Method for automatic frequency adaptation of a filter in a closed loop

Abstract

A method adapts a resonant frequency of a first filter of a closed control loop to a given frequency. The method includes feeding an output signal of a delta sigma modulator of the closed control loop into a frequency adaptation circuit and determining a first noise spectrum of the output signal in a first frequency band and a second noise spectrum of the output signal in a second frequency band. The first frequency band and the second frequency band are arranged symmetrically with respect to the given frequency. The method includes comparing the first noise spectrum with the second noise spectrum, generating an adaptation signal that causes a frequency adaptation of the resonant frequency if the first noise spectrum differs from the second noise spectrum, and outputting the adaptation signal from the frequency adaptation circuit to a control input of the first filter for adapting the resonant frequency.

Claims

1. A method for adapting a resonant frequency of a first filter of a closed control loop to a given frequency, comprising: feeding an output signal of a delta sigma modulator of the closed control loop into a frequency adaptation circuit; determining a first noise spectrum of the output signal of the delta sigma modulator in a first frequency band and a second noise spectrum of the output signal of the delta sigma modulator in a second frequency band in the frequency adaptation circuit, the first frequency band and the second frequency band are arranged symmetrically with respect to the given frequency; comparing the first noise spectrum in the first frequency band with the second noise spectrum in the second frequency band; generating an adaptation signal that causes a frequency adaptation of the resonant frequency if the first noise spectrum differs from the second noise spectrum; and outputting the adaptation signal from the frequency adaptation circuit to a control input of the first filter for adapting the resonant frequency.

2. The method of claim 1, wherein in the comparing step, the first frequency band and the second frequency band are each demodulated individually into a baseband.

3. The method of claim 2, wherein in the comparing step, a noise power in the first frequency band is compared with a noise power in the second frequency band or a signal strength in the first frequency band is compared with a signal strength in the second frequency band.

4. The method of claim 1, wherein the first frequency band with a predetermined width is determined with respect to a first frequency and the second frequency band with a predetermined width is determined with respect to a second frequency.

5. The method of claim 4, wherein the first frequency and the second frequency are provided by a further closed control loop having a second filter or a mechanical resonator with the given frequency as a further resonant frequency.

6. The method of claim 5, wherein the first filter represents a movement of a secondary mass and the second filter or the mechanical resonator represents a movement of a primary mass of a rotation rate sensor.

7. The method of claim 1, wherein the first noise spectrum and the second noise spectrum in the frequency adaptation circuit are determined in a digital domain or in an analog domain.

8. A signal processing device, comprising: a closed control loop with a first filter having a resonant frequency and a delta sigma modulator; and a frequency adaptation circuit receiving an output signal of the delta sigma modulator, determining a first noise spectrum of the output signal of the delta sigma modulator in a first frequency band and a second noise spectrum of the output signal of the delta sigma modulator in a second frequency band, the first frequency band and the second frequency band being arranged symmetrically with respect to a given frequency, comparing the first noise spectrum in the first frequency band with the second noise spectrum in the second frequency band, generating an adaptation signal that causes a frequency adaptation of the resonant frequency if the first noise spectrum differs from the second noise spectrum, and outputting the adaptation signal to a control input of the first filter for adapting the resonant frequency to the given frequency in response to the comparison result.

9. The signal processing device of claim 8, wherein the frequency adaptation circuit includes a first demodulator branch and a second demodulator branch that extract the first noise spectrum in the first frequency band and the second noise spectrum in the second frequency band.

10. The signal processing device of claim 9, wherein the first demodulator branch includes a first multiplier multiplying the output signal of the delta sigma modulator with a first comparison frequency that is higher than the given frequency.

11. The signal processing device of claim 10, wherein the second demodulator branch includes a second multiplier multiplying the output signal of the delta sigma modulator with a second comparison frequency that is lower than the given frequency, the first comparison frequency and the second comparison frequency are arranged symmetrically to the given frequency.

12. The signal processing device of claim 11, wherein the frequency adaptation circuit includes an adder and each of the first demodulator branch and the second demodulator branch has a respective squarer, the outputs of the squarers are connected to the adder, which determines a difference between the signals applied to the outputs of the squarer and outputs a differential value.

13. The signal processing device of claim 11, wherein the frequency adaptation circuit includes an adder and each of the first demodulator branch and the second demodulator branch has a respective absolute-value element determining an absolute value, the outputs of the absolute-value elements are connected to the adder, which determines a difference between the signals applied to the outputs of the absolute-value elements and outputs a differential value.

