Re: a running problem with the USRP, i am confused!


#1

Hi, Jason

Thank you for your help! Good man! Actually, after running lots of
experiments, I found the reason for my problem. One of my USRP is broken
and it can only serve as a TX, when playing as RX, that weird thing
would happen.

Actually, I am not transmitting one large batch file continuously. I was
transmitting 1000 packets 8 times, which means I started the
benchmark_rx.py 8 times. E.g., the PDR was 90% 1st time, 80% 2nd time,
60% 3rd time…

Anyway, thank you! I have fixed it now!

Bill

— On Fri, 5/8/09, Jason U. removed_email_address@domain.invalid wrote:

From: Jason U. removed_email_address@domain.invalid
Subject: Re: [Discuss-gnuradio] a running problem with the USRP, i am
confused!
To: “Bill S.” removed_email_address@domain.invalid,
removed_email_address@domain.invalid
Date: Friday, May 8, 2009, 4:35 PM

Does it mean USRP could not keep on running for
a long time or the USRP boards we are using have been broken? I do not know
the reason for this problem. Please let me know what’s wrong with the USRP.
Thanks a lot!

I believe there is an issue with the costas loop implementation in one
of the older mpsk_receiver implementations where the phase offset
correction gets clipped rather than rolled over (ie, -2pi->2pi). If
the offset is small it may take awhile for this error to show up. If
you use the verbose costas loop and see the phase correction going up
to 2pi or down to -2pi and not rolling over you may have this issue.

Jason