JP2006308074A - Magnetic bearing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a magnetic bearing device for eliminating demodulation processing while requiring less computing throughput regarding bearing control than conventional one. <P>SOLUTION: A carrier signal of a frequency fc is applied to a sensor 73 via a filter 32. The carrier signal is demodulated by the sensor 73 and a difference between amplitude modulated wave F<SB>AM</SB>(t) and a sensor reference signal Fstd(t) is computed by a differential amplifier 203. A difference signal Fsub(t) output from the differential amplifier 203 is A/D converted into a digitized sensor signal by an A/D converter 34. A sampling frequency fs at this time is set to be equal to fc=n×fs or fc=fs/2 in relation to the frequency fc of the carrier signal, where n=1, 2, .... So, the digitized sensor signal does not include a carrier and conventionally necessary demodulation processing can be eliminated. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a magnetic bearing device used for a turbo molecular pump, a machine tool, or the like.

  In a turbo molecular pump or a machine tool that requires high-speed rotation, a magnetic bearing device is often used. In the magnetic bearing device, the displacement of the rotating body is detected by a displacement sensor, and the detection result is fed back to the excitation current control of the electromagnet. An inductance type sensor is often used as the displacement sensor. A carrier wave is applied to the displacement sensor, and the carrier wave is amplitude-modulated by an impedance change caused by a gap change with the rotating body. The amplitude-modulated signal is demodulated after A / D conversion, and the excitation current of the electromagnet is controlled using the displacement signal obtained by the demodulation (see, for example, Patent Document 1).

JP 2004-144291 A

  By the way, the frequency of the modulated wave signal to be A / D converted is fc−fr to fc + fr, where fc is the carrier frequency and fr is the frequency band of the displacement signal of the rotating body. The largest fr band within this range is 2fr. In general, when an analog signal is converted into a digital signal by an A / D converter at a sampling frequency fs, it is necessary to satisfy 4fr <fs from the sampling theorem in order to prevent aliasing. For this reason, fr is limited to fs / 4, and depending on the processing capability of the DSP (Digital Signal Processor) being used, fr may not be very large, and the high-frequency characteristics are deteriorated.

  In order to solve this, measures such as increasing the sampling frequency fs or increasing the carrier frequency fc can be considered. However, since the processing for the DSP increases, a DSP with a high frequency operation is required, which increases the cost. It becomes a factor. Further, in the case of performing conventional digital demodulation, there is a case where the signal is attenuated during the demodulation process, the S / N is lowered, and the rotating body as the supported body is vibrated.

The invention of claim 1 is applied to a magnetic bearing device that supports a supported body in a non-contact manner by an electromagnet, and generates a carrier wave signal, and modulates and modulates the carrier wave signal according to the support position of the supported body. Detection means that outputs a wave signal, and detection at a sampling frequency fs satisfying fc = n · fs (where n is a natural number) or fc = fs / 2, where fc is the frequency of the carrier wave signal and fs is the sampling frequency A / D conversion means for converting the modulated wave signal output from the means into a digital signal, and the excitation current of the electromagnet is controlled based on the digital signal output from the A / D conversion means, and the support position of the supported body And control means for controlling.
According to a second aspect of the present invention, in the magnetic bearing device according to the first aspect, the carrier wave generating means includes a sine wave discrete value generation unit that generates a sine wave discrete value by digital arithmetic processing, and a sine wave discrete value as a D / A. A D / A converter that converts and generates a carrier wave signal, and includes a phase shift calculator that shifts the phase of a sine wave discrete value by digital calculation processing, based on the phase shift calculation result by the phase shift calculator, The timing at which the modulated wave signal is converted into a digital signal by the A / D conversion means is made to substantially coincide with the envelope of the modulated wave signal.
According to a third aspect of the present invention, in the magnetic bearing device according to the first aspect, the magnetic bearing device is a multi-axis control type magnetic bearing device having a plurality of control shafts, and the detecting means is a carrier wave signal for each of the plurality of control shafts. The A / D conversion means performs A / D conversion in synchronization with each carrier wave signal of each modulated wave signal output from the detection means, and (b) a plurality of modulation signals are output. As for two modulation signals in the modulation wave signal, one is sampled in synchronization with the timing at which the carrier component changing at the carrier frequency in the modulation wave signal is maximized, and the other is sampled in the modulation wave signal. Sampling is performed almost in synchronization with the timing at which the carrier component is minimized.
According to a fourth aspect of the present invention, in the magnetic bearing device according to the first aspect, the magnetic bearing device is a multi-axis control type magnetic bearing device having a plurality of control shafts, and the detecting means is a carrier signal for each of the plurality of control shafts. The A / D conversion means performs A / D conversion in synchronization with each carrier wave signal output from the detection means, and performs carrier wave frequency in the modulated wave signal. In this case, sampling is performed within a phase range of ¼ of the carrier cycle centered on the phase where the carrier component that changes at the maximum or minimum is centered.
According to a fifth aspect of the present invention, in the magnetic bearing device according to the third or fourth aspect, the carrier wave generating means generates a plurality of carrier signals having the same frequency and different phases, and the detecting means is provided for each of the plurality of control axes. One of a plurality of carrier signals is modulated and a modulated signal is output.
According to a sixth aspect of the present invention, in the magnetic bearing device according to the fifth aspect, the carrier wave generating means is generated by a sine wave discrete value generation unit that generates a sine wave discrete value by digital processing and a sine wave discrete value generation unit. A D / A conversion means for generating a carrier wave signal by D / A converting the sine wave discrete value, and a phase shift means having at least one phase shift circuit for shifting the phase of the carrier wave signal. is there.
According to a seventh aspect of the present invention, in the magnetic bearing device according to the fifth aspect, the carrier wave generating means is generated by a sine wave discrete value generation unit that generates a sine wave discrete value by digital processing and a sine wave discrete value generation unit. And a plurality of D / A conversion means for generating a carrier wave signal by D / A converting the discrete sine wave values.
According to a seventh aspect of the present invention, in the magnetic bearing device according to the third or fourth aspect, a plurality of A / D converting means are provided, and a plurality of modulated wave signals output from the detecting means and having the same carrier wave component are output. These are input to different A / D conversion means and A / D converted at the same timing.

