Forum: GNU Radio Problems with Mueller & Mueller clock recovery; not very rel

Announcement (2017-05-07): www.ruby-forum.com is now read-only since I unfortunately do not have the time to support and maintain the forum any more. Please see rubyonrails.org/community and ruby-lang.org/en/community for other Rails- und Ruby-related community platforms.
1b785981ac4fcfb087d2a074ab20c81f?d=identicon&s=25 Christian Sokolowski (Guest)
on 2007-07-04 14:20
(Received via mailing list)
Hi,

I have some trouble using the function gr_clock_recovery_mm_cc.cc
(Mueller &
Mueller Clock recovery). Using DPSK with
datarate 40 kbps, the communication system works perfect running a
simulation
without USRP board. The problem occurs when I switch to the board: when
calling/initialising the receiver client of the system, in approximately
70%
of the cases the following error message shows up.

scipio:/home/cso/da # ./mdpsk_rx_rrc_client_thread.py
>>> gr_fir_ccf: using SSE
Using RX d'board A: Flex 400 Rx MIMO B
python: gri_mmse_fir_interpolator_cc.cc:66: gr_complex
gri_mmse_fir_interpolator_cc::interpolate(const gr_complex*, float):
Assertion `imu >= 0' failed.
Aborted

Once the clock is synchronised (the other 30% cases), the receiver works
fine. The problem is 'temporarely' solved by unplugging the DC power
supply
of the board. Is it possible that some buffers on the hardware remain
filled
from former runs on the board?

I am calling the M&M clock recovery with the following parameters:
clock_sync = gr.clock_recovery_mm_cc(self.sw_decim, 0.25*0.009, 0.5,
0.03,
0.002)


parts of gri_mmse_fir_interpolator_cc.cc, which is called by
gr_clock_recovery_mm_cc.cc:

gr_complex
gri_mmse_fir_interpolator_cc::interpolate (const gr_complex input[],
float mu)
{
  int   imu = (int) rint (mu * NSTEPS);

  assert (imu >= 0);
  assert (imu <= NSTEPS);

  gr_complex r = filters[imu]->filter (input);
  return r;
}

I expect that "assert(imu >=0);" is not very robust for synchronising
the
signal for some unknown reason.

thanks for any help.

Chris

---------------------------------------------------------------------
Christian Sokolowski     Office: +49(0)-240-7575-7043
RWTH Aachen              FAX:    +49(0)-240-7575-7050
Wireless Networks Dept
Kackertstrasse 9         E-mail: cso@mobnets.rwth-aachen.de
D-52072 Aachen           Web: http://www.mobnets.rwth-aachen.de
---------------------------------------------------------------------
3596cfe1d579c65b9babd35e8787977c?d=identicon&s=25 Matt Ettus (Guest)
on 2007-07-05 19:13
(Received via mailing list)
Christian Sokolowski wrote:
>
> I expect that "assert(imu >=0);" is not very robust for synchronising the
> signal for some unknown reason.
>

Have you tried commenting that out and seeing what happens?

Matt
This topic is locked and can not be replied to.