14. The signal processing device of claim 12, further comprising a controller connected to the adder, the controller generating the adaptation signal from the differential value.

15. The signal processing device of claim 11, further comprising an additional closed control loop having a second filter or a mechanical resonator with the given frequency as a further resonant frequency.

16. The signal processing device of claim 15, wherein the additional closed control loop has a phase-locked control loop providing the first comparison frequency and the second comparison frequency.

17. The signal processing device of claim 8, wherein the frequency adaptation circuit determines the first noise spectrum and the second noise spectrum in a digital domain or in an analog domain.

18. A circuit arrangement for reading a capacitive rotation rate sensor, comprising: a signal processing device including a closed control loop with a first filter having a resonant frequency and a delta sigma modulator, and a frequency adaptation circuit receiving an output signal of the delta sigma modulator, the frequency adaptation circuit determining a first noise spectrum of the output signal of the delta sigma modulator in a first frequency band and a second noise spectrum of the output signal of the delta sigma modulator in a second frequency band, the first frequency band and the second frequency band being arranged symmetrically with respect to a given frequency, comparing the first noise spectrum in the first frequency band with the second noise spectrum in the second frequency band, generating an adaptation signal that causes a frequency adaptation of the resonant frequency if the first noise spectrum differs from the second noise spectrum, and outputting the adaptation signal to a control input of the first filter for adapting the resonant frequency to the given frequency in response to the comparison result; and a control unit configured to calculate and output a rotation rate signal based on the output signal of the delta sigma modulator.

19. The circuit arrangement of claim 18, wherein the given frequency is a frequency of a primary oscillation of the capacitive rotation rate sensor, the capacitive rotation rate sensor having a primary mass and a secondary mass connected to the primary mass, the primary mass being excited to the primary oscillation during operation and the secondary mass being deflected from a position of rest in a direction transverse to the primary oscillation if the rotation rate sensor rotates about a sensitive axis extending transversely to the direction of the primary oscillation and to the direction of deflection of the secondary mass.

20. A rotation rate sensor, comprising: a primary mass; a secondary mass connected to the primary mass, the primary mass being excited to a primary oscillation during operation and the secondary mass being deflected in a direction transverse to the primary oscillation if the rotation rate sensor rotates about a sensitive axis; and a circuit arrangement for reading a rotation rate signal, the circuit arrangement comprising a signal processing device including: a closed control loop with a first filter having a resonant frequency and a delta sigma modulator, and a frequency adaptation circuit receiving an output signal of the delta sigma modulator, the frequency adaptation circuit determining a first noise spectrum of the output signal of the delta sigma modulator in a first frequency band and a second noise spectrum of the output signal of the delta sigma modulator in a second frequency band, the first frequency band and the second frequency band being arranged symmetrically with respect to a given frequency, comparing the first noise spectrum in the first frequency band with the second noise spectrum in the second frequency band, generating an adaptation signal that causes a frequency adaptation of the resonant frequency if the first noise spectrum differs from the second noise spectrum, and outputting the adaptation signal to a control input of the first filter for adapting the resonant frequency to the given frequency in response to the comparison result; and a control unit configured to calculate and output the rotation rate signal based on the output signal of the delta sigma modulator.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention will now be described by way of example with reference to the accompanying Figures, of which:

(2) FIG. 1 is a block diagram of a rotation rate sensor according to the prior art;

(3) FIG. 2 is a block diagram of a secondary control loop for operating a rotation rate sensor according to the prior art;

(4) FIG. 3A is a graph of a spectral power density of an output of the secondary control loop of FIG. 2A in a case where a primary resonant frequency of a rotation rate sensor, a secondary resonant frequency of the rotation rate sensor, and a filter frequency of an electrical filter of a delta sigma modulator are mismatched;

(5) FIG. 3B is a graph of the spectral power density of the output of the secondary control loop of FIG. 2A in a case where the primary resonant frequency, the secondary resonant frequency, and the filter frequency match;

(6) FIG. 4 is a graph of the spectral power density of the output of the secondary control loop of FIG. 2A after an optimum adaptation of the filter frequency of the electrical filter to the primary resonant frequency and in a case where the secondary resonant frequency is mismatched;

(7) FIG. 5A is a graph of the spectral power density of the output of the secondary control loop after an optimum adaptation of a secondary resonant frequency of a rotation rate sensor as well as the frequency bands demodulated into a baseband according to the present invention;

