Title of Invention

AN ANALOG COMPENSATION METHOD AND SYSTEM FOR REDUCING DISTORTION IN A DATA ACQUISITION SYSTEM

Abstract

ABSTRACT "An analog compensation method and system for reducing distortion in a data acquisition system " The present disclosure relates to an analog compensation method for reducing the distortion in a data or signal acquisition system comprising the step of compensating the distorted frequency response characteristics of a sinc3 filter by cascading a compensation filter of complementary characteristics with the said sinc3 filter. The compensation filter can be implemented as an analog 2nd order type 1 Chebyshev filter. Also disclosed is system to reduce the distortion in a data/signal acquisition system comprising a sinc filter for decimation, band limiting and setting the output data rate, and an analog compensation filter with characteristics complementary to the said sinc3 filter wherein the said compensation filter is connected in cascade with the said sinc3 filter. Fig 2

Full Text

Technical Field: The present disclosure is from the field of Electronics and Communication and relates to the compensation of multiplier-less digital filters in data acquisition systems. It is more specifically related to compensation of sinc filters in sigma Delta ADCs. Background and prior art: Digital filters are extensively used in data acquisition and processing systems. A conventional digital filter makes use of multipliers. These multipliers are logic circuit blocks that perform the coefficient multiplication function in each stage of the filter structure. They may work with fixed point numbers (i.e. decimal numbers) supporting fixed precision or floating point numbers that support variable precision. In both the cases the circuit implementation of the multipliers are very complex thus requiring a lot of space on the silicon IC or system board. Further, these digital filters take up a lot of computation time due to delays of multiple logic stages thus slowing down the entire system and limiting the signal frequencies. To overcome these problems digital filter structures that do not employ multipliers have been researched into. Cascaded Integrator Comb (CIC) and sine are some such Multiplier-less digital filter structures that have been extensively discussed in literature and used in components such as sigma-delta ADCs to support data acquisition fimctions. Nevertheless, these filters suffer from a major drawback in their application to data acquisition and processing systems. The characteristic of the filter is distorted when compared to the ideal filter characteristic, as these filters have a droop in the pass-band and ripples in the stop-band. This makes them unsuitable for use in systems where signal fidelity (i.e. precise signal reconstruction in time domain) is important. This is especially so in systems where interfering noise signals in certain fi"equency bands are present. This disclosure intends to propose solutions to this problem so as to enhance the fiinctionality of data acquisition systems employing multiplier-less filters and sigma-delta ADCs in an efficient and cost-effective manner. A thorough search of literature on compensated filters could yield neither any articles nor any patents on analog compensation methods. Our invention is on analog compensation method for achieving flat pass band and ripple fi"ee stop band for the digital filters used in data acquisition systems. Objectives of the invention: The primary objective of the invention is to reduce the signal distortions in data acquisition systems that use sigma delta ADCs. Another objective of the invention is to reduce the droop in the pass-band and ripples in the stop band of a conventional multiplier-less filter in sigma deha ADCs. Another objective is to achieve versatility and cost effectiveness in the application of sigma delta ADCs in data acquisition systems. Summary of the invention: The present disclosure proposes an analog compensation method to overcome the limitations of multiplier-less filters as mentioned above. This is achieved by cascading the sine filter with a compensation filter that has the inverse or complementary characteristics so that the combination of the two filters provides a flat pass-band and a uniform high attenuation in the stop-band. This is done prior to the ADC in analog domain employing an analog, second order Type-I Chebyshev filter. A method for compensating the distorted frequency response characteristics of a sinc^ filter comprising the step of cascading a compensation filter of complementary characteristics with the said sinc^ filer is discussed. The method effectively annuls the pass-band droop and stop-band ripple characteristics of the sine filter to provide a substantially flat pass-band and uniformly attenuated stop-band for the composite filter. In one embodiment of the disclosure the analog compensation filter is of second order Type-I Chebyshev characteristics and is used to compensate a sine filter of 50 Hz bandwidth. In the second embodiment the same type of analog ■J compensation filter is used to compensate a sine filter of 100 Hz bandwidth. The disclosure also describes a system to reduce the distortion in a data or signal acquisition system comprising a sine filter for decimation, band limiting and setting the output data rate, and an analog compensation filter with characteristics complementary to the said sinc^ filter where the said compensation filter is connected in cascade with the said sinc^ filter. Brief description of the drawings: Fig 1 describes the comparison of characteristics of the ideal filter and the sine filter Fig 2 describes the compensated filter as per the present invention Fig 3 describes the implementation of a compensation filter in accordance with the first embodiment of the of the present invention Fig 4 describes the fi-equency response characteristics of the compensated filter and in accordance with the first embodiment of the invention Fig 5 describes the implementation of a compensation filter in accordance with the second embodiment of the of the present invention Fig 6 describes the describes the fi"equency response characteristics of the compensated filter and in accordance with the second embodiment of the invention Fig 7 describes the stop band characteristics of the compensated filter and in accordance with the first embodiment of the invention Detailed Description: Sigma-delta ADCs are a category of ADCs widely used in data acquisition systems especially when high resolution and a combination of signal conditioning and digitization fiincfions are desired. A sine filter is a kind of Cascaded-Integrator-Comb (CIC) filter structure most commonly employed in sigma-delta ADCs for decimation, signal band limiting and setting the output data rate. Figure 1 describes the characteristics of a sine filter in comparison with the 'brick-wair characteristics of an ideal low-pass filter. It may be noted that the pass band of the sinc^ filter is not flat due to a droop towards the pass band edge (1). Further there are ripples in its stop band (2) that lead to non-uniform, low attenuation of the stop band fi-equencies. Both these characteristic features are deleterious, especially for data acquisition applications where signal reconstruction in fime domain is of primary importance. Figure 2 describes the filter as per the present invention