discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Receiving ATSC using USRP2


From: Chuck Swiger
Subject: Re: [Discuss-gnuradio] Receiving ATSC using USRP2
Date: Sun, 28 Mar 2010 11:55:39 -0400

On Sat, 2010-03-27 at 13:36 -0700, Jared Casper wrote:
> Hi all,
> 
> I've got a USRP2 with a WBX attempting to receive ATSC signals.  I'm
> using the latest (git HEAD on 27 March 2010) version of everything
> (firmware and software).  usrp2_fft.py shows a strong signal for a
> local tv transmitter (680 - 686 MHz).
> 
> It looks like the ATSC code is written for the original USRP, using a
> decimation of 10 to get 6.4 Msps, interpolated by three to get 19.2
> Msps.  Since I have a USRP2, I can't get exactly 6.4 Msps.  Has anyone
> gotten this to work using a USRP2?
> 
> I've tried using 5, 10, and 20 Msps, interpolating up to 20 Msps (I've
> also tried various gains).  In addition to changing the interpolation
> rate in interp_short.py, I've gone through the code and changed 19.2e6
> to 20e6 in interp_short.py, fpll.py, xlate.py, atsc_fpll.cc, and
> atsc_bit_timing_loop.cc.  However, nothing is output from ./btl-fsd.py
> (a symptom seen a few times in this list with no obvious solutions).
> None of the programs output any error messages, just that they are
> using SSE, and fpll.py reports setting initial_freq to 3065000.  I'vek
> sanity checked the logged data in matlab and verified a signal exists
> from -3 MHz to +3 MHz.
> 

Lets see, can you guys capture at 20 Msps now?  Tuning to the center of
the channel you have signal from -4 to +4Mhz, then xlate.py shifts the
center up to 5.75Mhz. At 20Msps the top is just under 10Mhz (9.75Mhz). I
liked to just dump the output of each stage, look at it with a spectrum
display and make sure it's what you expect. After xlate.py you should
have the signal centered at 5.75Mhz. 

Then wire xlate.py to fpll.py and check it's output - that should have
the pilot carrier at 0 with signal up to the low pass cutoff at 5.75Mhz.

Maybe if someone makes a few seconds of raw data available I could play
around and compare what works with what doesn't.

BTW, I *just* moved back to an area with good over-the-air signals.

--Chuck






reply via email to

[Prev in Thread] Current Thread [Next in Thread]