|Title of Invention||
A SYSTEM AND METHOD FOR CALIBRATION OF SIGNAL PROCESSING ELECTRONICS
|Abstract||ABSTRACT (A system and method for calibrating signal processing electronics) The invention relates to a system having a random signal generator, which is fed to the input of a wide band voltage amplifier, which would amplify the noise in the microvolt level to a level specified as input signal level of the device under calibration (DUC). This amplified noise signal is a wide band signal having a wide range of frequencies. Hence the required band of interest that has to be used for calibration is filtered using appropriate filter. The filtered random signal is given as input to the DUC. A digital multi meter is used to exactly measure the rms voltage at the output of the filter. The output of the DUC in terms of digital count is acquired in a PC and the input-output ratio is computed as scale factor in niVrms/count. A method of calibrating the signal processing electronics is also disclosed. (Figl)|
System for digital calibration of signal processing electronics for
naturally occurring signals
The invention is in the field of electronics engineering and concerned with digital calibration systems. It is more specifically related to the design and generation of a truly random input signal that simulates the naturally occurring random signals for using in the digital calibration system of vibration and acoustic signal processing electronics. The design exploits the truly random nature of the thermal noise produced across a resistor for the purpose.
Background of the invention:
Thermal noise is produced in the resistor as the electrons in the resistor are subjected to random thermal motions. Because of the nature of the thermal noise, it has a constant power spectral density, i.e., mean square voltage per unit frequency up to very high frequencies. Noise, whose power spectral density is constant, is called white noise. In general, the thermal noise of any connection of passive elements is equal to the thermal noise that would result from the real part of the equivalent total impedance. If we are dealing with pure resistances, the thermal noise is equal to the thermal noise produced by an equivalent resistance. The thermal noise emf of a resistor can be calculated very accurately. The mean square noise voltage present across a circuit element is given by
Erms = rms noise voltage (V)
k = Boltzmann's constant = 1.38xl(T23 (JK"1)
T = the absolute temperature (K)
Z(f) = impedance of the circuit element at frequency f(Q)
fmax ~ maximum frequency
fmin = minimum frequency
If this circuit element is a simple resistor, then mean square noise voltage becomes,
E2rms=4kTR(fmttK- fmin) = AkTRB
B = frequency band being considered = fmax - fmin (Hz)
The power spectral density of this thermal noise is,
Sthermal(f) = 4k.T.R
In most systems, the minimum frequency fmin would be close to zero (DC) so the bandwidth is the maximum frequency fmax. The magnitude of thermal noise is generally in the ju V range. Thus some level of amplification is required to use it in our application. The amplifier output is thus related to input E^ by,
Where A is the amplification factor which can be made independent of frequency by using proper electronic components. The amplifier will be selected in such a way that the amplification factor is independent of frequency in our useful range.
This amplified thermal noise can be used for characterising the naturally occurring random signals like vibration, acoustics etc. by proper band limiting.
An exhaustive prior art search was conducted and the resulting patents and their drawbacks in comparison to the present invention are given below:
i) US Patent No: US4175258, High-level white noise generator
A wide band, stable, random noise source with a high and well-defined output power spectral density is provided which may be used for accurate calibration of Johnson Noise Power Thermometers (JNPT). The noise source is based on the fact that the open-circuit thermal noise voltage of a feedback resistor, connecting the output to the input of a special inverting amplifier, is available at the amplifier output from an equivalent low output impedance caused by the feedback mechanism. This high-level white noise generator is used for the calibration of thermometers that operate on the Johnson noise power phenomenon. Here the noise produced in the feed back resistor of an inverting amplifier is used as such. But, in our application a band limited random signal is used for the calibration of digital signal processing electronics.
ii) W09723032, Digital Calibration Of A Transceiver
Methods and systems for calibrating transceivers are described. Analog transceiver components create errors, such as gain and offset errors in a receiver or a transmitter portion, which are compensated by adjusting digital signal processing in the computing portion of the transceiver. Errors can be measured in a calibration procedure and the determined compensating values will be stored in a memory device of the transceiver. These values can then be retrieved when the transceiver is initialized for usage during signal processing of signals received by
and transmitted by the transceiver. Here the calibration of a transceiver is described for compensating the errors of gain, offset etc. But, our invention describes the calibration of a signal processing circuitry for naturally occurring signals.
