GB2424706A

GB2424706A - Solid-state gyroscopes

Info

Publication number: GB2424706A
Application number: GB0506479A
Authority: GB
Inventors: Ian Braithwaite
Original assignee: EUROP TECHNOLOGY FOR BUSINESS; European Technology for Business Ltd
Current assignee: EUROP TECHNOLOGY FOR BUSINESS; European Technology for Business Ltd
Priority date: 2005-03-30
Filing date: 2005-03-30
Publication date: 2006-10-04
Also published as: GB0506479D0

Abstract

A solid-state gyroscope has a piezo-electric sensing element provided with drive and detecting electrodes. A control system supplies a primary drive signal for a primary electrode, a primary input for a pick-off signal from an electrode on the sensing element, and a phase-lock loop (PLL) maintains a phase relationship between the primary drive signal and the signal at the primary input. The control system further supplies a secondary drive signal for an electrode on the sensing element at a node when resonating and a secondary input for a pick-off signal from an electrode on the sensing element where there is a further defined phase relationship with respect to the secondary drive signal such as a node. The secondary drive and pick-off signals are arranged in a negative feedback loop to suppress unwanted sensing element resonation. A synchronous detector acts on the primary and secondary drive signals to produce a rate signal indicative of rotation of the sensing element.

Description

SOLID-STATE GYROSCOPES

This invention concerns solid-state gyroscopes. In particular, this invention relates to a control system for a solid-state gyroscope, to a gyroscope device comprising the combination of a solid-state gyroscope and a control system therefor, and to a method of operating a solid-state gyroscope device.

The main attraction of a solid-state gyroscope (hereinafter referred to as a "gyro") is that it can be physically small, involving no rotating parts and requiring no mechanical bearings. A class of solid-state gyros is based on circularly-symmetric piezo-electric structures, such as rings, cylinders and discs, which form the sensing element or sensor).

This invention particularly, though not exclusively, concerns a solidstate gyro having a disc-shaped piezo-electric sensing element. As such, the invention will primarily be described with reference to a gyro having that form of sensing element though it is to be understood that the invention may be applicable to gyros having other kinds of piezo-electric sensing element.

Piezo-electric materials are well known and have the characteristic that a voltage applied to the material results in a mechanical strain in the material.

Conversely, mechanical strain generates free electric charges which produce a voltage.

With a disc-shaped sensor, it is possible to apply an ac voltage thereto, to make the sensor deform from its normal circular structure to an ellipse with a defined axis of vibration, as shown in Figure 1 (described hereinafter).

Vibrating structures can vibrate in many possible ways, known as "modes", and Figure 1 depicts what is known as the N=2 mode. It is possible to make a disc vibrate in other modes, but the N=2 mode has the advantage of simplicity.

In order to produce a stable vibration with adequate amplitude, the phenomenon of resonance must be used. The mode shapes are produced at particular frequencies, most strongly dependent on material characteristics and the diameter of the structure. For example, prototype solid-state disc gyros have used frequencies in the region of 60 to 120kHz. The resonant modes tend to be very sharply defined with narrow bandwidths, characterised by high- Q factors (the Q factor being the resonant frequency divided by the difference in the half-power frequencies), which for available piezo- electric materials can exceed 1000. In view of these properties, the sensor of a gyro is often referred to as the gyro resonator.

The deformation of the piezo-electric disc is brought about by applying a sinusoidally-varying electrical voltage to a pattern of electrodes deposited on one flat surface of the disc, in conjunction with a conductive plane or a matching electrode pattern on the obverse side of the disc. This sets up an electric field through the thickness of the disc and causes a force orthogonal to the disc plane, resulting in a minute deformation shown greatly exaggerated in Figure 1. The electrodes used to excite the disc lie along the vibration axis.

It can be seen in Figure 1 that along axes at 450 to the vibration axis, there are nodes where no vibration takes place. Electrodes deposited on the surface of the disc along these axes would produce no output voltage as there is no mechanical strain in the disc material at these nodes. This presumes perfect mechanical balance, with perfect circular symmetry in the disc and electrode pattern. In practice, this is impossible to achieve but the principle still applies and so only small output voltages would be seen at electrodes on the axes at 45 to the vibration axis, so long as the disc is not subjected to rotation.

If the disc is rotated about its own axis, the disc particles (at a microscopic level) obey Newton's laws and so tend to continue to move in a straight line in space. Relative to the rotation of the disc, the vibration axis becomes a curve. This is known as the Coriolis effect or force, resulting in a non-zero signal voltage along the 45 axes, the magnitude of which is proportional to the rotation rate of the disc.

