WSPR harmonics #txaudio


Steve marsh (M0NMA)
 

If receiving a very strong local wspr signal it often decodes as a main signal plus several harmonics at +/-100hz on 50MHz and 70MHz. On 2m they seem to be +/-50MHz. 


does anyone know what is causing these? Are they artefacts of the transmitted signal, or are they created in the sound card? Also, why is it exactly 100hz?

just curious.

Steve M0NMA


Alan G4ZFQ
 

does anyone know what is causing these? Are they artefacts of the transmitted signal, or are they created in the sound card? Also, why is it exactly 100hz?
Steve,

It is a common phenomena, probably more likely to be seen on the quieter VHF bands.
A few times I've noticed it on my own signals. It seemed to be greatly reduced by attending to grounding. Maybe by simply wiggling a nasty 3.5mm plug.
Therefore I attribute it to bad grounding/ground loops from 50Hz main power, bridge rectified? It is sometimes reported elsewhere, 120Hz in the States.

Sometimes it seems to be overmodulation, sometimes produced by a receiver on strong signals.

73 Alan G4ZFQ


Reino Talarmo
 

>If receiving a very strong local wspr signal it often decodes as a main signal plus several harmonics at +/-100hz on 50MHz and 70MHz. On 2m they seem to be +/-50MHz. 


>does anyone know what is causing these? Are they artefacts of the transmitted signal, or are they created in the sound card? Also, why is it exactly 100hz?

>just curious.

>Steve M0NMA

Hi Steve,
Thank you for a good question! Well, the answer is that you saw an European station, if it were an American one you would have seen 120 Hz. The reason is that the full wave rectification of DC voltage has generated a pulse or triangular100 Hz AC voltage and that modulates Tx frequency for some reason. One known reason is usage of a power amplifier run in C-class or D-class or any class where amplitude of the output signal depends  on the power supply DC voltage. So there is an additional amplitude modulation. There are also other mechanisms, but I assume that AM modulation is normal reason.
When those sidebands are at 50 Hz, then mains voltage enters at some point to the audio path and that path is unlinear e.g. due to modulator or audio amplifier overdriving.
By the way earlier wsjt-x versions quite often decoded also those sideband signals resulting multiple decodes of the same signal at 100 Hz steps. Current wsjt-x version seem to reject additional decodes and it more difficult to “measure” strength of the sidebands; sometimes I spotted only -20 dB down.

73, Reino OH3mA


Steve marsh (M0NMA)
 

Thanks Alan and Reino. 


I am helping another station local to me test his beacons on 10,6,4,2m and 70cm. They are all homebrew beacons and run off battery as the intention is to locate the remotely once testing has finished. 


in this case, the mains induced harmonics must be being created at my end. I’ll try decoding using a laptop and battery operated rig to see if the harmonics disappear. 


Steve M0NMA


Reino Talarmo
 

Steve,

Interesting case. Could you report results of the battery test, please?

There could be some less obvious transfer of main related interference from e.g. lab power supply to the homebrew transmitter even, when TX is battery powered, hi!

73, Reino OH3mA

 

>I am helping another station local to me test his beacons on 10,6,4,2m and 70cm. They are all homebrew beacons and run off battery as the intention is to locate the remotely once testing has finished. 

>in this case, the mains induced harmonics must be being created at my end.
I’ll try decoding using a laptop and battery operated rig to see if the harmonics disappear. 


Steve marsh (M0NMA)
 

Reino

i will do. My friend is also going to look at the output on a spectrum analyser. I attach a copy of what I am currently decoding.

steve


Alan G4ZFQ
 

i will do. My friend is also going to look at the output on a spectrum analyser. I attach a copy of what I am currently decoding.
Steve,

Not too frightening:-)
Better than 50dB down although your standards might be a little higher for VHF?

73 Alan G4ZFQ


Steve marsh (M0NMA)
 

Right - the results of today's experimentation. (You are right that 50dB down is not much of an issue, but it is the understanding that I am trying to get to as I can't work out where in the transmission chain the artefact is coming from).

I confirmed that the beacons were isolated from the mains. The first snapshot is what I see in WSPR with the mains in my shack turned on.


You can see the harmonics at +/- 100hz and there are also stronger ones at +/-300hz, I then turned off the mains supply to my shack (shed in the garden) and repeated the test:



You can see that it made no difference in the received signal. I also ran a test against the 70cm beacon which also had harmonics, except that these appear at +/-70hz and +/-210hz.

Conclusion: Mains interference is unlikely to be the culprit as a) there was none at either end of the transmission chain and b) 70hz doesn't fit!

My friend also built a circuit that generated a pure sine wave and this also displayed the same =/-100hz harmonics when viewed using WSJT-X.

My guess is that I am looking at a harmonic generated by the sound card processing or the algorithm used in WSJT-X, but am none the wiser as to what the precise mechanism is that produces them.

Steve M0NMA


Alan G4ZFQ
 

My guess is that I am looking at a harmonic generated by the sound card processing or the algorithm used in WSJT-X, b
Steve,

Yes, maybe the worst cases are caused by the power distribution frequency producing mixing.
Your low spurii certainly looks like something getting introduced by the processing. The difference on 70cm is also puzzling, first thing I wonder is RF? The audio output to the radio needs examining..
It could be the soundcard, maybe try with a high quality device?
You have a job on your hands!

73 Alan G4ZFQ


Martin G0HDB
 

Steve:

I can't recall where I read it or any of the specific details but some time ago someone, perhaps in the USA, identified the source of the AC mains-related sidebands he was observing to RFI emanating from (if I recall correctly) the fluorescent lamps in his shack.

What you're seeing aren't harmonics as such - they're not multiples of the centre frequency - but they're sidebands spaced at 100Hz from the 'primary' signal that are the result of something introducing some additional modulation onto either the primary signal being transmitted or onto the signal coming out of the receiver and being fed into the soundcard for processing by WSJT-X.  The question is, what's introducing that additional and unwanted modulation onto the primary signal?!  That's what you need to track down.

As has already been mentioned, the sidebands spaced at +/-100Hz and perhaps multiples thereof are pretty much without doubt related to the frequency of the local mains supply, but that presumably isn't the case for the sidebands spaced at +/-70Hz - I can't offer any suggestions as to the source of those unless there's something very peculiar operating somewhere in the vicinity of either the Tx or the Rx!

--
Martin G0HDB


Don Roden
 

As strong as the primary looks on the waterfall, could it be receiver overload, or soundcard overload ?
Don W4DNR
 
 

On 2021-01-20 14:00, Martin G0HDB wrote:
Steve:

I can't recall where I read it or any of the specific details but some
time ago someone, perhaps in the USA, identified the source of the AC
mains-related sidebands he was observing to RFI emanating from (if I
recall correctly) the fluorescent lamps in his shack.

What you're seeing aren't harmonics as such - they're not multiples of
the centre frequency - but they're sidebands spaced at 100Hz from the
'primary' signal that are the result of something introducing some
additional modulation onto either the primary signal being transmitted
or onto the signal coming out of the receiver and being fed into the
soundcard for processing by WSJT-X.  The question is, what's
introducing that additional and unwanted modulation onto the primary
signal?!  That's what you need to track down.

As has already been mentioned, the sidebands spaced at +/-100Hz and
perhaps multiples thereof are pretty much without doubt related to the
frequency of the local mains supply, but that presumably isn't the
case for the sidebands spaced at +/-70Hz - I can't offer any
suggestions as to the source of those unless there's something very
peculiar operating somewhere in the vicinity of either the Tx or the
Rx!

-- 
Martin G0HDB