iii) US Patent No: US5553623, Method for calibrating a system for recording and playing back ECG signals
An ECG recorder and playback unit that includes software-implemented digital signal processing filters which compensate for phase and magnitude distortion occurring when an ECG signal is recorded on a Hotter recorder and played back. This permits "tuning" a recorder and playback unit, which may be unrelated, as when made by different manufacturers. An impulse or step signal is recorded and played back to provide a system frequency response measurement. Coefficients for a digital correction filter are derived from the discrete Fourier transform of the impulse or step response and a desired system response. When recorded ECG data is played back, it is filtered on a substantially real-time basis with the digital correction filter to compensate for phase and magnitude distortion. Prior to recording, the high frequencies of the ECG signal are boosted to compensate for high frequency losses inherent in the recording and playing back of ECG signals. A calibration pulse and a system characterization pulse are derived by statistically selecting from a plurality of pulses and averaging the selected pulses in order to minimize the effects of noise and tape defects in the pulses. High frequency components of the ECG signal can be amplified prior to recording the signal. This results in decreased amplification, and therefore noise, when the recorded signals are played back. The ECG calibration proposed in the above patent employs the impulse or step response of the ECG recorder- playback system to compute the coefficients of a digital correction filter. This is entirely different from our application.
iv) US Patent No: US5583501, Digital-to-analog converter with digital linearity correction
Digital linearity correction is obtained by generating a digital calibration signal having at least one frequency component, converting the digital calibration signal to an analog signal, and detecting distortion in the analog signal generated from the calibration signal by non-linearity to produce a compensation coefficient used to digitally compensate the digital input of the digital-to-analog converter. The compensation coefficient is adjusted in a feedback loop so that the distortion is minimised. Preferably the calibration signal has two frequencies, and the distortion is an inter-modulation component having a substantially lower frequency. The calibration method can also be used to compensate for non-linearity in signal conditioning circuitry, or to obtain a specified non-linear transfer characteristic. This is not directly related to our application. Above cited calibration technique is for the compensation of non-linearity in the DAC. This is done by using sinusoidal signals of two different frequencies. But, in our application, random signal is used for the calibration of Signal processing electronics.
Objects of the Invention:
The primary object of the invention is to provide a Digital Calibration system for aerospace signal processing electronics.
An additional object of the invention is to have a system, which can provide a truly random signal that represents the naturally occurring random signal.
One more object of the invention is to have a system, which can provide a random source with variable Power Spectral density and Bandwidth.
Another object of the invention is to have a Calibration system that is easy to integrate with the automatic test system of any digital signal processing electronics for naturally occurring signals.
A further object of the invention is to avoid the computer-based complex and time-consuming algorithms that can generate only a psuedo random signal.
One another object is to provide a much compact, cost effective and simple system to generate random signal for calibrating the vibration and acoustics signal processing electronics.
Another object of this invention relates to design, development of a highly versatile system that can be used in any application that requires a random signal input for calibration or characterization.
Brief Description of the Drawings
Figure 1 shows the schematic of the experimental set-up of the calibration system according to this invention.
Figure 2 shows the circuit used to produce thermal noise.
Detailed Description of the Invention:
In many applications, digital signal processing techniques are used to process the information content of naturally occurring signals like vibration, acoustics etc. The calibration of such electronics using a known and measurable random signal is necessary to characterise the system. The space vehicles onboard digital signal processing electronics is used for processing the vibration and acoustics signals in order to reduce the bandwidth requirements of the space Vehicle telemetry system. The present invention relates to the digital calibration system for such electronics.
Schematic of the Calibration System
The present invention incorporates simple and a cost effective digital calibration system, using resistor induced thermal noise. The schematic of the experimental set-up is as shown below is shown in figure 1.
Thermal Noise generated by the noise source (1) is fed to the input of a wide band voltage amplifier (2), which would amplify the noise in the microvolt level to a level specified as input signal level of the device under calibration (DUC) (4). This amplified noise signal is a wide band signal having a wide range of frequencies.
Hence the required band of interest that has to be used for calibration is filtered using appropriate filter (3). The filtered random signal is given as input to the DUC (4). A digital multi meter (5) is used to exactly measure the rms voltage at the output of the filter. The output of the DUC (4) in terms of digital count is acquired with a data acquisition card on a computer (6), which also computes the input-output ratio as scale factor in mV^/count
The noise source (1) in the case of thermal noise is a resistor with some additional circuitry. The thermal noise source (1) used in the set up of the present invention is as shown in figure 2. The input resistor Rl is the noise source. Resistors R2 and R3 are provided for giving necessary amplification.