It is possible to produce a solid-state gyro using an "open-loop" technique as described above, but there is a number of practical difficulties.

The Coriolis-induced gyro signal is very small. Tiny imbalances give rise to signals (known as "bias") along these axes unrelated to rotation and which are typically much larger than the desired gyro signal, and prone to drift with time and temperature. The small gyro signal requires amplification, stable not only in magnitude but also in phase. The gyro bandwidth, that is, its ability to follow rapid changes in rotation, is limited by the bandwidth of the resonant structure.

It is a principal aim of the present invention to provide a control system for a solid-state gyro which does not suffer from the problems discussed above of an open-loop technique. A further, but subsidiary, aim is to provide both a solid-state gyro device with an improved control system and a method of operating such a gyro device.

According to one aspect of this invention, there is provided a control system for a solid-state gyroscope having a piezo-electric sensing element provided with drive electrodes arranged to cause the sensing element to resonate when supplied with a drive signal and at least one detecting electrode arranged on the sensing element at a location where a signal appears when the sensing element is subjected to rotation, which control system comprises: - a primary circuit having a primary drive signal output for supply to at least one primary drive electrode, a primary input for a primary pick-off signal from an electrode disposed on the sensing element where there is a signal with a defined phase relationship with respect to the primary drive signal, and a phase-lock loop to maintain the defined phase relationship between the primary drive signal output and the signal at the primary input when the sensing element is resonating; - a secondary circuit having a secondary drive signal output for supply to at least one secondary drive electrode provided on the sensing element at a node when resonating, and a secondary input for a secondary pick-off signal from a detecting electrode disposed on the sensing element where there is a signal with a further defined phase relationship with respect to the secondary drive signal, the secondary circuit and secondary pick- off signal being arranged as a negative feedback loop for the secondary drive signal output whereby the secondary drive signal serves to suppress unwanted sensing element resonation; and - a synchronous detector acting on the primary and secondary drive signal outputs to produce a rate signal indicative of the rotation of the sensing element.

The above control system is generally applicable to a gyro having a circularly-symmetrical piezo-electric sensing element, which does not necessarily have to be in the form of a disc. However, and as mentioned above, this invention is particularly applicable to a solid-state gyro having a sensing element in the form of a disc and which is caused to resonate in the N=2 mode. Consequently, this invention also provides a control system for use with a disc-shaped sensing element provided with at least one drive electrode disposed on a first diameter of one face of the disc and which drive electrode is arranged to cause the disc to resonate in the N=2 mode when supplied with an appropriate drive signal, and a detecting electrode disposed on said face of the disc on a second diameter at substantially 450 to said first diameter and at which detecting electrode a signal appears when the disc is subjected to rotation, which control system comprises: - a primary circuit having a primary drive signal output for supply to at least one primary drive electrode, a primary input for a primary pick-off signal from an electrode disposed on said face of the disc on a third diameter at substantially 900 to said first diameter and where there is a signal with a defined phase relationship with respect to the primary drive signal, and a phase-lock loop to maintain the defined phase relationship between the primary drive signal output and the signal at the primary input when the disc is resonating; - a secondary circuit having a secondary drive signal output for supply to at least one secondary drive electrode disposed on said face of the disc on a fourth diameter at substantially 900 to said second diameter, and a secondary input for a secondary pick-off signal from said detecting electrode disposed on said second diameter and where there is a signal with a defined phase relationship with respect to the secondary drive signal, the secondary circuit being arranged as a negative feedback loop for the secondary drive signal output whereby the secondary drive signal serves to suppress unwanted resonation; and - a synchronous detector acting on the primary and secondary drive signal outputs to produce a rate signal indicative of the rotation of the disc.

