[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] Frequency Xlating FIR filter: bug in gr_rotator
From: |
Matt Ettus |
Subject: |
Re: [Discuss-gnuradio] Frequency Xlating FIR filter: bug in gr_rotator |
Date: |
Thu, 10 Jul 2003 11:44:43 -0700 |
User-agent: |
Internet Messaging Program (IMP) 4.0-cvs |
Quoting Sérgio Rui Silva <address@hidden>:
> I have been using the "GrFreqXlatingFIRfilterSCF" and I notice
> that its output slowly increases in amplitude (or diminishes,
> depending on the frequency setting). I found that this is due to
> the "gr_rotator" class. If the absolute value of the
> "gr_rotator"s "phase"
> variable is printed during the processing cycle one can see that
> its value starts with 1.0, as it should, and then augments
> slowly. I think this is because the recursive multiplication
> used causes a precision error to increase in one direction.
Yes, that's exactly what is happening.
> I solved this by
> not using the gr_rotator class but the gr_nco class instead.
> This solution is not as elegant as the one with the gr_rotator
> class, but it does the trick by now... Any ideas one how to
> solve this correclty?
That's exactly how to solve the problem. I thought I had alread made that
change, but apparently, I had only done that in the FPLL class. If you send me
a patch, I'll integrate it.
Thanks
Matt