Re: [Discuss-gnuradio] Suse10.1 walkthrough and a question concerning crosstalk

[Mark Sullivan]

Matt,

Coupling problem solved. Doh! -45 dBm is *way* too low; only the very
low order A/D bits were toggling. When I put in something closer to +10
dBm I get a nice strong signal and the image on the other port of the
same daughtercard is down about 60 dB. I can't measure the crosstalk
from one daughtercard to the other. I guess that the tiny amount of
crosstalk at -45 dBm was just enough to flip the LSB of the unconnected
port often enough to give me a spectral line at a very low level. With
more amplitude the A/Ds are acting more "linear" and I can now see the
difference in levels more accurately. The upside is that I know a lot
more about the code now than when I started.

Best regards,

Mark Sullivan

On 9/24/06, Matt Ettus <address@hidden> wrote:

Mark Sullivan wrote:

> When I inject a signal at a -45 dBm level on the "RX-B" sma connector
> on a Basic RX daughtercard, I see a copy of the signal on the "RX-A"
> side when I use usrp_fft.py. The level is 10-20 dB below that of what
> I see if I move the signal from "RX-B" to "RX-A," but this seems like
> too much coupling. Using a 50 ohm termination on the unconnected sma
> connector helps a little bit. I see no copy of the signal from either
> input of the other Basic RX daughtercard. Any thoghts, anyone?
>

I don't know what's happening here, but I see about -80dB coupling
between the channels.

>
> Here is a walkthough I put together for Suse10.1. I found a few
> "gotchas" not described in the build instructions on the Wiki.

Thanks!  I put this on the wiki, but didn't "wikify" the formatting.
Feel free to edit it.  It is linked to from the Suse install page.

Matt