Loop-back on the same USRP

Hi all !

I have only one USRP2 and would like to use a loop-back cable to connect
the output to the input of the same USRP. So I would, for example, run
benchmark_tx and benchmark_rx in two different shells. Can I do it
without any worry ? I am asking the question because it seems that many
people always ran a such test with two USRP (one for Tx and the second
for Rx).

Thanks,

Chris

On Thu, Mar 4, 2010 at 12:37, [email protected] wrote:

I have only one USRP2 and would like to use a loop-back cable to connect the output to the input of the same USRP. So I would, for example, run benchmark_tx and benchmark_rx in two different shells. Can I do it without any worry ? I am asking the question because it seems that many people always ran a such test with two USRP (one for Tx and the second for Rx).

No. In general, USRP applications assume they “own” the USRP they are
communicating with, so the second script you run will stomp on the
first one.

There is no problem, however, using a single machine and having two
USRP1 devices on the USB. In this case, scripts can distinguish which
USRP they are talking to using a parameter during initialization. Not
all example scripts are written with this in mind; I’d have to check
if the digital example you reference does so (I think so.)

A single USRP, however, could be used with a loopback between TX and
RX if you were to write your own script that pulled in the
“transmit_path” and “receive_path” hierarchical blocks into a single
application, similar to what tunnel.py does.

As an aside, I think the recommendation is to use 30 dB of attenuation
between the transmitter and receiver boards to avoid damaging the
front end of the receiver. You may be able to get away with less
attenuation if you know what you are doing, and properly control the
transmit power.

Johnathan