  According to the present invention, when the frequency of the carrier signal is fc and the sampling frequency is fs, A / D conversion is performed at a sampling frequency fs satisfying fc = n · fs (where n is a natural number) or fc = fs / 2. By doing so, it is possible to omit the demodulation processing that has been conventionally required, to improve the high frequency characteristics, and to reduce the amount of calculation processing related to bearing control as compared with the conventional one. Further, when the A / D conversion is performed in synchronization with the carrier wave signal, the modulated wave signal is sampled almost synchronously with the timing at which the carrier wave component is maximized or minimized, as in the invention of claim 3. Sampling in a phase range of ¼ of the carrier cycle centered on the phase where the carrier component is maximum or minimum as in the invention can suppress a decrease in the S / N ratio.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings. FIG. 1 is a sectional view showing a schematic configuration of a pump body 1 of a magnetic bearing turbomolecular pump. Hereinafter, a case where the present invention is applied to a magnetic bearing device of a magnetic bearing type turbo molecular pump will be described as an example. The magnetic bearing type turbo molecular pump shown in FIG. 1 includes a 5-axis control type magnetic bearing, and a rotor 4 formed with a plurality of stages of rotor blades 21 includes electromagnets 51 and 52 constituting a radial magnetic bearing, and an axial. Non-contact support is provided by the electromagnet 53 constituting the magnetic bearing.

  Inside the casing 20, a plurality of rotor blades 23 are provided in the axial direction (the vertical direction in the figure). The magnetic bearings are provided with radial sensors 71 and 72 and an axial sensor 73 for detecting the flying position of the rotor 4 corresponding to the radial electromagnets 51 and 52 and the axial electromagnet 53. The receptacle 25 is connected to a cable connecting the pump main body 1 and a controller (not shown), and the pump main body 1 is driven and controlled by the controller. The rotor 4 is rotationally driven by the motor 6 while being supported in a non-contact manner by the electromagnets 51, 52, 53, thereby generating a pump action.

  27 and 28 are emergency mechanical bearings. When the rotor 4 is not magnetically levitated, the rotor 4 is supported by these bearings 27 and 28. The bearing 27 constrains the movement of the rotor 4 in the radial direction (x-axis and y-axis) in an emergency, and the bearing 28 includes two radial directions (x-axis and y-axis) and one axial direction (z-axis). ).

  FIG. 2 is a control block diagram of a 5-axis control type magnetic bearing. The radial electromagnet 51 includes an X1-axis electromagnet 51x and a Y1-axis electromagnet 51y, and the radial electromagnet 52 includes an X2-axis electromagnet 52x and a Y2-axis electromagnet 52y. Each of the electromagnets 51x, 51y, 52x, and 52y is composed of one electromagnet facing each other with the rotor 4 interposed therebetween. On the other hand, the sensor 71 includes an X1 axis sensor 71x and a Y1 axis sensor 71y corresponding to the X1 axis electromagnet 51x and the Y1 axis electromagnet 51y. Similarly, the sensor 72 includes an X2 axis sensor 72x and a Y2 axis sensor 72y corresponding to the X2 axis electromagnet 52x and the Y2 axis electromagnet 52y.