(8) FIG. 5B is a graph of the spectral power density of the output of the secondary control loop in a case where the secondary resonant frequency of the rotation rate sensor to be adapted is mismatched as well as the frequency bands to be demodulated into the baseband according to the present invention;

(9) FIG. 6 is a block diagram of a rotation rate sensor with a frequency adaptation circuit according to an embodiment of the present invention; and

(10) FIG. 7 is a block diagram of a rotation rate sensor with a frequency adaptation circuit according to another embodiment of the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENT(S)

(11) For making the present invention more easily understandable, it will be explained in more detail on the basis of the embodiments shown in the figures following hereinafter. Like components are here provided with like reference numerals and like component designations. In addition, also individual features or combinations of features from the depicted and described embodiments may, when considered individually, represent independent inventive solutions or solutions according to the present invention.

(12) The present invention will now be described in greater detail with reference to FIGS. 5A, 5B, 6, and 7. While the present invention can generally be used for adapting any kind of filters where a broadband excitation of the filters takes place and a transfer function of the filters is approximated on the basis of one of the signals in the application environment, and while the invention can be used for analog-to-digital converters with closed control loops, e.g. delta sigma modulator analog-to-digital converters, the use for rotation rate sensors will be considered in detail hereinafter. Other sensors for which the present invention can be used comprise e.g. Lorentz force magnetic field sensors.

(13) The frequency adaptation principle according to the present invention is here based on an evaluation of the noise shaping (determination of the noise spectrum) of a closed secondary control loop at the output Y of a ΔΣM in a differential band with respect to the primary resonant frequency f.sub.d (of a filter or a mechanical resonator H.sub.p(s)), frequency bands, which have been determined symmetrically to the primary resonant frequency f.sub.d, being here demodulated individually into the baseband, so that the noise power or the signal strength in the two symmetrically determined bands can be compared with each other. This makes it possible to accurately determine whether the current secondary resonant frequency f.sub.s is too high or too low in comparison with the primary resonant frequency f.sub.d, and to generate a signal for setting the secondary resonant frequency f.sub.s. This setting signal is then fed back to the secondary mass m of the sensor, and the secondary resonant frequency f.sub.s is set making use of the spring softening effect.

(14) In view of the differential approach in the frequency range, the setting process will not be interfered with by an existing rotation rate signal F.sub.C. Due to the two sidebands of the amplitude-modulated rotation rate signal F.sub.C, a respective component occurs equally in the f.sub.d+f.sub.l and f.sub.d−f.sub.l path, respectively. In addition, the rotation rate signal can be suppressed by appropriate filters after demodulation. It follows that the frequency adaptation can work in the background during operation, even if parts of a rotation rate signal F.sub.C occur in one of the two bands determined symmetrically to the primary resonant frequency.

(15) FIGS. 5A and 5B show exemplarily the curve of the spectral power density of the output Y of the ΔΣM and thus of the secondary control loop as a function of the frequency f. The output signal satisfies here the above described formula (3); the quantization noise n.sub.q may here be approximated by white noise, by way of example. FIG. 5A schematically illustrates the case where optimum adaptation f.sub.d=f.sub.s=f.sub.f has taken place, whereas FIG. 5B shows an example where the secondary resonant frequency f.sub.s is mismatched with the primary resonant frequency f.sub.d(f.sub.d<f.sub.s in the example shown). According to the present invention, the noise power or the signal strength in a first frequency band (band 1) and in a second frequency band (band 2), which are arranged symmetrically with respect to the primary resonant frequency f.sub.d with a predetermined bandwidth with respect to f.sub.d+f.sub.l or f.sub.d−f.sub.l, are compared with each other. In the event that f.sub.d=f.sub.s=f.sub.f, the curve of the power spectrum extends symmetrically with respect to the resonant frequency f.sub.d of the primary oscillation and the difference between the values for the power density spectrum of band 1 and band 2 disappears (see FIG. 5A). If, however, a differential signal is formed for the bands 1 and 2 in the power density spectrum shown in FIG. 5B, this signal is not equal to zero and can be used to generate an adaptation signal for the input of the filter H.sub.s for the secondary mass m, as will be explained in more detail hereinafter with respect to FIG. 6.

(16) FIG. 6 shows a rotation rate sensor 100 comprising a primary control loop 110, a secondary control loop 120, and a frequency adaptation circuit 130 according to an embodiment of the present invention. Signals with frequencies f.sub.a=f.sub.d+f.sub.l and f.sub.b=f.sub.d−f.sub.l can be provided by a phase-locked control loop PLL (which comprises an oscillator) of the primary control loop 110, these signals thus exhibiting a positive or a negative frequency shift with respect to the primary resonant frequency f.sub.d.

