotor 1 and the stator 2, large salient poles Al, A2, A3 . A
The reluctance of part 4 is smaller than that of parts CI, C2, C3, and C4.

Furthermore, large salient poles A1 to A4. Coils a1 to a4. are wound around C1 to C4, respectively. Looking at the inductances of cl to c4, the inductances of the coils a1 to a4 are larger than the inductances of C1 to C4.

Here, if the rotor 1 is shifted in either direction by 1/4 of one pitch of the salient poles 11 of the rotor 1, the phase relationship between the small salient poles 2 of the stator 2 and the salient poles 11 of the rotor 1 will be Large salient poles A1 to A4 of stator 2. In all of C1 to C4, the salient poles are half aligned. Therefore, both reluctance and inductance are large salient pole A1
~A4° All of C1 to C4 are equal.

Furthermore, if the rotor 1 is further shifted by 1/4 pitch in the same direction as before, the small salient poles of the large salient poles A1 to A4 and the salient poles of the rotor 1 will not overlap at all, and the large salient poles C1 to A4 will not overlap at all. C
The small salient pole of rotor 4 and the salient pole of rotor 1 are in phase and completely overlapped. Therefore, the inductance of the coils 01 to c4 becomes larger than the inductance of the coils a1 to a4.

Further, the small salient poles of the remaining large salient poles IMBI to B4 of the stator 2 are shifted from the small salient poles of the large salient poles A1 to A4 by 1/4 pitch of the salient pole pitch of the rotor 1, and the large salient poles D1 to Since the small salient pole of D4 is similarly shifted by 1/4 pitch from the small salient poles of large salient poles C1 to C4, the coils b1 to b4. and coil d
The changes in inductance of coils 1 to d4 are respectively
The change in inductance of 1 to a4 and C1 to c4 is 1/4 pitch of the pitch of salient poles 11 of rotor 1 (90 in electrical angle).
This means that the phase is shifted by an amount (°).

Since the coils a1 and C3 and C1 and C3 are connected in series and connected to the high frequency AC power supply e, when the rotor 1 rotates, the coil a1 and C3 are connected in series.

The inductances of C3, cl, and C3 change with a correlation. In other words, when the inductances of the coils al and a3 are maximum (when the small salient poles 21 of the stator 2 and the salient poles 11 of the rotor 1 overlap), the inductances of the coils cl and c3 are the minimum (when the small salient poles 21 of the stator 2 and the salient poles 11 overlap). When the salient poles 11 of the rotor 1 do not overlap, the coil cl
, c3 exhibits a minimum value.

When rotor 1 rotates by 1/4 pitch, coil al
, a3 and cl, C3 are equal in inductance, so the potential difference between the ends of the coils cl, c3 shows an intermediate value (a voltage that is 1/2 of the voltage of the high-frequency AC power supply e). Further, when the rotor 1 rotates by 1/4 pitch, the inductance of the coils al and a3 becomes the minimum, and the inductance of the coils cl and c3 becomes the maximum, so the potential difference between the ends of the coils cl and c3 becomes the maximum. Further, when the rotor 1 rotates by 1/4 pitch, the inductances of the coils al, a3 and the coils cl, c3 are equal, and the potential difference becomes an intermediate value. Furthermore, the rotor 1 rotates by 1/4 pitch and returns to the same state as before. This cycle corresponds to 360'' electrical angle.

Therefore, the coil cl in this cycle.

The potential difference between both ends of C3 has a sinusoidal curve in which the peak value of the AC having the same frequency as that of the high-frequency AC power source e undulates.

Next, since the coils C2 and C4, C2 and C4 are connected in the opposite order to the aforementioned coils a1 and C3, and cl and C3, the potential difference between the ends of the coils a2 and C4 is the same as that of the coil C1. An alternating current voltage that is inverted from the potential across C3 is output. When such two high-frequency AC voltages are extracted, the difference between the two is output as a difference between the two high-frequency AC voltages.

