On Mon, 2009-06-29 at 19:43 +1000, Nick Withers wrote:
G'day everyone,
I'm a bit worried... I no longer seem to be able to get anything but
zeroes on the quadrature channel out of gr-radar-mono (and my derived
wind profiler code).
I've attached an image of a plot of the in-phase samples received from
my wind profiler code... It may well be horribly misleading, given that
it's highly pre-alpha stuff, but it might also give someone a clue.
Certainly looks bizarre to me! Looks like a mix of the real and
quadrature samples I want / am expecting to see (~10 kHz wave; we're
sampling over 0.2 ms), along with their sum(?) and some zeroes thrown in
for good measure.
But yeah, this may not actually be as a result of the same problem, but
more my in-development work.
(Warning: This is pretty much all FPGA-side code)
Short version: I seem to be getting only zeroes for quadrature samples.
Could this be a software problem? Have I mucked up my hardware?
I'm using a USRP rev. 4 with a BasicRX. GNUradio 3.2, Ubuntu 9.04.
I get quadrature samples as expected if I run with debugging (-D
flag to usrp_radar_mono), which chucks through FPGA-generated data. If I
modify my wind profiler code to throw quadrature samples to both the
real and imaginary output channels, I get nothin' but zeroes on both
real and imaginary.
I'm examining data by using read_complex_binary
(gnuradio-core/src/utils/read_complex_binary.m) and plotting with
plot(real(samples)) / plot(imag(samples)) in GNU Octave.
It's taken me a while to notice this problem and a couple of things have
happened lately that may be related:
- I've physically moved my USRP / PC / signal generator; and
- I've updated over a hundred packages on the Ubuntu 9.04 PC to bring
it up to date.
Could this be a software issue? From looking at synaptic it did look
like I had registered GNUradio / USRP packages (libgnuradio, libusrp,
etc.), which I imagine I had installed yonks ago before I'd built
GNUradio from SVN. I've tried removing all those and rebuilding GNUradio
3.2 from scratch, but the problem persists. I can rebuild the box, but
if it's just gonna reinstall the same required libraries and so forth
that I may be having a problem with it may not give me very helpful. I
can rebuild the box with another OS but I'm trying to stay as boring and
mainstream as possible.
...Or maybe I've stuffed my USRP hardware? I'm working almost
exclusively in the FPGA...
If someone with an up-to-date Ubuntu 9.04 machine with GNUradio 3.2
could try to reproduce this I'd owe them a beer!
Any pointers appreciated, cheers all!
--
Nick Withers
email: n...@nickwithers.com
Web: http://www.nickwithers.com
Mobile: +61 414 397 446
attachment: usrp_wind_profiler real look.png
signature.asc
Description: This is a digitally signed message part
___
Discuss-gnuradio mailing list
Discuss-gnuradio@gnu.org
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