  The sensors 71x, 71y, 72x, 72y, and 73 are inductance type sensors, and the gap displacement is converted into an electric signal by using a change in sensor unit impedance caused by a change in the gap displacement. The carrier wave signal having the frequency fc generated by the digital control circuit 30 is converted into an analog signal by the D / A converter 31 and applied to each of the sensors 71x, 71y, 72x, 72y and 73 via the filter 32.

  The carrier wave signal applied to each of the sensors 71x, 71y, 72x, 72y and 73 is amplitude-modulated in accordance with the sensor section impedance change caused by the gap displacement. This amplitude-modulated modulated wave signal (hereinafter simply referred to as amplitude modulated wave) is input to the A / D converter 34 via the sensor circuits 33a to 33e. Analog signals from the sensor circuits 33 a to 33 e are sequentially converted into digital values by the A / D converter 34 and input to the digital control circuit 30.

  The digital control circuit 30 calculates an excitation current to be passed to each of the electromagnets 51x, 51y, 52x, 52y, 53 based on the magnetic levitation control constant stored in advance and the position information converted into a digital value. Output a control signal. The excitation current control signal is converted into an analog value by the D / A converter 35 and then input to the excitation amplifier 36. In FIG. 2, only one excitation amplifier 36 is shown, but in actuality, there are provided as many control axes as there are excitation currents from each excitation amplifier to each electromagnet 51x, 51y, 52x, 52y, 53. Is supplied.

FIG. 3 shows in detail a portion related to the radial sensor 73 (z-axis direction) in the control block diagram shown in FIG. The sine wave discrete value generated by the sine wave discrete value generation unit 313 of the digital control circuit 30 is converted into an analog signal by the D / A converter 31, and the analog signal is output to the filter 32. Since the output carrier wave signal includes harmonics and has a stepped shape, a smooth carrier wave signal can be obtained by filtering with the filter 32 including a low-pass filter and a band-pass filter. The carrier wave signal is applied to a sensor 73 connected in series through a resistor R. The carrier signal Fcarrier (t) output from the filter 32 is expressed by the following equation (1) if the carrier frequency is fc.
Fcarrier (t) = Asin (2πfct) (1)

The carrier wave signal applied to the displacement sensor 73 is amplitude-modulated by an impedance change that changes in accordance with the position of the rotor 4 to become an amplitude-modulated wave F _AM (t). Here, assuming that the position information signal is Fsig (t), the amplitude-modulated wave F _AM (t) is expressed by the following equation (2). Note that φ is the phase difference from the carrier signal.
F _AM (t) = (A + Fsig (t)) sin (2πfct + φ) (2)

FIG. 4 shows an example of a signal waveform. (A) shows the position information signal Fsig (t), and (b) shows the carrier signal Fcarrier (t). When the carrier signal Fcarrier (t) in FIG. 4B is modulated with the position information signal Fsig (t) in FIG. 4A, an amplitude-modulated wave F _AM (t) as shown in FIG. 4C is obtained. . This amplitude-modulated wave F _AM (t) is input from the displacement sensor 73 to the differential amplifier 203.

A sensor reference signal Fstd (t) represented by the following equation (3) is input to the differential amplifier 203 together with the amplitude modulation wave F _AM (t), and the differential signal Fsub (t) is input to the differential amplifier 203. Is output from. For the sensor reference signal Fstd (t), the gain of the carrier signal Fcarrier (t) is adjusted by the gain adjusting unit 202, and the phase of the sensor reference signal Fstd (t) is adjusted by the phase shift circuit 204 so as to be in phase with the amplitude modulation wave F _AM (t). Is formed.

The sensor reference signal Fstd (t) has a waveform as shown in FIG. 4D, and the differential signal Fsub (t) shown in the following equation (4) has a waveform as shown in FIG. The difference signal Fsub (t) output from the differential amplifier 203 is subjected to bandpass processing with the carrier frequency fc as the center frequency in the filter 205.
Fstd (t) = Csin (2πfct + φ) (3)
Fsub (t) = F _AM (t) −Fstd (t)
= (A + Fsig (t) -C) sin (2πfct + φ) (4)

  The difference signal Fsub (t) input from the filter 205 to the A / D converter 34 is converted into a digital value by the A / D converter 34. The frequency of the discretized signal sampled by this digital sampling is represented by a signal having a frequency (fa−fb) when, for example, a sine wave of the frequency fa is sampled at the sampling frequency fb.