(17) As shown in FIG. 6, an output signal Y of the secondary control loop 120 is fed into an optional digital prefilter of the frequency adaptation circuit 130 and converted into an analog signal. The analog signal is pre-amplified in a filter H.sub.v(s) and frequencies that lie above the primary resonant frequency f.sub.d by more than a predetermined threshold are suppressed in this filter. In the next step, the signal will be demodulated by means of two multipliers and the frequencies f.sub.a=f.sub.d+f.sub.l and f.sub.b=f.sub.d−f.sub.l so as to further evaluate the resultant baseband signals. The multiplication has the effect that signals with frequencies that had previously been above or below f.sub.d by f.sub.l now appear at 0 Hz (DC). This allows other signal components to be removed very easily with downstream low-pass filters. The frequencies f.sub.a,b may be applied to the multipliers either by sinusoidal voltages or, in order to reduce the complexity, by rectangular voltages.

(18) In the following step, either the power or the signal strength of both signals is measured by squaring or approximated by absolute value formation, and, subsequently, the difference between the powers or signal strengths is formed. The difference variable serves as an error value for a controller that advantageously has an I term. The controller may either be operated continuously or may be reset periodically in combination with sign evaluation and a digital logic unit. It will be advantageous to store the value for the control variable V.sub.t obtained after the initial adjustment, so as to obtain the correct secondary resonant frequency f.sub.s adapted to the primary resonant frequency f.sub.d, more quickly during a subsequent start. Finally, the voltage V.sub.t is fed back to the electrodes as a control variable via an amplifier, which typically operates in the high-voltage range (voltage higher than the standard supply voltage of the chip), for setting the secondary resonant frequency f.sub.s of the sensor 100.

(19) While the method according to the present invention takes place with analog signal processing for frequency adaptation in the embodiment shown in FIG. 6, FIG. 7 shows an embodiment in the case of which an implementation with digital signal processing for frequency adaptation is carried out. The rotation rate sensor 200 shown in FIG. 7 comprises a primary control loop 210, a secondary control loop 220, and a frequency adaptation circuit 230. However, unlike in the case of the embodiment shown in FIG. 6, the output signal Y of the secondary control loop 220 is not initially converted into an analog signal in the frequency adaptation circuit 230, but the above-described calculation of the power or signal strength in bands 1 and 2 and of the difference therebetween is executed in the digital range. Again, the difference serves as an error value for a controller and the respective signal output by the controller serves, after having been converted into an analog signal V.sub.t and after amplification of the latter, as a control variable for setting the secondary resonant frequency f.sub.s of the sensor 200. Digital signal processing can be executed equivalently to analog signal processing, or more complex methods, based e.g. on the Goertzel algorithm, can be used to extract the noise power.

(20) The present invention is based, inter alia (in particular in connection with the resonance adaptation of rotation rate sensors; more generally, the task is to be seen in adapting the frequency of a filter upstream of the electrical filter H.sub.f), on the task of adapting the resonant frequency f.sub.s of the secondary mass automatically and during operation to the primary resonant frequency f.sub.d of the rotation rate sensor. The method provided is based on the evaluation of the noise shaping of the closed control loop at the output of the ΔΣM in a differential band with respect to the primary resonant frequency f.sub.d. The noise spectrum of the output signal of the closed control loop is ascertained for two frequency bands determined approximately symmetrically to the primary resonant frequency as a given frequency value, and a frequency adaptation is executed based on an ascertained difference in the noise spectra.

(21) The present invention provides a structural design with a minimum additional investment in circuit technology and low space requirements. In addition, by including a large noise signal band, the frequency adaptation accomplished is significantly faster and more accurate than in cases where discrete signal components in the spectrum of the output Y of the secondary control loop are considered. Since, for this purpose, neither any additional signals have to be fed into the secondary control loop nor is it necessary to separate the secondary control loop for a short time, the functionality for signal readout cannot be impaired.

(22) As has already been mentioned, the field of use of the present invention is not limited to rotation rate sensors alone. The invention is also applied for operating other sensors, such as acceleration sensors, Lorentz force magnetic field sensors, etc.. In addition, especially frequency adaptation during operation can also be used for analog-to-digital converters with closed control loops (e.g. ΔΣM-ADC). Furthermore, the concept can be used with any type of filters in the case of which broadband excitation of the filter takes place and the transfer function of the filter is thus approximated on the basis of one of the signals in the application environment.