LimeSDR USB test generated different output. Faulty limeSDR HW? Fixed

As per quick test procedure, I load self_test.ini and run test procedure, however my results look totally different. Does this indicate HW problem of the board? Any hints to troubleshooting would be really appreciated.

Hello,

I see that you said you have loaded the self_test.ini, did you also load the waveform file?

Regards,
Bevan

Hi, Bevan,

Yes, I’ve strictly followed test procedure - configured SXT, CLKGEN, enabled RF loopback, loaded the waveform.

In terminal window there are many libusb messages left after I close LimeSuiteGUI.

libusb: warning [darwin_abort_transfers] aborting all transactions on interface 0 pipe 2
libusb: warning [darwin_abort_transfers] aborting all transactions on interface 0 pipe 2
libusb: warning [darwin_abort_transfers] aborting all transactions on interface 0 pipe 2
libusb: warning [darwin_abort_transfers] aborting all transactions on interface 0 pipe 2
libusb: warning [darwin_abort_transfers] aborting all transactions on interface 0 pipe 2

Regards,

Dmitry

I got it working finally. Now it looks OK. The way I did it is quite simple. I work under Mac OS. Test sequence described in the Wiki looks like this:

  1. run LimeSuiteGUI,
  2. connect to the board,
  3. open self_test.ini,
  4. open SXT, press Calculate+Tune,
  5. open CLKGEN, press Calculate+Tune,
  6. open Modules -> Board Controls -> tick RF loopback ch.A -> RF loopback ch.B
  7. open Modules -> FPGA controls -> tick MIMO -> tick WCDMA (load waveform)
  8. open Modules -> FFT viewer -> Data Reading -> Packets MIMO, Graphs -> Display Channel -> A&B, START.

That would produce the output seen in the first post.

What fixed it was adding step 4a) open SXR, press Calculate+Tune.
Seems like without doing so limeSDR was out of tune on either Rx or Tx path. That what produced the odd result seen in the beginning of the thread.
Now I am very happy to know my SDR board works fine and I can gently move to learning how to run various applications on it and maybe even write SW programs. :smile:

I find that there is FFT signal separation between channels A+B if the GUI-> Chip button is not pressed (before or after Starting data reading). Pressing it multiple times sometimes causes separation of the channels and back to overlap. LimeSuite has problems.

As per wiki, gui->chip should be pressed once after opening *.ini file, e.g. self_test.ini, for the values from the file to take effect. However, it usually takes me several times to open the same file for all tabs like SXT, SXR, CLKGEN to get corresponding values.