Filter block for compensating droop in a frequency response of a signal
View Patent ↗The invention may provide a method and filter block for compensating droop in a frequency response of a signal. The filter block may include a decimator, which decimates a high frequency input signal to a set frequency output signal. The set frequency can be, for example, the Nyquist frequency for the input signal. Further, the filter block may include a droop compensator that compensates the droop in the frequency response of the output signal from the decimator. The droop compensator may be made using recursive filters, as opposed to large tap FIR filters, which may result in less memory consumption and decreased power consumption.
1. A filter block comprising:
a decimator configured to decimate an input signal to an output signal having a set frequency; and
a droop compensator configured to compensate for pass band droop in the output signal from said decimator and comprising at least one recursive filter.
2. The filter block according to claim 1 wherein said decimator is configured to decimate the input signal into the output signal with set frequency in a single step.
3. The filter block according to claim 1 wherein said decimator comprises a Cascaded Integrator-Comb (CIC) filter.
4. The filter block according to claim 1 further comprising an all pass filter section downstream from said droop compensator.
5. The filter block according to claim 4 wherein said all pass filter section is configured to correct a phase distortion from an output of said droop compensator.
6. The filter block according to claim 4 wherein said droop compensator is cascaded with said all pass filter section.
7. The filter block according to claim 1 wherein said droop compensator comprises at least one finite impulse response (FIR) filter.
8. The filter block according to claim 1 wherein the set frequency comprises a Nyquist frequency of the input signal.
9. A filter block comprising:
a decimator configured to decimate an input signal to an output signal having a set frequency;
a droop compensator configured to compensate for pass band droop in the output signal from said decimator and comprising a plurality of low order recursive filters; and
an all pass filter section downstream from said droop compensator and configured to correct a phase distortion from an output from said droop compensator.
10. The filter block according to claim 9 wherein said decimator is configured to decimate the input signal into the output signal with the set frequency in a single step.
11. The filter block according to claim 9 wherein said decimator comprises a Cascaded Integrator-Comb (CIC) filter.
12. The filter block according to claim 9 wherein said droop compensator is cascaded with said all pass filter section.
13. The filter block according to claim 9 wherein said droop compensator comprises at least one finite impulse response (FIR) filter.
14. The filter block according to claim 9 wherein the set frequency comprises a Nyquist frequency of the input signal.
15. A method implemented by a filter block to compensate for pass band droop in a filtered signal, the method comprising:
decimating an input signal to an output signal having a set frequency; and
compensating for the pass band droop in the decimated output signal using a recursive filter.
16. The method according to claim 15 wherein the input signal is decimated into the output signal with the set frequency in a single step.
17. The method according to claim 15 further comprising correcting a phase distortion in an output of the recursive filter.
18. The method according to claim 15 wherein compensating for the pass band droop comprises calculating a first pole position; and wherein the first pole position is calculated so that an absolute value of a slope of transfer characteristics of the pass band droop is maximum beyond a pass band edge of the pass band droop.
19. The method according to claim 18 further comprising determining whether a magnitude of the calculated pole position is unity.
20. The method according to claim 19 further comprising calculating a second pole position; and wherein the second pole position is calculated so that the magnitude of the first pole position becomes unity.
21. The method according to claim 18 further comprising choosing a first zero and a first pole for maintaining a set pass band ripple threshold.
22. The method according to claim 21 further comprising determining whether the pass band ripple threshold has been achieved.
23. The method according to claim 22 further comprising choosing a second zero and a second pole to achieve the pass band ripple threshold.
24. The method according to claim 23 further comprising adjusting a gain value of an all pass filter for providing an overall gain of unity.