Spurios waves on 2.4GHz

[dj7th]

For radio operation on 2.4GHz via the QO-100 satellite I use the Adalm-Pluto Rev-B and Rev-C (brand new) + PA and on the narrow band SDR Console 3.x for control. Some of my radio friends have found through measurements with high-quality spectrum analyzers that the Adalm-Pluto at 2.4GHz generates impermissibly high spurious waves, which are up to 40dB below the TX signal, but max. -60dB are permissible.

I checked the output signals with two other Adalm-Plutos + Satsagen Spectrum Analyzer program and found that these spurious waves are in the +/- 1MHz range around 2.4GHz, no matter which SDR Console version I use (3.0.26 - 3.1 Jan 17).
As a test, I operated the Adalm-Plutos with SDRangel and found that these spurious waves do not occur with it.
Even Satsagen as Signal Generator and Plutos DATV operation in stream mode with F5OEO firmware generates a clean signal without spurious waves. (see attachments)
Question: is there any information about this effects?

[dj7th]

Addendum:
OM Rüdiger DK6JL also examined the spectrum of the Adalm-Pluto with a professional spectrum analyzer on various frequencies and allowed me to publish his results below. He also used the January 17 beta version for this, as did I.

[Max]

I checked the output signals with two other Adalm-Plutos + Satsagen Spectrum Analyzer program and found that these spurious waves are in the +/- 1MHz range around 2.4GHz, no matter which SDR Console version I use (3.0.26 - 3.1 Jan 17).

I believe this problem is fixed in the latest Beta (Build 2496, January 26th).

Download here:
https://www.sdr-radio.com/download#Beta

Release notes here:
viewtopic.php?f=66&t=1474

73

Max

[dj7th]

Of course, we immediately installed this new Beta version (Jan. 17) and hoped that it would fix the problem. However, our measurements have shown that there is practically no difference between versions 3.0.26 to the current one as far as the spurious waves are concerned.
It is puzzling why SDRangel and the RF-Generator from Satsagen, as well as the DATV firmware from F5OEO deliver a clean signal, but SDR Console generates spurious waves that are far above the permissible limit.

[Max]

Of course, we immediately installed this new Beta version (Jan. 17) and hoped that it would fix the problem. However, our measurements have shown that there is practically no difference between versions 3.0.26 to the current one as far as the spurious waves are concerned.

Rolf, are you sure you are using latest Beta? It's Jan 26th, not Jan 17th.

I don't own Pluto so cannot confirm. Hopefully Sigi will see this. Sigi?

Did you see this thread on the IO Group chat?

https://sdr-radio.groups.io/g/main/topic/88479723

73

Max

[dg9bfc_sigi]

Simon worked a bit on the code (tx filter) in one of the last versions ... (so use last beta to measure spurii)
yes .. the pluto has some spurious emissions (mixing products etc) ... with FULL POWER they are around 40 to 44 db surpressed (i did not measure again after simon made those changes in tx filter so now they may be even lower) ... but ... if you lower the power the distance from main signal to spurious signal becomes bigger and bigger ... so to have a cleaner output i highly recommend to turn the drive down (below - 20 dbm at best) and add a gain stage of 20 db behind the pluto to have again the around 0 dbm output (but with a lot cleaner signal)

the chips differ a bit (production spreading) and also on every boot the pluto does a new iq calibration ... and so you may have different numbers after each start

but ... if you keep your power low then the signal is clean enough to be used

the "mixing products" (spurious emissions) are from a not perfect iq calibration on start ... a good warming up and then restart the radio ... plus a stable reference (example a good tcxo) and the results are much better

imagine how the calculation results will come out at the end when your reference drifts a bit DURING the iq calibration!!! ... if there is just a tiny "wobble" on the reference then the iq calibration will not work good (not to mention the "rough" modulation when you tx with such an unstable reference)

hope that helps a bit
dg9bfc sigi
ps see picture how the spurious go down with lower power setting ... and yes i used satsagen to test that and used an external well working spectrum analyzer

[Simon G4ELI]

FWIW the problem was the optional extra gain in the FIR filter. Disabling this greatly reduced the spurious transmission.

[dj7th]

After today's installation of the latest 3.1 beta, I connected two brandnew Plutos Rev C via cable + 10dB attenuator and measured the output signals with Satsagen at 2,400MHz, once with 100% output (Drive=100) and once with reduced output (Drive=50 ). As can be seen, the spurious level changes only minimally in relation to the signal on 2,400MHz.
Rüdiger DK6JL came to the same result in his measurements with a high-quality spectrum analyzer, regardless of the frequencies and output level of the measured Pluto.

Since very few users of SDR Console are aware of this problem and can therefore unknowingly generate impermissibly high spurious waves outside the band limits when transmitting, I believe that a remedy should be found as soon as possible.

Be that as it may, the question is still: why are there no or hardly any visible spurious waves when using SDRangel or Satsagen as an RF-Generator and what is different in their software?

[MikeC]

I usually use a LimeSDR but I do also have an unmodified rev B Pluto that I intend to use with Satsagen as a spectrum analyser. I thought I would see if the effect described by DJ7TH could be reproduced. I have to say I get exactly the same spurs as he does, as measured on an external Anritsu spectrum analyser. They are at about -40dBc and change linearly with the output level even when the output is down at -20dBm. The version of Console is from 26th Jan and I checked that the boost is turned off. There is no sign of them when the Pluto is run from Satsagen or SDRAngel.

Now, I have limited experience of the Pluto and I may have some Console setting wrong so all I am saying is that in my setup the effect described by DJ7TH can be reproduced. I do not seem to get the effect that Sigi described where the higher the output level the worse it gets. I am not doubting his results, or anyone else's, but here, and presumably at DJ7TH's, these spurs seem to exist.

Mike Christieson

[ted211]

These spectral lines occur only in "Tune" or "Tone", not when transmitting.

Attached is a spectrum using AM

Detlev