  When digital conversion is performed by the A / D converter 34, sampling is performed based on the sine wave discrete value generated by the sine wave discrete value generation unit 313. However, when the carrier signal is modulated by the displacement sensor 73, the phase is changed. shift. For this reason, the phase shift computing unit 312 shifts the phase of the sine wave discrete value according to the shift, and inputs it to the A / D converter 34. The A / D converter 34 makes the timing for converting the modulated wave signal into a digital signal substantially coincide with the envelope of the modulated wave signal. That is, it is synchronized with the maximum position of the carrier wave component.

Here, a case where the sampling frequency fs in the A / D converter 34 is made equal to the frequency fc of the carrier wave signal will be described. At this time, the discretized sensor signal e (t) obtained by sampling the difference signal Fsub (t) at the frequency fc is expressed by the following equation (5). Note that P = A−C, Q = sinφ, and P and Q are constants.
e (t) = (A + Fsig (t) −C) sin {2π (fc−fc) t + φ}
= (A + Fsig (t) -C) sinφ
= QP + QFsig (t) (5)

  As can be seen from the equation (5), the discretized sensor signal e (t) does not include a carrier wave and has a feature that there is no need to perform demodulation calculation processing. FIG. 4 (f) shows the discrete sensor signal e (t), and the discrete sensor signal e (t) is offset and gain adjusted by the gain / offset adjustment unit 310 of the digital control arithmetic circuit 30. The original position information signal Fsig (t) can be extracted. FIG. 4G shows the discretized sensor signal e (t) after gain / offset adjustment, and the broken line shows the position information signal Fsig (t) superimposed. In the control calculation unit 311, the excitation current control amount is calculated based on the extracted position information signal Fsig (t).

In the above example, the case where fc = fs has been described. However, even when the relationship between the sampling frequency fs and the frequency fc of the carrier signal is set to fc = n · fs, the carrier component is included in the discretized signal. Similarly, demodulation operation processing and filter processing can be omitted. The discrete value signal obtained by sampling the difference signal Fsub (t) at the sampling frequency fs is expressed as the following equation if the value at t = mTs (where m = 0, 1, 2,...) Is sampled. Is done. However, Ts = 1 / fs.
(A + Fsig (mTs) -C) sin (2πfc · mTs + φ)

Considering the case of fc = n · fs, the result is the same as the case of fc = fs, as shown in the following equation.
(A + Fsig (mTs) -C) sin (2πfc · mTs + φ)
= (A + Fsig (mTs) -C) sin (2πn · fs · m / fs + φ)
= (A + Fsig (mTs) -C) sin (2πn · m + φ)
= (A + Fsig (mTs) -C) sinφ
= QP + QFsig (mTs)

  FIG. 5 is a diagram for explaining the sampling timing when sampling is performed in synchronization with the carrier wave signal. In FIG. 5, (a) is a displacement signal corresponding to the above-described position information signal Fsig (t), (b) is a carrier signal, (c) is a sensor signal obtained by modulating the carrier signal with the position information signal. Show. The sensor signal has a carrier component that varies at the carrier frequency. Further, (d) to (e) show sampled discrete value signals (indicated by circles and triangles) when fc = fs (that is, when n = 1), and sampling starts. The timing is different.

  When n = 1, sampling is performed every one cycle Tc of the carrier component. In FIG. 5D, the sampling timing is synchronized with the position where the carrier component is maximum, and in FIG. 5E, the carrier component is minimum. It is synchronized with the position. The discrete value signal acquired at the sampling timing shown in (e) can obtain the displacement signal (a) only by inverting the sign of the signal. In (f), the sampling timing is synchronized with the position shifted from the maximum position and the minimum position of the carrier wave component.

  In addition, when n = 2, sampling is performed every two periods of the carrier wave component, so that every other circle and triangle in FIG. 5 are sampled. Further, every two samples are sampled when n = 3, and every three samples when n = 4. The same applies when n is 5 or more.

On the other hand, when the sampling frequency fs is set as fc = fs / 2, the sampled discrete value signal is expressed by the following equation.
(A + Fsig (mTs) -C) sin (2πfc · mTs + φ)
= (A + Fsig (mTs) -C) sin (π · m + φ)
In this case, sampling is performed every half cycle of the carrier component, and if sampling is started so as to synchronize with the maximum position of the carrier component, the first sampling is performed with a circle in FIG. In the second round, sampling is performed at the position of the triangle mark in FIG. That is, sampling is performed in the order of maximum position, minimum position, maximum position, minimum position, and so on. In this case, the envelope of the sensor signal shown in FIG. 5C, that is, the signal shown in FIG. 5A is obtained by vertically inverting the discrete value signal at the minimum position.