The control systems of this invention allow for novel operating methods for a solid-state gyro with a disc-shaped piezo-electric sensing element. As such, yet another aspect of this invention provides a method of operating a solid-state gyroscope device including a sensing element in the form of a disc of piezo-electric material provided with at least one primary drive electrode disposed on a first diameter of one face of the disc, a detecting electrode disposed on said face of the disc on a second diameter at substantially 45 to said first diameter, a primary pick-off electrode disposed on said face of the disc on a third diameter at substantially 90 to said first diameter, and at least one secondary drive electrode disposed on said face of the disc on a fourth diameter at substantially 90 to said second diameter, in combination with a control system having primary and secondary circuits both connected to the electrodes of the disc, in which method: - a primary drive signal is output by the primary circuit to said at least one primary drive electrode so as to cause the disc to resonate in the N=2 mode whereby a signal appears at the detecting electrode when the disc is subjected to rotation; - the primary drive signal is maintained at a defined phase relationship to a primary pick-off signal from said primary pick-off electrode at resonance of the disc by a phase-lock loop in the primary circuit; - a secondary drive signal is output by a secondary circuit to said at least one secondary drive electrode, and a secondary pick-off signal from said detecting electrode is used in a negative feedback loop to control the secondary drive signal whereby the secondary drive signal serves to suppress unwanted resonation; and - the primary and secondary drive signals are supplied to a synchronous detector to produce a rate signal indicative of the rotation of the disc.

It will be appreciated that the negative feedback loop provided in this invention may be enhanced in order to optimise the signal processing and so largely to eliminate resonances of the disc in modes other than the N=2 mode.

Most preferably, this is achieved by operating with relatively high gain and narrow bandwidth in the feedback path. In the particularly preferred feedback technique, the signal to be processed is sampled down to dc or "zero frequency" and then a low-pass filter is applied. Then, by resampling back to a high frequency, a bandpass filter operation will have been performed on the signal. Low-pass filters of low bandwidth are relatively easy to construct, though as the bandwidth is reduced, the components become physically larger.

Such processing techniques are known in a radio environment where the signals to be processed come from a distant transmitter. By contrast, in the control system of this invention, expressly for a solid-state gyro, the excitation signal is always present at the N=2 operating frequency, so that the system will always be truly synchronous and will track perfectly any change in the operating frequency.

The defined phase relationship between the primary drive signal output and the primary pick-off signal, and also the further defined phase relationship between the secondary drive signal output and the secondary pick-off signal, are both preferably one of in-phase or phase-quadrature, though it will be appreciated that it would be possible to arrange a system where other phase relationships apply.

The negative feedback loop should have a relatively high loop gain at the resonant frequency of the sensing element and a relatively low loop gain at other frequencies. Preferably, the loop includes an amplifier having a high open-loop gain for processing the secondary pick-off signal, the signal from the amplifier then being filtered by a high frequency lowpass filter.

In order to maintain full phase information in the feedback loop, the secondary circuit preferably has at least two channels for processing the secondary pick-off signal and which operate in-phase and in phasequadrature with respect to the primary drive signal output. Each of these channels may have a respective first sampling gate operating on the secondary pick-off signal, the two first sampling gates being driven in phase-quadrature. The sampled signals may then be passed through respective low frequency low- pass filters to reduce the frequency of the respective output to zero.

After filtering, the output of each low-pass low frequency filter may be supplied to a respective further sampling gate, the further sampling gate of the in-phase channel operating at 1800 out of phase with respect to the first sampling gate of the in-phase channel, and the further sampling gate of the phase-quadrature channel operating at 180 out of phase with respect to the first sampling gate of the phase-quadrature channel. The outputs of these two further sampling gates may be combined and passed through a high frequency low-pass filter. Finally, a phase-shifter may be arranged to process the output of that filter to compensate for phase shifts in the processing of the secondary pick-off signal.

It is highly preferred that the control system includes a single oscillator arranged to provide the drive signal for causing the sensing element to resonate at the N=2 frequency, and also to control operation of the negative feedback loop of the secondary circuit. In this way, precise synchronicity can be assured, with precise tracking of the secondary drive signal.

Synchronous detectors are well known and the control system may employ a suitable form of detector, to produce the rate signal indicative of the rotation of the sensing element. For example, the preferred embodiment utilises a voltage-controlled switch operating at the frequency of the primary drive signal output and acting on the secondary drive signal output of the - 10- secondary circuit; the output of the switch may then be amplified and filtered to yield the required rate information.

