It's all about the Software

I have a strange problem which is really bending my brain. I have an Adalm-Pluto RevB and LibreSDR Rev5 connected to the latest SDR-Radio release. Over time I upgraded both SDR's to v0.3.5 of the tezuka firmware. Both work well with SDR-Console and SDR-Television, no problems.

When the v0.3.11 firmware was released, I upgraded them but have issues all of a sudden. The Adalm-Pluto is transmitting about 10kHz lower than before and on the LibreSDR I do not get any transmit power, on TX1 and TX2. Tried various suggestions on the groups without success.

I downgraded back to v0.3.5 and all is working perfectly. When v0.3.12 was released, I decided to upgrade again. The problem is back again. On the Adalm-Pluto the transmit frequency is lower than before and on the LibreSDR I have no transmit signal on either TX1 or TX2. Tried some suggestions on the group again without success.

I downgraded back to v0.3.5 and all is back to working perfectly again. I am using the correct build file for the SDR device and firmware upgrades are performed as per the instructions from F5OEO.

Any suggestions on why this is happening will be appreciated.

73
Johan ZS6JPL

As summarized it occurs to me that if you generate different definitions for different firmware the definition might have an offset in one or more of them. I am betting this is not in the least likely; but, I've learned to listen when my brain coughs up silly hairballs. In the definition converter setting you might find the error if it is constant at all frequencies far enough apart a "ppm" type error would reveal itself it is in the Pluto. Tuning should not affect the conversion oscillator. Heat or "RF in the shack" might do something. Are the RF levels the same between firmware builds? Is the offset stable with time or does it drift? Is one firmware accepting external GPS and the other is not? 10 kHz at 10 GHz is only 1 ppm which could be from failure to GPS lock. But you're TX is about 2.5 GHz, so 10kHz is maybe still within the range of a TCXO with external trimming.

I'm just brainstorming here thinking of even unlikely things that could affect the shift. Stability and exact frequency change if it can be determined, could rule out some elements of the system. So far I like the up converter clock or the Pluto clock drifting for some reason if it is GPS locked.

{^_^}