  As a comparative example, consider a case where the sampling frequency fs is set to 4/3 times the carrier frequency fc. Here, a case where a simple sine wave (shown by a solid line) as shown in FIG. 6 is sampled will be considered. When the sine wave signal shown in FIG. 6 is A / D converted with fs = (4/3) fc, sampling is performed at the position of the triangle mark S1 in FIG. The sampled discrete value signal has a periodicity as indicated by a broken line, and has a frequency that is ¼ that of a sampled signal (a sine wave signal indicated by a solid line).

In this case, demodulation calculation processing is required after A / D conversion, and the A / D converted data is multiplied by a sine wave having a quarter frequency of the sampled signal. The signal to be sampled at this time is represented by the following equation (6), and a signal obtained by A / D conversion is represented by equation (7). Ts is a sampling period.
Fsample (t) = Ksin (2πfct + ξ) (6)
FADin = Ksin {2π (fs / 4) · nTs + ξ ′}
= Ksin {π · n / 2 + ξ ′} (7)

If the demodulation multiplication sine wave signal Fdecode at this time is expressed by the following equation (8), the demodulated signal Fdetect is expressed by the following equation (9).
Fdecode = Lsin {2π (fs / 4) · nTs + ξ ′}
= Lsin {(π / 2) · n + ξ ′} (8)
Fdetect = FADin x Fdecode
= KLsin ₂ {(π / 2) · n + ξ ′}
= KL {1-cos (πn + ξ ′)} / 2 (9)

  Here, considering the case of L = K = 1 and ξ ′ = 0, it can be seen that the signal of amplitude 1 is attenuated to Fdetect = 1/2 at Fsample (t) = sin (2πfct). In the case of this processing, low-pass filter processing is necessary to extract a DC component (= KL / 2) after demodulation processing. In the case of FIG. 6, since sampling is performed at the position of the triangle mark S1, the acquired signals are 0, -1, 0, 1, 0, -1, 0, 1, 0,. When this is multiplied by a synchronized signal of the same frequency, it becomes 0, 1, 0, 1, 0, 1, 0, 1,..., And when these are averaged, the signal is attenuated to 0.5.

  On the other hand, when sampling is performed at fc = fs as in the present embodiment, the signal is sampled at the position of the square mark S2 in FIG. 6, and the captured signal is used as it is as the position information signal. it can. As can be seen from FIGS. 5D and 5E, the sampling frequency fs is set to fc = nfs (n ≠ 1) and fc = fs / 2 with respect to the carrier frequency fc. The same applies to the case, and if the sampling is performed in synchronization with the maximum position or the minimum position of the carrier wave component, the S / N ratio does not decrease.

  However, when the sampling start position deviates from the maximum position and the minimum position of the carrier wave component as shown in FIG. 5 (f), the gain of the signal decreases, and this is the gain in the case of maximum position synchronous demodulation. When the adjustment is made, the S / N ratio decreases. This decrease in the S / N ratio increases as the phase shift between the sampling timing and the maximum carrier position increases, and no signal component is detected at a position 90 degrees out of phase from the maximum carrier position.

  By the way, when a carrier wave having the same frequency and the same phase is applied to each sensor as in the five-axis control type magnetic bearing device shown in FIG. The carrier component of the signal has the same phase. FIG. 7 shows in detail the sensors 71x to 73 and the sensor circuits 33a to 33e of each axis in FIG. The X1 axis sensor 71x, Y1 axis sensor 71y, X2 axis sensor 72x, and Y2 axis sensor 72y in the radial direction have the same configuration, and a difference is taken between output signals from the pair of sensors.

  On the other hand, as for the sensor 73 in the axial direction, the difference between the output signal from the sensor and the sensor reference signal Fstd (t) is calculated as described above. As described above, since the sensing method is different between the radial direction and the axial direction, the X1 axis, the Y1 axis, the X2 axis, the Y2 axis, and the Z axis often have large phases. Therefore, the A / D converter 34 can sample the Z-axis signal in synchronization with the maximum position of the carrier component, and can also perform sampling on the other axis in synchronization with the maximum position.

  However, since the X1, Y1, X2, and Y2 axes have the same structure as shown in FIG. 7, signals having the same phase are input to the A / D converter 34. Therefore, it is not possible to sample all of the X1, Y1, X2, and Y2 axes in synchronization with the maximum or minimum position of the carrier component. For example, if the X1 axis is sampled in synchronization with the maximum position and the Y1 axis is synchronized with the minimum position, the other X2 and Y2 axes are sampled at positions that deviate from the maximum and minimum positions of the carrier component. , S / N ratio decreases.

