LimeSDR Mini Dead on Arrival?

Projects LimeSDR
1

First time trying my new LimeSDR Mini (even though I got it back in late Feb).

Installed the Windows driver for the USB as tutorials suggested.
Ran the LimeSDR Quick Test app and got a fail on one of the RF Loopback tests.
Still within the 90day return window (just barely). Crowdsource told me to post here to confirm it was a device that needed to be exchanged (or to learn what the fix should be for it).
Here’s the log output from the test. Ran it multiple times, same result every time.

[ TESTING STARTED ]
->Start time: Wed May 6 00:25:49 2020
->Device: LimeSDR Mini, media=USB 3, module=FT601, serial=1D4C28EDDF0C2A,
index=0
Serial Number: 1D4C28EDDF0C2A
[ Clock Network Test ]
->REF clock test
Test results: 44987; 58184; 5845 - PASSED
->VCTCXO test
Results : 6710995 (min); 6711152 (max) - PASSED
->Clock Network Test PASSED
[ FPGA EEPROM Test ]
->Read EEPROM
->Read data: 13 02 0C 13 02 0C 02
->FPGA EEPROM Test PASSED
[ 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:(-56.8 dBFS, -1.35 MHz) - FAILED
->Run Tests (TX_1 -> LNA_H):
CH0 (SXR=2100.0MHz, SXT=2105.0MHz): Result:(-13.5 dBFS, 5.00 MHz) - PASSED
->RF Loopback Test FAILED
=> Board tests FAILED <=
Elapsed time: 3.86 seconds

Thanks in advance for the feedback!
I also tried the self-test.ini using the GUI app and got some other error.
I should add that I also tried to update the firmware. The firmware update went fine but same RF loopback test FAIL.

2

I had the same, see other topics about the quick test.
I had better results with an older firmware than the last one but even then i had sometimes quick test fail and always in the loopback.
But…when using it to transmit and receive on Oscar-100 satellite ( 2400MHz tx uplink and around 730MHz rx downlink) i never had problems except that i.m.o. the lime sdr mini runs a bit hot.
So i put a 5V fan on it, and now the temp rises to maximum 38 degrees C (was 56 degrees C and even more). I have done some more quick tests and now the tests are ALWAYS PASSED.
So…if you want to do the quicktest than the lime must be cold and no antennas connected.

3

Ok, I tried it while the device was cold and had no antennas plugged in. (Previously I had the two stubby antennas that came with it screwed onto the device). This time the QuickTest passed. I left it plugged into the computer and tried the same QuickTest again after 6 minutes and now it’s back to failing.

So, I’m new to this but generally if a device cannot reliably pass it’s own hardware test, I’d consider it a faulty device. If this is a heat issue I’d still argue it’s a faulty device as it doesn’t feel warm enough to warrant a failure. Especially since it’s not actually doing anything but running the quicktest.

If I unplug the LimeSDR Mini and let it cool down I’ll get a few passes until it gets a little bit warm then it starts to fail consistently. When I say warm, it’s just barely “warm” enough to tell it’s not “cold” this is nowhere close to being hot when it starts to fail.

[ TESTING STARTED ]
->Start time: Thu May 7 20:03:16 2020

->Device: LimeSDR Mini, media=USB 3, module=FT601, serial=1D4C28EDDF0C2A, index=0
Serial Number: 1D4C28EDDF0C2A

[ Clock Network Test ]
->REF clock test
Test results: 48907; 62104; 9765 - PASSED
->VCTCXO test
Results : 6710980 (min); 6711138 (max) - PASSED
->Clock Network Test PASSED

[ FPGA EEPROM Test ]
->Read EEPROM
->Read data: 13 02 0C 13 02 0C 02
->FPGA EEPROM Test PASSED

[ 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:(-12.4 dBFS, 5.00 MHz) - PASSED
->Run Tests (TX_1 -> LNA_H):
CH0 (SXR=2100.0MHz, SXT=2105.0MHz): Result:(-12.7 dBFS, 5.00 MHz) - PASSED
->RF Loopback Test PASSED

=> Board tests PASSED <=

Elapsed time: 5.49 seconds

[ TESTING STARTED ]
->Start time: Thu May 7 20:09:23 2020

->Device: LimeSDR Mini, media=USB 3, module=FT601, serial=1D4C28EDDF0C2A, index=0
Serial Number: 1D4C28EDDF0C2A