In addition, coils b1 and b3. di and C3 and C2 and C4
, b2 and b4, the above-mentioned left half coils a1 and C3, cl and C3, C2 and C4, C2
and C4, but as mentioned above, the phase is shifted by 1/4 pitch of the pitch of the salient poles 11 of the rotor 1, so the COS curve shape is shifted by 1/4 pitch.

The sin θ output obtained as described above and c.

It is amplified to a predetermined voltage by Sθ ratio output amplifiers 33 and 34, and is applied to R/D converters 35 and 36. The R/D converters 35 and 36 are configured to convert the sin obtained by the detector 31 into
It has a correction function for correcting output errors caused by waveform distortion of the θ output and cos θ output. This correction function involves checking for errors caused by distortion of the input signal to the R/D converter 35.36, and inputting a signal to the R/D converter 35.36 that cancels out the error. Accordingly, correct position signals are output from the R/D converters 35 and 36.

To perform this correction, the correctors 37 and 38 use the n-bit position signal output from the R/D converter 35 and 36 as an address, and store a predetermined binary digital quantity at that address. R/D converter 35.3
A digital signal is output based on the position signal output from 6. This digital signal is converted into an analog signal by D/A converters 39 and 40 and then sent to R/D converters 35 and 3.
6 and correction is made to the correct position.

In one embodiment of the present invention, each of the R/D converters 35 and 36 outputs n-bit position signals, a phase difference is provided to the n-bit outputs, and these are combined by an EXOR gate 41 to produce one bit of position signals. The number of bits is increased so that the output is n+1 bits.

This operation will be explained in more detail with reference to FIGS. 2 and 3. FIG. 2 shows the corrector 37 in FIG.
A constant digital signal is output from 38, and R/D
Error values measured without correction are shown when a constant voltage signal is input to the converters 35 and 36. The horizontal axis represents the absolute angle expressed in pulses, and the vertical axis represents the error determined as the deviation from the absolute angle, indicating the curves SAI and SA2. Assuming that the errors of the two channels show the same value, in order to correct this error, a correction amount as shown in FIG. 2(b) is set. At this time, if the error is 0, the correction amount is 0, and if the error is 0, the correction amount is -, which is an amount commensurate with the error.

In this way, the curve HA shown in FIG. 2(b) is set for each pulse, and this value is the 1st P1! It is stored in the corrector 37 of J. Further, the curve Ha shown in FIG. 2(b) is a correction amount set such that there is an error of +0.5 pulses from the above-mentioned correction amount. This correction amount is stored in the corrector 38 shown in FIG. If the correctors 37 and 38 set in this way are used to measure the errors using the absolute angle as a reference, the output of the R/D converter 35 will be as shown in FIG. 2(c). A channel is Sal
As shown in , the error is +0. In addition, R/D converter 3
The B channel, which is the output of No. 6, always shows an error of 10.5 pulses from the absolute angle like Sa2.

FIG. 3 shows n bits a of the A channel. +al...a
n-+ and n bits of the B channel. ,b,...
The output waveform of bn-1 is shown. These A, B
As can be seen by comparing the channels, each bit shows a waveform that is advanced by 0.5 pulses in the B channel. Here, ao is the least significant bit of both channels. When the exclusive OR of is calculated, a waveform like C8 in FIG. 3 is synthesized. This synthesis is performed by the EXOR gate 4 shown in Figure 1.
This is done by 1. The composite waveform C8 obtained here
By adding the n bits to the lowest order of the n-bit output of the a channel, a continuous output of n+1 bits is obtained.

This n+1 bit output is given to register 42.

Note that the output error of the B channel is ideally +0.5 pulses, but it is sufficient that the lowest bit outputs of the A channel and B channel are output alternately, and 0.5 pulses can be set to different values. There is no problem in practical terms.

As described above, according to one embodiment of the present invention, the detection unit 31
Although the adjustment section 32 and amplifiers 33 and 34 are the same as in the prior art, the number of output bits can be increased by one by adding an R/D converter 36 or the like for one channel. In other words, the resolution can be doubled, and the frequency characteristics are R/D
Since it is determined by the characteristics of the converter 35 and 36, it will not drop. In other words, this is equivalent to doubling the frequency characteristics of the R/D converters 35 and 36. When the number of salient poles of the rotor of the detection unit 31 is 150, the resolution and the maximum rotation speed are shown in the table as an example. The frequency characteristics of the R/D converters 35 and 36 are assumed to be 800 Hz at 10 bits.

