USRP tuning

I think I have a misunderstanding regarding tuning the USRP.

I have connected the TX-A directly to RX-A. I am plotting the FFT of
both the transmitted and received signal. I have tuned RX to 0 Hertz
so I expect the positive frequencies to look the same up to noise
affects.

What I get instead is mostly noise, with occasional blips at an offset
frequency. I.E. 5 KHz input gives me occasional energy at about 3
KHz. But most often I don’t see any signal at all.

I have posted the GRC graph and FFT plots at
http://wyoinnovation.blogspot.com/.

Thanks in advance for any help you can offer.

Justin S.

On 9/7/07, Justin S. [email protected] wrote:

I think I have a misunderstanding regarding tuning the USRP.

I have connected the TX-A directly to RX-A. I am plotting the FFT of
both the transmitted and received signal. I have tuned RX to 0 Hertz
so I expect the positive frequencies to look the same up to noise
affects.

I am assuming this means you used a cable to connect two different
cards together? If these have any transmitting power to them, there
should be some attenuators in line to make sure you don’t blow
anything up. This is just an FYI.

What I get instead is mostly noise, with occasional blips at an offset
frequency. I.E. 5 KHz input gives me occasional energy at about 3
KHz. But most often I don’t see any signal at all.

I have posted the GRC graph and FFT plots at
http://wyoinnovation.blogspot.com/.

So I’ve been looking at your GRC graph and I have some fundamental
USRP questions as well as some questions with regards to your setup.

First, your signal source has a frequency of 0 and an amplitude of 16,
so you just want to transmit carrier? I am not sure if there is a
constant block but using that to make a DC offset can accomplish the
same thing.

Second, your sampling rate appears to be 50ksps and your USRP
interpolation rate is 200. This gives a final sampling rate of
10Msps. Does the USRP final sampling rate have to be 64Msps, or can
the USRP drive the DAC at 10Msps?

Next, your USRP source has a decimation rate of 1. I don’t think this
is a legal decimation rate. You should probably decimate your signal
down a bit. A decimation rate of 128 would give a sampling rate of
500ksps for a signal sampled at much less than that which is really
just a DC offset.

Lastly, there seem to be some sliders in the GRC window which are
controlling the aforementioned values. In those screen shots, I see a
tuning frequency of 0 and a signal frequency of 5kHz.

I am probably just getting confused. On the other hand, getting
answers about the interpolation and decimation rates of the USRP would
be good.

Sorry if I haven’t been much help.

Brian

Sounds like you are on the right track. I’ll add those details
tomorrow.

Thanks,
Justin

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs