Problem with simple test of USRP2+WBX

Hello:

I’m stuck, and I need some help.

I’m running GNU Radio 3.3.0 on openSUSE 11.2. I have two USRP2 boards,
each with a WBX daughterboard, connected to two Intel PRO/1000 GT NICs,
on eth1 and eth2. I’m using the “txrx_wbx_raw_eth_20100608.bin”
firmware, not the from-the-factory firmware. I made a simple flowgraph
which connects an FFT window to a USRP2 Source, tuned to 900 MHz at a
sample rate of 3.125 MHz (decimation factor 32). I input a 901 MHz sine
from an HP signal generator into the TX/RX connector on the WBX
daughterboard. When I run the flowgraph with the USRP2 Source set to
eth1 (for the first USRP2), it runs fine, and I see a nice clear tall
spike in the FFT at 1 MHz. However, when I set the USRP2 Source to eth2
(for the second USRP2) and re-run the flowgraph, I get a spike at 901
MHz that is barely noticeable and greatly attenuated. What’s going wrong
here? I used that second USRP2 a few weeks ago with a different
daughterboard and it ran fine, so I don’t think that USRP2 is bad.
Is there something wrong with the WBX daughterboard? Am I doing
something wrong in GRC? Any ideas?

Any help would be GREATLY appreciated.
Thanks a lot!!

Steve McMahon

On 11/19/2010 03:13 AM, Steve M. wrote:

Steve McMahon

What power level did you use for the signal generator?

The GaAs MMIC LNAs on these boards don’t like high power input, and the
gate insulator
will often go “phhhttttttt!” if you input more than -15dBm for any
length of time.


Principal Investigator
Shirleys Bay Radio Astronomy Consortium