Re: [Discuss-gnuradio] Variation in Tx power across USRPs

[Marcus Müller]

Hi Pavan,

thanks for sharing these observations! To qualifiedly answer, though, we'd need to know a couple more things:

Which USRPs and, if applicable, which daughterboards are we talking about, which frequencies and which filter bandwidths, and maybe also which kind of signal?

Generally, a little inter-device difference is expected, but 5dB would be, according to my gut feeling, above what I'd expect. Generally, none of the USRP devices are calibrated measurement/reference systems, so absolute power follows no strict scale, and temperature, and different component values might contribute. I'd also assume that you didn't use a torque-controlled wrench to screw on the cable... Really, as a rule of thumb, the higher the frequency, the larger inter-device differences might affect power, but the B2xx devices, for example, aren't in their "sweetest" spot at the very low end of their frequency range, so variations might be severe there, too.

As this isn't really a GNU Radio question, I cross-refered the usrp-users list. I think that might be a better forum for discussion on this!

Best regards,

Marcus

On 13.07.2016 00:15, Pavan Yedavalli wrote:

Hi,

I was doing a quick test to see whether all of the USRPs are transmitting at the same level, so I did a single setup where I connected each USRP's Tx/Rx output to a spectrum analyzer (with attenuations). After sequentially doing it, I am getting up to 5 dB of variation across different USRPs. Is this expected?

For instance:

USRP1: -16.x dB

USRP2: -19.x dB

USRP3: -14.x dB

USRP4: -16.x dB

Note that I'm disconnecting the same cable and reconnecting to each USRP, so that is constant, as well as the exact same code (with just changing the address of each USRP). Should I double check some connections or is this variation expected? It seems too high for it to be expected. Please keep me posted. Thanks.

--

Pavan

_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio