LimeSDR mini : RF loopback test FAILED!

So i downloaded and ran the latest LimeQuickTest executable to test the device and got the following log:

->Start time: Fri Sep 17 16:36:32 2021
->LimeSuite version: 20.10.0-g1480bfea

->Device: LimeSDR Mini, media=USB 2, module=FT601, serial=1D53AFDB8144F9, index=0, HW=2, GW=1.28
Warning: USB3 not available
Serial Number: 1D53AFDB8144F9
Chip temperature: 36 C

[ Clock Network Test ]
->REF clock test
Test results: 43026; 56201; 3884 - PASSED
->VCTCXO test
Results : 6711030 (min); 6711187 (max) - PASSED
->Clock Network Test PASSED

->Read data: 13 07 0F 13 07 0F 02

[ LMS7002M Test ]
->Perform Registers Test
->External Reset line test
Reg 0x20: Write value 0xFFFD, Read value 0xFFFD
Reg 0x20: value after reset 0x0FFFF
->LMS7002M Test PASSED

[ RF Loopback Test ]
->Configure LMS
->Run Tests (TX_2 → LNA_W):
CH0 (SXR=1000.0MHz, SXT=1005.0MHz): Result:(-20.9 dBFS, 5.00 MHz) - FAILED
->Run Tests (TX_1 → LNA_H):
CH0 (SXR=2100.0MHz, SXT=2105.0MHz): Result:(-21.0 dBFS, 5.00 MHz) - FAILED
->RF Loopback Test FAILED

=> Board tests FAILED <=
Elapsed time: 5.87 seconds

Also, There is nothing connected the SMA connectors. (Do I need to add an antenna here?)

Also I am pretty much new to SDR, my previous background has been in developing FPGA systems, so can anybody also please guide me in how to get started with this device? (My aim is to develop signal processing systems and deploy them on limeSDR)

The above made no sense to me I always assumed that -15 +/- 6dBFS was a PASS
Turns out I’m wrong, a different test is used for different hardware, which makes sense, and for the LimeSDR-mini (and the LimeNET-Micro), a PASS is defined by the range -14 +/- 6BFS (-8dBFS to -20dBFS).

So to me it looks like your LimeSDR-Mini about 1 dB outside of the design specification.

  1. Is this including that fact that there were no antennae connected?
  2. Is this normal ?
  3. Will I have to do an RMA then?

Srry for the rapid fire, I’m very new so I have no idea…

My guess would be that the LimeQuickTest application is how the factory validate the assembled boards work. So after the boards are manufactured, they would be sent to testing, and the boards would be plugging in (with no antenna) at room temperature and LimeQuickTest ran and passed or failed would be decided in a handful of seconds, and then the next board would be tested, rinse and repeat.

My guess would be if you repeated the exact same test the way that it would have been done in the factory plugged your board in at room temperature and ran the test within a few seconds it would pass. Although 36 C is not particularly hot (~body temperature), it is probably enough to move your board from inside the specified specification to just outside it.

To reiterate what I’ve read here many times from people who work for Lime Microsystems posted on this forum the board must be cold (initially at room temperature) with no antenna or other hardware connected for a valid test. If the board has been plugged in for a few minutes then it has had time to warm up and the test is no longer a valid test.

So to answer your questions

  1. If there was anything other than a USB cable connected to the board the test would be invalid.
  2. At a higher temperature being outside of specification, which is as far as I am aware only valid at room temperature, would be normal.
  3. The very first thing that will be asked is to repeat the test with a cold board at room temperature (which you should do and post, I would - just for peace of mind), but I suspect that your board will pass with a valid test.

My understanding is that LimeQuickTest is to rapidly identify boards with faults (in the factory), and if they plugged in a board for minutes before each test not as many boards could be tested every day, or it would require more test stations and staff which would increase costs which would raise the price (and it would also requires a different test specification than for a cold board). So currently the validity of a test is within a very narrow range of parameters - nothing connected, cold board, rapid testing.

1 Like

We should probably have made clear that such results are anything but a hard failure and even if the odd dB out of spec when cold, it’s unlikely to have any discernible effect on the application. 10 or 20dB of course is a different matter.