[USRP-users] E100 Sequence Errors

McKeever, Kenneth R. Kenneth.McKeever at jhuapl.edu
Tue Feb 19 15:51:57 EST 2013


I am currently working with some E100s which appear to be exhibiting
sequence errors whenever I try to transmit ("S" is flooded on the
output).  This seems to happen only when I run a flowgraph with a
transmitting component, whether it is code that I have written or is
example code, such as benchmark_tx.py.  I see the same issue on both E100s
that I am working with, which have the same hardware and software

This was a problem when I first started development using the factory SD
card images (e1xx-002).  Regardless of the code being used, the flowgraph
would run for a while with good RF output from the USRP, then after some
random amount of time (usually minutes), the RF output would halt and "S"
characters would flood the screen.

I have since updated to the latest image (e1xx-003) and have updated all
of the software using opkg. Not only does the problem persist, but it
almost immediately (less than 0.5 second) produces errors upon running any
transmitting code.  The time it takes to cut out is related to the sample
rate -- lower sample rates make it cut out quickly (~15 ms) whereas larger
sample rates take a bit longer to cut out (100's of ms).

I am using E100s with WBX RF daughtercards and GPSDOs running up-to-date
software via opkg (GNURadio v.3.6.2 with UHD v.003.005.000).

I came across a discussion on a USENET post from last May talking about a
similar problem, but I'm not sure if a solution reached the official
GNURadio or UHD builds.

If anyone has run into this before or has more information about a
solution or work around, I'd really appreciate any pointers.


More information about the USRP-users mailing list