and the analog compensation filter. The present disclosure of an analog compensation method is to cascade a compensation filter with the sine filter. The compensation filter has inverse or complementary characteristics to that of a sine filter. When cascaded, the combination of the two filters provides a flat pass band and a uniform high attenuation in the stop band. This cascaded filter combination is referred to as a compensated filter. This compensation of the filters is done in an analog domain prior to the Analog to Digital Conversion. The analog compensation filter can be a second order Type-I Chebyshev filter. The characteristics of this filter are complementary to the sine filter and thus it can be used to compensate the sine filter. This combination is called the compensated filter. Figure 3 describes the implementation of a compensation filter in accordance to first embodiment of the invention. The compensation filter is implemented by means of a Chebyshev filter. The sine filter with 50 Hz cut off Irequency is cascaded with a Chebyshev filter of complementary characteristics. The circuit implementation of the analog compensation filter uses a single operational amplifier and two resistors (Rl, R2) and capacitors (CI, C2) in a non-inverting amplifier configuration. The cut-off frequency (where the filter provides 3-dB attenuation) is taken as 50 Hz for the sine filter and proportionally, 88 Hz for the Chebyshev filter. The cutoff frequency of the Chebyshev filter is designed for 44/25 times the 3dB frequency of the sine' filter. Figure 4 shows the frequency response characteristics of the compensated filter along with the characteristics of the sinc^ filter and the Chebyshev. It can be noted that the pass band of the Chebyshev filter is 'swollen up' towards the edge where the transition band begins shown here with marking (a) and this compensates for the droop in pass-band of the sine filter shown with marking (b) in the figure 4. It can be seen that the compensated filter has a substantially flat pass-band which is shown with marking (c) in the figure 4. The ripple magnitude of the Chebyshev filter (d) is found to be 2.77 dB for optimum pass band flatness of the compensated filter Therefore for a optimum design of the compensation filter the cut-off frequency is designed as 88 Hz (for 50 Hz sinc^ filter), Pass-band ripple, r = 2.77 dB and the filter order, n = 2 Figure 5 describes the implementation of a compensation filter in accordance to second embodiment of the invention. The compensation filter is implemented by means of a Chebyshev filter. The sine filter with 100 Hz cut off frequency is cascaded with a Chebyshev filter of complementary characteristics. In this instance when the cut-off frequency (where the filter provides 3-dB attenuation) is taken as 100 Hz for the sine filter the Chebyshev filter is designed at a cutoff frequency of 176 Hz, which is 44/25 times the 3dB frequency of the sinc^ filter. The pass-band ripple remains the same at 2.77 dB for optimum compensation. In the filter circuit implementation, this implies a halving of the resistor values. This shows that the method can be extrapolated to any filter bandwidth according to the requirement by scaling the 3-dB frequency and the circuit resistor values appropriately. Figure 6 describes the fi-equency response characteristics of the compensated filter and in accordance with the second embodiment of the invention. This is similar to the characteristics noted in the first embodiment except the scaling of fi*equency axis by a factor of 2. Figure 7 describes the stop band characteristics of the compensated filter and in accordance with the first embodiment of the invention. The characteristics of the compensated fiher along with the Chebyshev filter and sine filter is described. It can be noted that the Chebyshev filter suppresses the ripples in the stop-band of the sinc^ filter (e) to a large extent due to its uniform high attenuation (f). The stop band characteristics of the compensated filter in accordance with the second embodiment of the invention is also similar except the scaling of fi-equency axis by a factor of 2. It may be noted that since the compensation filter is an analog one in this method, it is best incorporated prior to the ADC along with preamplifier or antialias filter stages. Simulation studies indicate that the linear phase characterisfics of the sine filter are preserved to a large extent even after cascading with the compensation filter. Thus, the scheme improves the output signal fidelity significantly. Industrial application: The compensation method described above may be applied to data acquisition systems with sigma-delta ADCs. They will help to render the filter characteristics close to ideal so as to cater to situations where the performance requirements demanded of the data acquisition system are stringent. Need for accurate reconstruction of acquired signals in time-domain and high-fidelity signal acquisition in presence of interfering noise signals that fall in pass-band edge and stop-band ripple peak frequencies are typical examples of such situations. Such requirements are frequently encountered in biomedical applications like ECG/EEG monitoring and in industrial instrumentation such as analog controller monitoring. The compensation method in analog domain is best implemented along with the preamplifier/anti-alias filter stages preceding the sigma-delta ADC and costs only one Operational Amplifier IC and a few passive components. ' We Claim: 1. An analog compensation method for reducing the distortion in a data or signal acquisition system comprising the step of compensating the distorted frequency response characteristics of a sinc^ filter by cascading a compensation filter of complementary characteristics with the said sinc3 filter. 2. The method as claimed in claim 1 wherein the said compensation filter is an analog 2nd order type 1 Chebyshev filter. 3. The method as claimed in claim 1 or 2, wherein the said compensation filter is a 2nd order filter with 2.77dB pass band ripple and a cutoff frequency of 44/25 times the 3dB frequency of the said sinc3 filter. 4. A system to reduce the distortion in a data or signal acquisition system comprising a sinc3 filter for decimation, band limiting and setting the output data rate, and an analog compensation filter with characteristics complementary to the said sinc^ filter wherein the said compensation filter is connected in cascade with the said sinc3 filter. 5. The system as claimed in claim 4, wherein the analog compensation filter is an 2nd order analog Chebyshev filter. * 6. The system as claimed in claim 4 or 5, wherein the analog compensation filter is a 2nd order filter with 2.77dB pass band ripple and a cutoff frequency of 44/25 fimes the 3dB frequency of the sinc3 filter. 7. The system as claimed in any of the preceding claims wherein the analog compensation filter is implemented by means of an Operational Amplifier and two resistors (Rl, R2) and capacitors (CI, C2) in a non-inverting amplifier configuration. 8. The system as claimed in any of the preceding claims wherein the said system is implemented prior to analog to digital conversion. 9. The system as claimed in any of the preceding claims wherein the analog compensation filter for double cutoff fi-equency can be implemented by means of halving the resistor values (Rl, R2).