A method for solving such a problem will be described below.
<First Method>
In the first method, a plurality of types of carrier signals having different phases are generated and applied to each sensor. FIG. 8 is a block diagram showing an example thereof, and two D / A converters 31A and 31B are provided to generate two types of carrier signals having different phases.

  The sine wave discrete values generated by the sine wave discrete value generation unit 313 (see FIG. 3) of the digital control circuit 30 are input to the D / A converters 31A and 31B. The carrier wave signal from the D / A converter 31A is input to the Z-axis (axial) sensor 73, the X1-axis sensor 71x, and the Y1-axis sensor 71y via the filter 32. On the other hand, the carrier wave signal from the D / A converter 31B is input to the X2 axis sensor 72x and the Y2 axis sensor 72y via the filter 32.

  In this case, the carrier components of the X1-axis and Y1-axis sensor signals input to the A / D converter 34 have the same phase, and the carrier components of the X2-axis and Y2-axis sensor signals have the same phase. Yes. However, the X1 axis, Y1 axis and the X2, Y2, and Z axes are out of phase. The A / D converter 34 performs sampling in the order of the channels 1 to 5, and for example, the X1 axis and the X2 axis can be synchronized with the maximum position of the carrier component, and the Y1 axis and the Y2 axis can be synchronized with the minimum position of the carrier component. . Since the Z axis is out of phase with the X1 and Y1 axes, it can be synchronized to the maximum position independently of them.

  FIG. 8 shows an example in which the carrier wave signals of the D / A converters 31A and 31B are input to the sensors. The D / A converters 31A and 31B and the sensors 71x, 71y, 72x, 72y, The correspondence with 73 is not limited to the case of FIG.

  In the example shown in FIG. 8, a plurality of D / A converters are provided to generate carrier signals having different phases, but the phase may be shifted using a phase shift circuit as shown in FIG. In the block diagram shown in FIG. 9, the carrier wave signal output from the filter 32 is input to the phase shift circuits 37A to 37E, respectively. The carrier signal shifted in phase by the phase shift circuit 37A is input to the X1-axis sensor 71x. Similarly, the carrier signals shifted in phase by the phase shift circuits 37B to 37E are input to the sensors 71y, 72x, 72y, 73, respectively. Is done.

  The phase shift amounts in the phase shift circuits 37A to 37E are set so that the sensor signals input to the channels 1 to 5 of the A / D converter 34 can be sampled in synchronization with the maximum position or the minimum position of the carrier component. Is done. As a result, for example, sensor signals input to channels 1, 3, and 5 are sampled in synchronization with the maximum position of the carrier component, and sensor signals input to channels 1, 3, and 5 are synchronized to the minimum position of the carrier component. And can be sampled.

  In the example shown in FIG. 9, the phase shift circuit is provided on all of the five axes so that the phase can be adjusted most easily. However, it is not always necessary to provide the five axes. For example, a carrier wave signal output from the filter 32 is directly input to the X1 axis and the Y1 axis, and a carrier wave signal whose phase is shifted by 90 degrees using a phase shift circuit is input to the X2 axis and the Y2 axis. With respect to the X1, Y1, X2, and Y2 axes, sampling can be performed in synchronization with the maximum position or the minimum position of the carrier component.

<Second method>
In the second method, by using a plurality of A / D converters, the sensor signal of each axis is sampled at the maximum position and the minimum position of the carrier wave component. FIG. 10 is a block diagram in the case of the second method, and two A / D converters 34A and 34B are provided. The X1-axis and X2-axis sensor signals are input to the A / D converter 34A, and the Y1-axis, Y2-axis, and Z-axis sensor signals are input to the other A / D converter 34B.

  In the case of the second method, the signals input to channel 1 and the signals input to channel 2 of the A / D converters 34A and 34B can be sampled at the same timing. For example, the upper X1 axis sensor 71x and the Y1 axis sensor 71y sample in synchronization with the maximum position of the carrier component, and the lower X2 axis sensor 72x and Y2 axis sensor 72y sample in synchronization with the minimum position of the carrier component. I do. In addition, since the sensor signal of the Z-axis sensor 73 is out of phase with the sensor signals of the Y1-axis sensor 71y and the Y2-axis sensor 72y, the sampling timing of the channel 3 is set to the maximum position or minimum position of the carrier wave component or their position. It is possible to make it a near position.

