[USRP-RIO][urgent] Calibration of Rx gain

[Nicola_TUD]

Hi,

we are using the following setup:

• A USRP-RIO is connected via PCI card to a desktop PC (transmitter)
• Another USRP-RIO is connected via PCI card to another desktop PC (receiver)
• A modified version of the stock streaming example is used to transmit and receive.
• The modifications are limited to waveform generation and display, i.e. 2 independent Tx channels and 2 independetnt Rx channels are employed
• The transmitter generates two signals, which are combined and attenuated (30dB). Then the combined signal is ditributed to Ch0 and Ch1 of the receiver.
• LVC version 1.0.0.50009

Initially, when working with the stock streaming example, there was a warning message that popped up when pressing the "STOP" button in the GUI (Tx).

Following the instructions, I run the "Update Device Correction Data.gvi". Ever since then, there is a strong mismatch of the gains in Ch0 and Ch1.

I did not pay attention if this was a problem __before__ running the "Update Device Correction Data.gvi". But I have the feeling that it might have actually messed up the channels.

This behavior is systematic. It was reproduced on two USRP-RIOs.

Any suggestion how to fix this?

Thanks,

Nicola

[Peter_W]

Hi Nicola,

From the screenshots you posted it looks like Ch0 has changed but Ch1 hasn't (tough to tell with the Y-axis blocked in the top image but it looks as though auto-scale is off). Are you sure the correction was applied to both channels? Were there any errors when running the Update Device Correction Data GVI? Can you check the "Coerced Configuration" tab on the front panel to see if the coerced levels are similar to each other and to the set Reference Level?

Best Regards,

Peter Williams

[Nicola_TUD]

Hi Peter,

Thanks for looking into this. The first screenshot is from before applying the update. As you stated it is not quite clear if the problem was occuring at that point in time already. I did not pay attention to the axis. The second screenshot is specifically from after the update.

Further, I do not know if both channels were corrected. The update VI has only one input, which is the RIO device identifier. There was no error running it.

The coerced configuration looks like this now:

If I set the reference levels to 10 dBm on each channel, then the coerced levels are approximately 3 dBm and 10 dBm.

Maybe there is a way to read the correction data directly from the device?

Nicola