Documents:

3209-CHE-2008 CORRESPONDENCE OTHERS 13-08-2013.pdf

3209-CHE-2008 AMENDED CLAIMS 02-12-2014.pdf

3209-CHE-2008 AMENDED CLAIMS 20-08-2014.pdf

3209-CHE-2008 AMENDED PAGES OF SPECIFICATION 02-12-2014.pdf

3209-CHE-2008 AMENDED PAGES OF SPECIFICATION 20-08-2014.pdf

3209-CHE-2008 EXAMINATION REPORT REPLY RECEIVED 02-12-2014.pdf

3209-CHE-2008 EXAMINATION REPORT REPLY RECEIVED 20-08-2014.pdf

3209-CHE-2008 FORM-13 11-08-2014.pdf

3209-CHE-2008 POWER OF ATTORNEY 20-08-2014.pdf

3209-che-2008 abstract.jpg

3209-che-2008 abstract.pdf

3209-che-2008 claims.pdf

3209-che-2008 description (complete).pdf

3209-che-2008 drawings.pdf

3209-che-2008 form-1.pdf

3209-che-2008 form-26.pdf

3209-che-2008 form-3.pdf

3209-che-2008 form-8.pdf

3209-che-2008correspondence-others.pdf

3209-CHE-2008_Form 13.pdf