[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] Re:Real Data only Transfer?
From: |
Eric Blossom |
Subject: |
Re: [Discuss-gnuradio] Re:Real Data only Transfer? |
Date: |
Wed, 7 Jun 2006 10:16:41 -0700 |
User-agent: |
Mutt/1.5.9i |
On Wed, Jun 07, 2006 at 07:04:00PM +0200, Kilian Timmler wrote:
> Ouch!
>
> This sounds like some of those previous threads.
>
> >Why do you want 16MS/sec real instead of 8MS/sec complex?
> >
> >Eric
>
> I didn't want to go there, the part I was worried about is to have no
> processing
> since the CIC filter has some unpleasant properties (not very flat, aliasing
> at the borders).
The final stage of the 8MS/sec path is run through a halfband filter
and thus is pretty flat. That is, the final decimation-by-two is
performed using a 31-tap halfband, not a CIC.
> From my point of view the PC is much more accessible than the
> FPGA. To get an understanding of the algorithms I don't have to try
> to read someones Verilog code, and to change any algorithms I don't
> have to write Verilog. On the other hand the PC should be quite
> capable in handling most signal processing tasks.
Yep.
> I'm generally interested in working with FPGAs but I don't have a
> logic analyzer, so that actual experiments might be cumbersome.
Simulation goes a long way. Or just "get it right" the first time ;)
See also icarus verilog.
> Anyway, I saw today that there is an fpga version with a halfband
> filter now, very nice.
Eric