By way of example only, one embodiment of control system for a solidstate gyro having a piezo-electric sensing element in the form of a disc arranged to resonate in the N=2 mode will now be described in further detail, reference being made to the accompanying drawings, in which: Figure 1 shows vibration of a circular structure in the N=2 mode, with Figure 1(a) showing the major axis lying along the axis of vibration, and Figure 1(b) a half cycle later, at right angles to it; Figure 2 is an outline system diagram of a solid-state gyro and a control system of this invention; Figure 3 is a block diagram of the secondary loop used in the system of Figure 2; Figure 4a is a virtual trace showing sampling gate operation (sampling to dc) of the in-phase channel of the secondary loop; Figure 4b is a virtual trace showing sampling gate operation (sampling to dc) of the quadrature channel of the secondary loop; Figure 5 is a circuit diagram of the low frequency low-pass filter used in the secondary loop; Figure 6 shows the output of the second set of sampling gates for two different amplitudes and phases of secondary pick-off signal, the period of the waveform being that of the operating frequency; Figure 7 is a circuit diagram of sampling pulse generator; Figure 8 is a virtual trace of input and sampling pulses obtained by the circuit of Figure 7, showing the 00, 90 , 180 and 270 pulses, in sequence, upwards from the bottom of the screen; and Figure 9 illustrates dc offset nulling applied to the secondary loop of Figure 3.

As discussed above, Figure 1 shows a piezo-eiectric disc sensor resonating in the N=2 mode, with the major axis in line with the axis of rotation in Figure la and at right angles thereto in Figure ib, haifa cycle later. On axes at 45 to the major axis, there are nodes where there is no mechanical strain.

Figure 2 shows the outline block diagram of a closed-loop gyro system, having a secondary, negative feedback loop. The resonator disc 10 is shown diagrammatically as having a pattern of electrodes arranged equally spaced around the disc, on one flat face thereof. The opposed flat face has a common "ground" (alternatively "earth" or "zero volt" or "OV") electrode, but in the interests of clarity this is not shown nor the connections thereto. Further, Figure 2 shows drive electrodes 1 & 5 and 4 & 8 connected for "double-driving" while pick-off electrodes 6 & 3 are used singly. In practice, all combinations of single and double driving or single and double pick-off will work.

In use, the disc is set into vibration by a sinusoidal primary drive voltage applied to electrodes 1 and 5. in order to maintain this at the correct frequency, a feedback system, the primary loop, is employed. A signal is picked off the disc by electrode 3, and a phase-locked loop adjusts the drive frequency until the drive and pick-off signals have the correct phase relationship. The primary loop system also contains an amplitude control, which keeps the signal from the disc at constant amplitude, compensating for changes in coupling of signals from drive to pick-off electrodes.

In a well-balanced disc, signals at diametrically opposed electrodes should be the same, so, for example, electrode 7 could be used as primary pick-off, or electrodes 3 and 7 could be connected in parallel. Similarly, electrode 1 or electrode 5 could be used as a drive electrode on its own, but electrodes 1 and 5 are connected together to increase the excitation of the disc.

A signal could be taken from the disc at any even-numbered electrode and processed as the gyro signal in an open-loop manner, as discussed above.

The system shown in Figure 2, however, has a secondary loop the function of which is to amplify the signals picked off the disc ("secondary pickoff', electrode 6 in Figure 2), and apply the amplified signals as a drive ("secondary drive", electrodes 4 & 8 in Figure 2). This drive signal acts to suppress vibration along the 450 axes - the lines joining electrodes 2 & 6 and 4 & 8 in Figure 2. In this negative feedback loop, the secondary pick-off signal now becomes a loop error signal, which can be reduced in proportion to the open- loop gain, in line with classical feedback theory. The secondary drive signal becomes the output signal, and by operation of the feedback loop, is of the correct amplitude and phase to reduce the error signal to a small level.

If the resonator is rotated, a gyro signal is generated, consisting of fluctuations in the secondary drive signal which, it can be demonstrated both theoretically and practically, are in phase with the primary drive signal.

Extraction of the gyro signal may be performed by using synchronous detection (essentially consisting of a switch followed by a low-pass filter and dc amplifier), with the primary drive operating the switch.

The desirable qualities for this, as for all, negative feedback loop is: 1) High open loop gain (which will be referred to generally as simply "loop gain"), which reduces the error signal to a low level and makes the system insensitive to changes in open-loop gain and phase, caused by such influences as temperature, component spread and ageing.

2) Stability under all conditions, with an adequate margin of stability. At worst, stability can be sufficiently poor that the loop oscillates. If the stability margin is not adequate, the response to dynamic changes, such as impulses, will be poor.

In all real systems, these two qualities are in conflict, and the gyro secondary loop has presented a significant challenge to the development of a suitable gyro electronic control system.

The intended meaning of "high loop gain" depends to a degree on the context, the aim being "high enough" for the application. Lower than this, and adequate performance will not be attained. Higher, and the design will be over- engineered and probably costly.