[ Clock Network Test ]
->REF clock test
Test results: 26862; 40059; 53256 - PASSED
->VCTCXO test
Results : 6710996 (min); 6711153 (max) - PASSED
->Clock Network Test PASSED

[ FPGA EEPROM Test ]
->Read EEPROM
->Read data: 13 02 0C 13 02 0C 02
->FPGA EEPROM Test PASSED

[ 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:(-57.1 dBFS, -2.46 MHz) - FAILED
->Run Tests (TX_1 -> LNA_H):
CH0 (SXR=2100.0MHz, SXT=2105.0MHz): Result:(-13.1 dBFS, 5.00 MHz) - PASSED
->RF Loopback Test FAILED

=> Board tests FAILED <=

Elapsed time: 6.22 seconds

– Additional Edit –

If I blast the device with a high power fan to keep the temperature down, it will pass the test several times more than without the fan before it starts to fail (and fails while the Aluminum case is cold to the touch because of the fan).

If I remove the board from the Aluminum case and blast the board directly with the high powered fan it seems to pass the QuickTest consistently. Got tired of trying to make it fail in this config.

Conclusion: This thing is hyper-sensitive to any temperature increase. The Aluminum case looks nice but actually holds in the heat and contributes to the failure problem.

4

Yes, i have the same conclusion. But remember that a self test with loopback is not the same as using the device on air. In the past i worked professionally with Ericsson microwave links who also had a software loopback test and i have seen many times a loopback test was perfect but errors when using it on air.
I think that the loopback test is indeed very sensitive for temperature rising.
The good thing is that i use the lime sdr mini from the beginning for 2400MHz tx and 730MHz rx and it never fails. Even without fan and with rising temperatures till almost 60 degrees the unit was still transmitting perfectly. So maybe we are too focussed on the self test.
Looks like there is a temperature bug in the software parameters for the selftest but luckily for us the device (in my case) is doing what he has to do.

5

I don’t have the version with the aluminium enclosure but i bought a 3D printed plastic enclosure on ebay and I glued the 5V fan on it, connected to the fan soldering points on the lime.

6

LimeQuickTest is configured with pass/fail parameters that are only valid when two conditions are met:

  1. There is nothing connected to the RF ports
  2. The board is cold
7

A few thoughts:

  • After HansKarel’s initial comment, I’ve done the tests without any antennas connected. I don’t recall seeing that guidance anywhere in the docs (why not put that note in the QuickTest app next to the start button).

  • “The board is cold” is a very subjective requirement. I let it sit till it’s room temperature. Plug it in and get a couple pass results. Within about about 1 minute it starts to fail. If I was in the tropics I wonder if it would work “cold”.

  • From an SDR newbie’s perspective, it looks like a bad piece of hardware (because the manufacture’s own test tells me that). What technical argument can you offer to offset my concern.

  • So the LimeSDR Mini reports its own temperature. If the QuickTest is temperature dependent, why not set a threshold over which the QuickTest will either not run that part of the test or report it (with a nice technical description of why) as something other than a failure. I confess, I’m a bit hung-up on the idea that a manufactures hardware diagnostic test should consistency pass under all operating conditions. Or at least be smart enough to flag a test that’s out of spec.

  • So how about some technical info on why this loopback test is so thermally sensitive? Component placement on board?

  • HansKarel offered some experience that puts me at ease but his experience is no guarantee my device will behave the same. Plus, I see a similar instance on the forum where the person was told to RMA it.
    (i.e. LimeSDR mini unreliable)

  • The comments in this post (New LimeSDR Mini - loopback test failed) seem consistent with what I’m seeing. If I aggressively hold the temperature down by blasting a fan on the bare board, the QuickTest appears to run as many times as you care to run it. However, any minor increase in temperature triggers a failure (and usually on the first of the two loopback tests). Is this an oscillator stability problem? Where do the oscillator(s) live in relationship to the components that generate heat?

  • I setup SDRConsole last night and was able to receive broadcast FM with the device. It runs a bit hot but seemed to work for RX. This gives me a bit more comfort that the device isn’t completely dead but still concerned I have a device with Thermal problems.

  • I won’t have time to try the TX functionality before the 90-day return window expires. This was a big ticket item for me, part of my angst is frustration that I waited so long before giving this a try and now I’m concerned if the TX functionality has issues I’ll get the “sorry too late” response when I try to resolve the problem.

1 Like