(Left below) In the above-mentioned embodiment, the phase of the poles provided on the stator 2 is shifted by 1/4 pitch of the salient pole pitch of the rotor 1 from a coil wound around four salient poles of the rotor 1. The present invention was applied to a resolver that produces two outputs having a phase difference of 2, ie, sin θ and cos θ output waveforms. However, in contrast, the salient poles of the stator are arranged so as to be shifted by 1/3 pitch of the salient pole pitch of the rotor, and three coils having a phase difference of 120° are generated from the coils wound around the salient poles of the three stators. Even in a so-called synchro device that obtains an output, by using two synchro/digital converters in parallel, the frequency characteristics can be doubled in the same way as in the above embodiment. Such a synchronizer is described, for example, in Japanese Patent Laid-Open No. 125153/1983.

[Effects of the Invention] As described above, according to the present invention, the resolution can be doubled by adding one channel of R/D converters and the like.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of one embodiment of the present invention. Second
The figure is a diagram for explaining how to determine the amount of correction set by the corrector in one embodiment of the present invention. FIG. 3 is a timing diagram for explaining the operation of one embodiment of the present invention. FIG. 4 is a diagram showing a detecting section of a resolver which is a prior art of the present invention. FIG. 5 is a block diagram showing the overall configuration of the resolver. In the figure, 31 is a detection section, 32 is an adjustment section, 33.34
is an amplifier, 35.36 is an R/D converter, 37.38 is a corrector, 39.40 is a D/A converter, and 41 is an EXOR gate. Figure 2 (a) (c) Figure 3

Claims (3)

[Claims] (1) A rotor whose circumference is divided into n equal parts (n is an integer of 3 or more) with salient poles provided at each equally divided point, and a stator with 4 m (m is an integer of 1 or more) salient poles arranged concentrically with the rotor. , and each pole of the stator is provided with a small salient pole opposing the salient pole of the rotor, and the phase of the small salient pole of the stator is adjusted in the order of the arrangement of the stator poles to the salient pole of the rotor. 1/4 of the pitch
The inductor type resolver, which is arranged at different pitches, has an output error correction function, and converts two output signals having a phase difference of 90 degrees in electrical angle obtained from the coils wound around the poles of the stator into digital signals. The first to convert to
and second resolver/digital conversion means, and the first and second resolver/digital conversion means such that the pulse signals output from each of the first and second resolver/digital conversion means have a phase difference of approximately 0.5 pulse. A rotation angle detection device comprising a correction amount setting means for setting a correction amount in the second resolver/digital conversion means. (2) A rotor whose circumference is divided into n equal parts (n is an integer of 5 or more) with salient poles provided at each equally divided point, and a stator with salient poles of 3m (m is an integer of 1 or more) arranged concentrically with the rotor. , and each pole of the stator is provided with a small salient pole opposing the salient pole of the rotor, and the phase of the small salient pole of the stator is adjusted in the order of the arrangement of the stator poles to the salient pole of the rotor. 1/3 of the pitch
The synchronizers arranged at different positions have an output error correction function and convert two output signals having a phase difference of 120 degrees in electrical angle obtained from the coils wound around the poles of the stator into digital signals. such that the pulse signals output from each of the first and second synchro/digital conversion means have a phase difference of approximately 0.5 pulse. , a rotation angle detection device comprising correction amount setting means for setting correction amounts in the first and second synchro/digital conversion means. (3) Exclusive OR of the respective least significant bits of the n-bit output signals of the first and second resolver/digital conversion means or the first and second synchro/digital conversion means. Claim 1, further comprising logic circuit means for adding the obtained signal to the lowest order of one of the two n-bit output signals to obtain an n+1-bit output signal. Or the rotation angle detection device according to item 2.