<Third method>
In the first method described above, the phase of the carrier component of the sensor signal is shifted by providing a plurality of D / A converters or a phase shift circuit. In the second method, by providing a plurality of A / D converters, a plurality of sensor signals having the same phase can be sampled at the same timing. As a result, sampling can be performed in synchronization with the maximum position and the minimum position of the carrier wave component, and the S / N ratio can be prevented from being lowered. Therefore, in the third method, a method for suppressing a decrease in the S / N ratio due to misalignment as much as possible even when the sampling timing is shifted from the maximum position or the minimum position of the carrier wave component will be described.

  FIG. 11 is a diagram for explaining the concept of sampling by the third method. FIG. 11 is a diagram illustrating a sensor signal when the displacement signal in FIG. 5A is only a DC component. The sensor signal is only a carrier wave component having the same frequency as the carrier wave signal, and the maximum position / minimum position of the carrier wave component corresponds to the maximum position / minimum position of the sensor signal. That is, when the sampling timing deviates from the maximum position or the minimum position of the sensor signal, the S / N ratio decreases according to the phase shift.

  In the third method, sampling is performed in a predetermined phase range centered on the maximum position and the minimum position of the carrier wave component, thereby suppressing a decrease in the S / N ratio. Here, the predetermined phase range is set to within ¼ of the carrier cycle Tc (that is, within 90 deg), and sampling is performed at two points by the A / D converter 34 within the range.

  When sampling is performed at the maximum sampling speed of the A / D converter 34, the number of samplings per second is Nsamp. In this case, 1 sampling takes 1 / Nsamp seconds. This time 1 / Nsamp is fc / N times the carrier frequency 1 / fc of the frequency fc and corresponds to (fc / N) · 360 deg in terms of the phase deg. If the carrier frequency Tc is set such that Tc> 4 / Nsamp, sampling can be performed twice within the range Tc / 4.

  In the example shown in FIG. 11, the first point M1 (circle) and the second point M2 (triangle) are sampled at positions symmetrical with respect to the maximum position, and 3 at positions symmetrical with respect to the minimum position. Sampling point M3 (square mark) and fourth point M4 (star mark). In this case, the decrease in signal level is sin (90− (fc / N) · 180). As described above, when sampling is performed twice within the range Tc / 4, even when sampling is performed at a position of ± 45 deg farthest from the maximum position / minimum position, the decrease in signal level is sin (45) = It can be suppressed to a decrease of about 0.707 = 3 dB. In the case of 5 axes, sampling of each axis of X1, Y1, X2, and Y2 is performed by the method as shown in FIG. 11, and the phase shift differs due to the difference in structure for the Z axis whose sensor structure is different from other axes. Is used to perform sampling in synchronism with the maximum or minimum position.

[Modification]
When configuring the magnetic bearing control system, any one of the first to third methods described above can be employed, or a plurality of methods can be employed in combination. For example, as described in the first method, a carrier wave signal whose phase is shifted by 90 degrees is input to the X2 axis and the Y2 axis using a phase shift circuit. If the phase of the Z axis is the same as that of the X1 axis and the Y1 axis, the range of the Tc / 4 period centering on the maximum position of the carrier wave component is used for the X1 axis and the Y1 axis as in the third method. Sampling is performed in synchronism with the minimum position of the carrier component with respect to the Z axis. The same processing is performed when the Z-axis phase is the same as the X2-axis and Y2-axis. On the other hand, if the phase of the Z axis is different from the X1 axis, the Y1 axis, and the X2 axis and the Y2 axis, the Z axis may be sampled near the maximum position of the carrier component. As a result, A / D conversion is possible with the maximum S / N ratio for all five axes.

  In the correspondence between the embodiment described above and the elements of the claims, the rotor 4 is a supported body, the sensors 71 to 73, 71x, 71y, 72x, 72y and the sensor circuits 33a to 33e are detection means, The control circuit 30 and the excitation amplifier 36 constitute control means, and the sine wave discrete value generation unit 313, the D / A converters 31, 31A, 31B, and the phase shift circuits 37A to 37E constitute carrier wave generation means. The above description is merely an example, and when interpreting the invention, there is no limitation or restriction on the correspondence between the items described in the above embodiment and the items described in the claims.

