Reference Level fo NI-USRP 2954R Simple Rx Streaming

[AlePiccio]

Hi everybody,

I'm working with the USRP 2954R (connected to a pc with the PCIe x4 cable), and i'm executing, as an example to better understand the environment, the VI "Rx Streaming (Host)", from the project "NI-USRP Simple Streaming".

The VI works fine, but i've some doubts about the Reference Level RL and the Coerced RL; in particular, i've noticed that for value of RL>-5dBm (indicatively), the Coerced still remain at the value of -5dBm, and for values of RL less than -5dBm, the Coerced decrease at steps of 0.5dBm, until RL>-35dBm; when RL<-35dBm, the coerced is blocked at -35dBm. I've investigate the project and the VIs deeper, and i've found that the RL coincides with the Gain:

In fact, the VI "Calculate Level Range" gives me the values (indicatively) -5 (MAX) and -35 (MIN).

The problem is that if i receive a pilot signal (whose amplitude was measured with a VNA), the RL does not behave like a gain: the only diffferences that i've noticed is that with the decreasing of the RL (and so the Coerced), the noise background decreases, and for a Coerced RL of -5, the maximum, the pilot signal is plotted with the expected amplitude.

So my questions are:

1. What actually is the Reference Level RL?
2. What is its relationship with the RX Gain?
3. How could i manually handle the gain for the received signal?

Thanks a lot !

Alex

[alamanna]

It doesn't make much sense to talk about reference levels, as the USRP is not a calibrated instrument, and therefore the only relevant parameter is the gain. In the "Level" case of the case structure, depending on the gain level set in the front panel, the optimal level between -5 dB and -35 dB is calculated, thus avoiding any faults in the electronic front end of the USRP.

In case you want to set the gain manually (but risking to break the USRP for too high values), just change the enum of the case structure, and put yourself in the "Gain" state which bypasses the calculation of the optimal gain.

Regards,

Alessia

[AlePiccio]

Thank you for your reply.

"It doesn't make much sense to talk about reference levels, as the USRP is not a calibrated instrument, and therefore the only relevant parameter is the gain. In the "Level" case of the case structure, depending on the gain level set in the front panel, the optimal level between -5 dB and -35 dB is calculated..."

So, i set a gain level in the front panel, and the system computes the optimal level, which will be used; the point is: what is basically the optimal level? because it isn't a gain (as mentioned before, with the VNA i didn't see any differences in the pilot varying the gain level, but only in the background noise), furthermore its dynamic is different: the gain level dynamic is 30dBm, while the receiver gain dynamic, as in the 2954R specifications, is 37.5 dB (both with steps of 0.5dB).

Thank you for your help.

[Mlandi]

Basically the USRP allows you only to set a gain, you should know in advance which is your signal level so that applying the proper gain you bring the input signal at the optimal level to be sampled by the ADC without any compression (too much gain) or without pushing the signal close to the USRP noise floor (too low Gain)

"the only diffferences that i've noticed is that with the decreasing of the RL (and so the Coerced), the noise background decreases, and for a Coerced RL of -5, the maximum, the pilot signal is plotted with the expected amplitude."

If I understand properly your sentence In Paragraph 2 of the following link you can find the reason why the only effect that you see is on the background noise level:

http://www.ni.com/product-documentation/53640/en/

Not sure this answer to your question, if not please provide us more details on the input signal level, measured levels and configured gain setting.

br