In the context of a disc gyro, a determination of adequate loop gain has been arrived at by experiment. A low gain secondary loop prototype, designed along more conventional lines than the technique to be described, may operate in many ways satisfactorily, but demonstrates sensitivity to changes in gain and phase within the secondary loop.

- 14 - Obtaining a high loop gain at a frequency in the region of 115kHz is not in itself difficult with modern devices: what is difficult is achieving a stable system. This is so because the resonator has, in addition to the wanted N=2 mode, other resonant modes. Some of these modes, due to unwanted out-of- plane motion of the disc and referred to as "bending modes", can be quite close in frequency to the N=2 mode, perhaps within 20%, and are more easily excited than the N=2 mode. This means that a given excitation voltage applied to a drive electrode at the resonant frequency of the unwanted mode will produce a greater voltage at the pick-off electrode than for the N=2 mode.

With these unwanted resonant modes come phase shifts, which vary rapidly with frequency. If a feedback loop possesses a loop gain of greater than unity and a total phase shift round the loop of 360 or a multiple thereof, it will inescapably oscillate. Much of the art of feedback loop design is in achieving high gain where it is needed and disposing of gain at other frequencies in order to achieve stability.

The disc gyro requires high secondary loop gain at its operating N=2 mode frequency and less than unity loop gain at all other resonant modes, this being hardest to achieve at the modes closest in frequency. This dictates that the open loop bandwidth must be small, in order that the gain is sufficiently reduced at the nearest resonant mode. A further problem is that the N=2 mode frequency varies between samples of discs, due to manufacturing tolerances. C)

- 15 - Figure 3 shows an embodiment of the secondary loop system of this invention, in block diagram outline form to aid clarity. Some additional details will be given following an explanation of the block diagram.

The secondary pick-off signal is amplified by a large factor, nominally the open-loop gain of the loop, by amplifier 11. The signal is filtered by a high frequency low-pass filter 12 to remove harmonics of the operating frequency, which result from two sources: distortion in the drive voltage, and non-linearity within the piezo-electric material itself. Since harmonics can affect the result of the sampling operation, it is prudent to minimise them, though not absolutely necessary; working prototypes have been built without this facility. As an example, low-pass filter 12 has a cut-off frequency in the region of 135kHz, for an operating frequency of around 115kHz.

Low-pass filter 12 can be made from inductors and capacitors (LC filters) or may be an active filter based on an operational amplifier, resistors and inductors.

The amplified and filtered secondary pick-off signal is then sampled, using two voltage-controlled switches or sampling gates 13. The opening and closing of the switches 13 is done by sampling pulses which are in phase quadrature (900 apart), provided by divider logic 14. In tests, the pulse width was one quarter of the period of the frequency of operation (again, the N=2 mode), but longer or shorter pulse widths could be used, provided they have the same pulse width and are phased apart by a quarter of a cycle.

The sampling switch outputs are termed the I (In-phase) and Q (Quadrature) channels. It is essential to use at least two channels in order to be able later to re-constitute a sinusoidal signal of the correct phase. If only one channel were used, all phase information would be lost.

The sampling gate outputs contain high frequency energy, but the signal of interest is at zero frequency (or "dc"). This is selected, and high frequencies rejected, by low frequency, low-pass filters 15, one for each channel.

Sampling is a very common process in electronics, but the exact conditions vary, so those existing in a prototype of the circuit of Figure 3 are illustrated, for the I-channel. Figure 4a shows a simulated oscilloscope screen.

The upper waveform is the rectangular sampling pulse with a positive duration of one quarter of a cycle, and the central trace is the sinusoidal waveform (at the same frequency) to be sampled. The sampling switch is closed only for the positive sampling pulse duration, allowing a portion of the sinusoidal input waveform to appear at the switch output the lower trace.

Figure 4b shows in more detail the sampling gate output waveform, which is magnified, and with zero volts as the centre line. The steady trace is this signal after low-pass filtering, which removes the varying component to leave only dc. It should appear intuitively reasonable that the dc output of the sampling gate will change if there is a change in either (or both) the amplitude or phase of the sinusoidal signal being sampled.

As noted, the operation of only one channel's sampling gate has been illustrated. Operation of the other channel is identical, and results in another dc output. Since the two sampling gates are supplied with the same sinusoidal signal input, differences in sampling gate outputs of the two channels reflect the fact that the same signal is being sampled at different times, and so having - 17- both I and Q channels provides information not only on amplitude, but also phase.

The low-pass filters 15 determine the stability of the loop, and can be made using resistors and capacitors (an "RC" filter). They set the frequency at which the open-loop gain starts to decrease and typically also contain components chosen to tailor the open-loop frequency response in order to produce a stable design, which entails avoiding excessive phase shift. Figure shows an implementation of the low-pass filter 15; components Ri and Cl set the low frequency "cut-off', the point at which the filter output starts to fall with increasing frequency. R2 reduces the phase shift of the filter in the frequency range where the open loop gain falls through unity, and C2 reduces the output further at higher frequencies to keep the open loop gain below unity at the unwanted resonance frequencies.

Referring back to Figure 3, the sampled dc voltages, filtered to define the loop characteristics, contain all the information needed to reconstitute a signal at the operating frequency, and this is done by resampling, using a further pair of sampling gates 16, identical to gates 13 and again driven at phase quadrature by the divider logic 14. Example voltage waveforms, as seen on an oscilloscope monitoring the outputs of the further pair of sampling gates 16, are shown in Figure 6. The two waveforms (a) and (b) have a period equal to that of the operating frequency, and both contain signals at the fundamental N=2 operating frequency, plus harmonics. The fundamental components differ in both phase and amplitude, determined unambiguously by the dc outputs of the first pair of sampling gates. - 18-

If the gyro resonator had only one pure resonant mode, this waveform could be applied to it, and the resonator would pick out or accept the fundamental, and reject the harmonics. However, the disc has higher frequency resonant modes or "overtones", which would be excited by thewaveform's harmonics. This is undesirable, so the waveform is low-pass filtered by a further high frequency low-pass filter 17, leaving, for practical purposes, only the fundamental sinusoidal component.

The signal processing applied thus far is accompanied by a phase shift, which could make the secondary loop unstable. To accommodate this, a phase-shifter 18 is employed as the final signal-processing element, the output of which is applied to the disc as the secondary drive. The phaseshifter may comprise an "all-pass filter" with phase adjustment provided by variable resistors. Many other implementations of phase-shifter are possible, including low- or high-pass filters using resistors and capacitors or inductors and capacitors, and active filters of various kinds.

If the secondary loop were to be broken following the phase-shifter, and a signal at the operating frequency applied to the disc secondary drive electrode(s), the phase-shifter output would ideally be in anti-phase (that is, 1800) with respect to this drive signal, and the loop can be set up for this condition. In practice, the loop is stable and functions satisfactorily over a significant range of phase shifts, of the order of tens of degrees, making the system quite tolerant of phase shift adjustment.

An embodiment of the sampling technique and low-frequency low-pass filter system described above has a cut-off frequency of 0.07Hz. Thus, the equivalent of an extremely narrow band-pass filter has been created, having twice this bandwidth, 0.144Hz, and which tracks the operating frequency of the resonator perfectly. As explained earlier, this narrow bandwidth is not required for its intrinsic virtue, but so that a high gain, stable negative feedback can be established, which does not oscillate at a close unwanted resonator mode. The open-loop gain used in the most advanced prototype was 70dB, or somewhat over a factor of 3000. This has proved sufficient to render negligible any changes in gyro output due to phase-shifter adjustment.

A few further details of a practical implementation will now be given.

The divider logic 14 that provides the sampling pulses may comprise integrated circuits of a common logic family to produce the pulses from an input signal at four times the operating frequency, produced by an oscillator. The circuit, based on D-type flip-flops and AND gates, is shown in Figure 7, though there are many other possible implementations. The input and four sampling pulse outputs produced by the circuit are shown in the simulated oscilloscope display of Figure 8.

As noted earlier, the pulses need not be one quarter of the period of the operating frequency, provided an accurate phasing of 90 is maintained between the I and Q channels. The pulses could be produced in a different order, but this would alter the phase shift through the loop, requiring the phase- shifter to be adjusted differently or modified. The phase shifts could, less elegantly, be obtained using 90 phase-shifters.

As shown in the circuit diagram, the primary drive signal is also derived from the logic 14 and is applied to the resonator after low-pass filtering by filter - 20 - 19 to reject harmonics. The oscillator frequency (divided by four) is maintained at the correct resonator frequency by means of a phase-locked loop incorporating the resonator, in a manner known per so.

A detail omitted from the outline system diagram of Figure 3 concerns dc offsets. The first sampling system produces dc outputs offset from their correct values due to component imperfections. It is easiest to visualise their effect by considering the secondary loop with no input signal. Ideal components would produce I and Q channel voltages of zero, but in reality there will be small dc voltages, which, applied to the second sampling gates and filtered, result in a small signal at the operating frequency. It is straightforward to add small offsets to cancel these and null this signal, as shown diagrammatically in Figure 9. These offsets can conveniently be derived using potentiometers fed from regulated power supply rails. They are applied at summing stages 20, conveniently implemented by operational amplifiers (op-amps).

Figure 3 shows all the gain being applied prior to the sampling operation.

This is good practice, since it helps to minimise the effect of dc offsets in the sampling system. However, the sampling process and real low-pass filters incur some loss, which is made up by appropriate gain stages added after the loss.

Though the invention has been described with reference to a disc having an electrode structure on one face and a simple earth electrode on the opposed face (a so-called single-sided disc), it is equally applicable to a disc having corresponding electrode structures on its opposed external faces and an internal central earth electrode - a so-called double-sided disc.

Claims

1. A control system for a solid-state gyroscope having a piezo-electric sensing element provided with drive electrodes arranged to cause the sensing element to resonate when supplied with a drive signal and at least one detecting electrode arranged on the sensing element at a location where a signal appears when the sensing element is subjected to rotation, which control system comprises: - a primary circuit having a primary drive signal output for supply to at least one primary drive electrode, a primary input for a primary pick-off signal from an electrode disposed on the sensing element where there is a signal with a defined phase relationship with respect to the primary drive signal, and a phase-lock loop to maintain the defined phase relationship between the primary drive signal output and the signal at the primary input when the sensing element is resonating; - a secondary circuit having a secondary drive signal output for supply to at least one secondary drive electrode provided on the sensing element at a node when resonating, and a secondary input for a secondary pick-off signal from a detecting electrode disposed on the sensing element where there is a signal with a further defined phase relationship with respect to the secondary drive signal, the secondary circuit and secondary pick- off signal being arranged as a negative feedback loop for the secondary drive signal output whereby the secondary drive signal serves to suppress unwanted sensing element resonation; and 22 - - a synchronous detector acting on the primary and secondary drive signal outputs to produce a rate signal indicative of the rotation of the sensing element.

2. A control system for a solid-state gyroscope having a piezo-electric sensing element in the form of a disc provided with at least one drive electrode disposed on a first diameter of one face of the disc and which drive electrode is arranged to cause the disc to resonate in the N=2 mode when supplied with a drive signal, and a detecting electrode disposed on said face of the disc on a second diameter at substantially 450 to said first diameter and at which detecting electrode a signal appears when the disc is subjected to rotation, which control system comprises: - a primary circuit having a primary drive signal output for supply to at least one primary drive electrode, a primary input for a primary pick-off signal from an electrode disposed on said face of the disc on a third diameter at substantially 900 to said first diameter and where there is a signal with a defined phase relationship with respect to the primary drive signal, and a phase-lock loop to maintain the defined phase relationship between the primary drive signal output and the signal at the primary input when the disc is resonating; - a secondary circuit having a secondary drive signal output for supply to at least one secondary drive electrode disposed on said face of the disc on a fourth diameter at substantially 900 to said second diameter, and a secondary input for a secondary pick-off signal from said detecting electrode disposed on said second diameter and where there is a signal with a defined phase - 23 - relationship with respect to the secondary drive signal, the secondary circuit being arranged as a negative feedback loop for the secondary drive signal output whereby the secondary drive signal serves to suppress unwanted resonation; and - a synchronous detector acting on the primary and secondary drive signal outputs to produce a rate signal indicative of the rotation of the disc.

3. A control system as claimed in claim 1 or claim 2, wherein secondary circuit has a relatively high loop gain at the resonant frequency of the sensing element and a relatively low loop gain at other frequencies.

4. A control system as claimed in claim 3, wherein the negative feedback loop includes an amplifier having a high open-loop gain, for processing the secondary pick-off signal.

5. A control system as claimed in claim 4, wherein the secondary circuit includes a low-pass filter for processing the signal from the amplifier.

6. A control system as claimed in any of the preceding claims, wherein the secondary circuit has two channels for processing the secondary pick- off signal and which operate in-phase and in phase-quadrature with respect to the primary drive signal output.

7. A control system as claimed in claim 6, wherein each of the two channels has a respective first sampling gate operating on the secondary pick- off signal, the two first sampling gates being driven in phase-quadrature.