Typical RMS voltages of Johnson noise signals are of the order of several microvolts. Most of the signal processing circuits will be capable of processing both periodic and random signals over a dynamic range of somewhat more than 10 from several milli volts to several volts. Thus the noise signals have to be
amplified sufficiently to rise to this level. A wide band amplifier is selected so that the gain of the amplifier will be constant throughout the frequency band based on the requirement and the device under calibration.
Thermal noise produced in this method is a broadband noise that stretches over a wide band of frequencies. As the object of the invention is to use this thermal noise for calibration of vibration and acoustics signal processing electronics, which operates only over a specified band of frequencies, the desired band of signals is filtered out. Here a six pole elliptical filter is used to get sharp cut off. At the output of the filter, the mean square voltage of the noise is measured using a Digital Multi Meter (5).
The thermal noise from the thermal noise source (1) produced by the above-explained method will be given as input to the system (4) to be calibrated. Since this noise is purely random in nature, it can be characterised by the power spectral density, which can be computed using equation (1). Moreover since the noise output is band-limited, the bandwidth is also known. Here the gain of the amplifier (2) can be adjusted to get different PSD values so that calibration can be done over its full input range. The PC based data acquisition system used to acquire the output of the device under calibration involves a data acquisition card and its associated software. A digital multi meter (5) is used at the output of the filter to exactly measure the rms voltage. The output of DUC (4) in terms of digital count is acquired in a PC and the input-output ratio is computed as scale factor in mVrms/count.
The random noise source has a simple configuration using inexpensive electronic parts and can generate the true physical random noise with the required spectral properties resulting in a simple and cost effective system.
The system described above can be used for calibrating different types of digital signal processing electronics for naturally occurring signals of any kind like vibration, acoustics etc.
1. A digital calibration system for calibrating signal processing electronics
a random signal generator (1) for generating and supplying random input signals representing the naturally occurring random signals to the device under calibration;
means (2, 3) for feeding the noise generated as input signal to the device under calibration at a specified frequency band and amplification level;
means (6) for acquiring the output of the device under calibration and determining the input/output ratio as scale factor in mV^/count
2. The digital calibration system as claimed in claim 1, wherein the random signal generator is a thermal noise source, which comprises a resistor and circuitry for necessary amplification.
3. The digital calibration system as claimed in claim 1, wherein the means (2,3) for feeding the noise generated as input signal to the device under calibration at a specified frequency band and amplification level is a wide band analog amplifier and an output filter, said wide band amplifier selected such that the gain of the amplifier will be constant throughout the said frequency band.
4. The digital calibration system as claimed in claim 3, wherein the said filter is a six pole elliptical filter.
5. The digital calibration system as claimed in claims 3 or 4, wherein a digital multi meter is provided at the output of the filter to measure the mean square voltage of the noise.
6. The digital calibration system as claimed in claim 1, wherein the means for acquiring the output of the device under calibration is a data acquisition card with associated software for determining the input/output ratio.
7. A method for calibrating signal processing electronics comprising the steps of:
generating and supplying random input signals representing the naturally occurring random signals to the device under calibration;
feeding the noise generated as input signal to the device under calibration at a specified frequency band and amplification level;
acquiring the output of the device under calibration; and
determining the input/output ratio as scale factor in mV^/count.
8. The method as claimed in claim 7, wherein a thermal noise source
comprising a resistor and additional circuitry generates the said random input
9. The method as claimed in claim 7, wherein the step of feeding comprises
amplifying the signal to a level specified as input signal of the device under
calibration and filtering the amplified signal to provide signals of a specified
10. The method as claimed in claim 9, comprising the step of measuring the
mean square voltage of the noise at the output of the said filter.
|Indian Patent Application Number||763/CHE/2007|
|PG Journal Number||08/2012|
|Date of Filing||11-Apr-2007|
|Name of Patentee||INDIAN SPACE RESEARCH ORGANISATION|
|Applicant Address||ANTRIKSH BHAVAN, NEW BEL ROAD, BANGLORE 560094, INDIA|
|PCT International Classification Number||H04N7/01|
|PCT International Application Number||N/A|
|PCT International Filing date|