It is sectional drawing which shows schematic structure of the pump main body 1 of the magnetic bearing type turbo molecular pump to which the magnetic bearing apparatus by this invention is applied. It is a control block diagram of a 5-axis control type magnetic bearing. FIG. 3 is a block diagram showing in detail a portion related to a radial sensor 73 (z-axis direction) in the control block diagram shown in FIG. 2. It is a figure which shows an example of a signal waveform. It is a figure explaining sampling timing. It is a figure explaining the improvement of S / N ratio. It is a figure which shows in detail the sensors 71x-73 and sensor circuit 33a-33e of each axis | shaft in FIG. It is a block diagram explaining a 1st method. It is a block diagram which shows the structure in the case of performing phase shift using a phase shift circuit. It is a block diagram explaining the 2nd method. It is a figure explaining the 3rd method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Pump main body 4 Rotor 30 Digital control circuit 31,31A, 31B, 35 D / A converter 33a-33e Sensor circuit 34,34A, 34 A / D converter 36 Excitation amplifier 37A-37E, 204 Phase shift circuit 51-53,51x , 51y, 52x, 52y, 53z Electromagnets 71 to 73, 71x, 71y, 72x, 72y Sensor 313 Sine wave discrete value generation unit

Claims (8)

In a magnetic bearing device that supports a supported body in a non-contact manner by an electromagnet,
Carrier wave generating means for generating a carrier wave signal;
Detecting means for modulating the carrier wave signal according to a support position of the supported body and outputting a modulated wave signal;
The modulation output from the detection means at a sampling frequency fs satisfying fc = n · fs (where n is a natural number) or fc = fs / 2, where fc is the frequency of the carrier signal and fs is the sampling frequency. A / D conversion means for converting a wave signal into a digital signal;
A magnetic bearing device comprising: control means for controlling an excitation current of the electromagnet based on the digital signal output from the A / D conversion means and controlling a support position of the supported body. The magnetic bearing device according to claim 1,
The carrier wave generating means includes a sine wave discrete value generator for generating a sine wave discrete value by digital arithmetic processing, and a D / A converter for generating the carrier signal by D / A converting the sine wave discrete value. And providing a phase shift calculation unit that phase shifts the sine wave discrete value by digital calculation processing,
The timing at which the A / D conversion means converts the modulated wave signal into a digital signal based on the result of the phase shift calculation by the phase shift calculation unit is made to substantially coincide with the envelope of the modulated wave signal. Magnetic bearing device. The magnetic bearing device according to claim 1,
The magnetic bearing device is a multi-axis control type magnetic bearing device having a plurality of control shafts,
The detection means modulates the carrier signal for each of the plurality of control axes and outputs a modulated signal,
The A / D conversion means performs (A) A / D conversion in synchronization with each carrier wave signal from each modulation wave signal output from the detection means, and (b) among the plurality of modulation wave signals. One of the two modulation signals is sampled in synchronism with the timing at which the carrier component changing at the carrier frequency in the modulation wave signal is maximized, and the other is sampled in the modulation wave signal. A magnetic bearing device characterized by sampling in synchronization with timing. The magnetic bearing device according to claim 1,
The magnetic bearing device is a multi-axis control type magnetic bearing device having a plurality of control shafts,
The detection means modulates the carrier signal for each of the plurality of control axes and outputs a modulated signal,
The A / D conversion means performs A / D conversion on each modulated wave signal output from the detecting means in synchronization with the carrier wave signal, and a carrier wave component that changes at the carrier frequency in the modulated wave signal. A magnetic bearing device that samples within a phase range of ¼ of a carrier wave period centered on a phase that is maximum or minimum. In the magnetic bearing device according to claim 3 or 4,
The carrier wave generating means generates a plurality of carrier wave signals having the same frequency and different phases,
The magnetic bearing device according to claim 1, wherein the detecting means modulates any one of the plurality of carrier signals for each of the plurality of control axes and outputs a modulated signal. The magnetic bearing device according to claim 5,
The carrier wave generating means is a sine wave discrete value generator for generating a sine wave discrete value by digital processing, and the sine wave discrete value generated by the sine wave discrete value generator is D / A converted to convert the carrier wave signal A magnetic bearing device comprising: a D / A conversion means to be generated; and a phase shift means having at least one phase shift circuit for shifting the phase of the carrier signal. The magnetic bearing device according to claim 5,
The carrier wave generating means is a sine wave discrete value generator for generating a sine wave discrete value by digital processing, and the sine wave discrete value generated by the sine wave discrete value generator is D / A converted to convert the carrier wave signal A magnetic bearing device comprising a plurality of D / A conversion means to be generated. In the magnetic bearing device according to claim 3 or 4,
A plurality of A / D conversion means are provided, and a plurality of modulated wave signals output from the detection means and having the same phase of the carrier wave component are respectively input to the different A / D conversion means, and A / D conversion is performed at the same timing. A magnetic bearing device that performs D conversion.

Images