8. A control system as claimed in claim 7, wherein the output of each first sampling gate is passed through a respective low frequency, lowpass filter to reduce the frequency of the respective output to zero.

- 24 -

9. A control system as claimed in claim 8, wherein the output of each low- pass low frequency filter is supplied to a respective further sampling gate, the further sampling gate of the in-phase channel operating at 1800 out of phase with respect to the first sampling gate of the in-phase channel, and the further sampling gate of the phase-quadrature channel operating at 180 out of phase with respect to the first sampling gate of the phase-quadrature channel.

10. A control system as claimed in claim 9, wherein the outputs of the two further sampling gates are combined and passed through a high-frequency low- pass filter.

11. A control system as claimed in claim 10, wherein a phase-shifter is arranged to process the output of the high-frequency low-pass filter to compensate for phase shifts in the processing of the secondary pick-off signal.

12. A control system as claimed in any of the preceding claims, wherein a single oscillator is arranged to provide the primary circuit with a signal at the required frequency for resonance of the sensing element, and to control operation of the negative feedback loop for the secondary drive signal.

13. A control system as claimed in claim 12, wherein a divider logic is provided to operate on the output of the single oscillator and to produce the primary drive signal and control signals for the negative feedback loop of the secondary circuit.

14. A control system as claimed in any of the preceding claims, wherein the secondary circuit includes compensation means for the negative feedback loop, consequent upon a non-zero error signal when the sensing element is not subjected to rotation.

- 25 -

15. A control system as claimed in claim 14 when appendent to claim 8, wherein there is provided means to add respective dc biases to the outputs of the first sampling gates.

16. A control system as claimed in any of the preceding claims, wherein the synchronous detector comprises a voltage-controlled switch operating on the secondary drive signal output under the control of and at the frequency of the primary drive signal output.

17. A control system as claimed in claim 16, wherein a low-pass filter and dc amplifier are arranged to operate on the output of the frequencycontrolled switch, to produce the rate signal indicative of the rotation of the sensing element.

18. A control system as claimed in any of the preceding claims, wherein the phase-lock loop of the primary circuit includes amplitude control, to maintain substantially constant the amplitude of the primary pick-off signal.

19. A solid-state gyroscope device comprising a control system as claimed in claim 2 or any claim appendent thereto in combination with a sensing element in the form of a disc of piezo-electric material, the disc having electrodes formed on the faces thereof and to which the control system is connected.

20. A solid-state gyroscope device as claimed in claim 19, wherein one face of the disc has a plurality of electrodes formed thereon and to which the control system is connected, and the other face of the disc has a ground electrode formed thereon.

- 26 -

21. A method of operating a solid-state gyroscope device including a sensing element in the form of a disc of piezo-electric material provided with at least one primary drive electrode disposed on a first diameter of one face of the disc, a detecting electrode disposed on said face of the disc on a second diameter at substantially 450 to said first diameter, a primary pick-off electrode disposed on said face of the disc on a third diameter at substantially 900 to said first diameter, and at least one secondary drive electrode disposed on said face of the disc on a fourth diameter at substantially 90 to said second diameter, in combination with a control system having primary and secondary circuits both connected to the electrodes of the disc, in which method: - a primary drive signal is output by the primary circuit to said at least one primary drive electrode so as to cause the disc to resonate in the N=2 mode whereby a signal appears at the detecting electrode when the disc is subjected to rotation; - the primary drive signal is maintained at a defined phase relationship to a primary pick-off signal from said primary pick-off electrode at resonance of the disc by a phase-lock loop in the primary circuit; - a secondary drive signal is output by a secondary circuit to said at least one secondary drive electrode, and a secondary pick-off signal from said detecting electrode is used in a negative feedback loop to control the secondary drive signal whereby the secondary drive signal serves to suppress unwanted resonation; and - 27 - - the primary and secondary drive signals are supplied to a synchronous detector to produce a rate signal indicative of the rotation of the disc.

22. A control system for a solid-state gyroscope as claimed in claim 2 and substantially as hereinbefore described, with reference to and as illustrated in the accompanying drawings.

23. A solid-state gyroscope device as claimed in claim 19 and substantially as hereinbefore described, with reference to and as illustrated in the accompanying drawings.

24. A method of operating a solid-state gyroscope device as claimed in claim 21 and substantially as hereinbefore described, with reference to the accompanying drawings.