locked Weird drift in Odd Tx signals #decode #transmit


Andrew Buskov
 

I tried searching the group threads and Google, but haven't found anything similar. Can someone explain to me why the odd tx cycles has some seriously drifting signals? It started out as one or two, but by the time I got some video it was quite a fe. Video was taken about 09:45 UTC, and I was initially thinking it was something atmospheric, but it doesn't seem to happen on the even tx cycles.
https://imgur.com/a/pQr7Sih


Reino Talarmo
 

I tried searching the group threads and Google, but haven't found anything similar. Can someone explain to me why the odd tx cycles has some seriously drifting signals? It started out as one or two, but by the time I got some video it was quite a fe. Video was taken about 09:45 UTC, and I was initially thinking it was something atmospheric, but it doesn't seem to happen on the even tx cycles.
https://imgur.com/a/pQr7Sih

Hi Andrew
Do you mean those signals that are e.g. from about 1640 Hz to 2040 Hz i.e. signal width 400 Hz. There are faint traces connecting those vertical lines together i.e. typical GFSK signal with smooth frequency transitions as FT8. Most probably those are 8th harmonics of an audio FT8 signal. From the pattern I see in the first three instances that the same message is sent.
The same signal at about 17th harmonics is present from about 800 hz upwards; the same pattern is present.
Why that high order harmonics are present is most probably due to a serious audio path overdriving. It is difficult to tell on which frequency the actual FT8 signal is, but one at 2450 Hz may the original signal.
What rig you have? Is it possible that the is some overdrive inside your rig or PC? If you see similar situation again, you may try to reduce RFF gain or some audio gain.

73, Reino OH3mA


Michael Black
 

I assume you are referring the fast deviations seen on each signal?
That's the FT8 protocol shifting frequencies.  If you slow down the waterfall you should see a more solid line.
Mike W9DMB

On Tuesday, December 6, 2022 at 06:51:13 AM CST, Andrew Buskov <andrew.buskov@...> wrote:

I tried searching the group threads and Google, but haven't found anything similar. Can someone explain to me why the odd tx cycles has some seriously drifting signals? It started out as one or two, but by the time I got some video it was quite a fe. Video was taken about 09:45 UTC, and I was initially thinking it was something atmospheric, but it doesn't seem to happen on the even tx cycles.
https://imgur.com/a/pQr7Sih


Michael Black
 

I didn't notice those the 1st time I watched this video.

In WSJT-X you will see vertical bars.  What you can do is look at the width of the vertical bars to determine the harmonic involved.

3rd order = 150Hz wide
5th order = 250Hz wide
7th order = 350Hz wide

You can take the starting offset of the vertical bars and divide by the harmonic to find the offset of the offending transmitter.

e.g. a signal at offset 300 will generate 900Hz 3rd order (150Hz wide) and 1500Hz 5th order (300Hz wide)
So if you see 150Hz wide vertical bars at 900hz you do 900/3=300 to find the transmitter.
Look them up on QRZ and point them to our paper on how to set digital audio levels.
https://www.dropbox.com/s/hlt50x44qs2ua1f/FT8Noise8.pdf?dl=0

We've been working to clean up signals for a few years now but they show up after every contest now.  I only see 1 or 2 per month now.

Mike W9MDB







On Tuesday, December 6, 2022 at 10:18:20 AM CST, Reino Talarmo <reino.talarmo@...> wrote:





I tried searching the group threads and Google, but haven't found anything similar. Can someone explain to me why the odd tx cycles has some seriously drifting signals? It started out as one or two, but by the time I got some video it was quite a fe. Video was taken about 09:45 UTC, and I was initially thinking it was something atmospheric, but it doesn't seem to happen on the even tx cycles.
https://imgur.com/a/pQr7Sih

Hi Andrew
Do you mean those signals that are e.g. from about 1640 Hz to 2040 Hz i.e. signal width 400 Hz. There are faint traces connecting those vertical lines together i.e. typical GFSK signal with smooth frequency transitions as FT8. Most probably those are 8th harmonics of an audio FT8 signal. From the pattern I see in the first three instances that the same message is sent.
The same signal at about 17th harmonics is present from about 800 hz upwards; the same pattern is present.
Why that high order harmonics are present is most probably due to a serious audio path overdriving. It is difficult to tell on which frequency the actual FT8 signal is, but one at 2450 Hz may the original signal.
What rig you have? Is it possible that the is some overdrive inside your rig or PC? If you see similar situation again, you may try to reduce RFF gain or some audio gain.

73, Reino OH3mA


Robert Lorenzini
 

So that's why I like you. I have to give Mike a lot of credit for all the effort
he selflessly puts into helping out hams to such an extent that he does.
Logging in to strangers computers to fix their problems is beyond the pale.
And there is the Linux development work to keep this humming.

Bob - wd6dod

On 12/6/2022 8:48 AM, Michael Black via groups.io wrote:
I didn't notice those the 1st time I watched this video.

In WSJT-X you will see vertical bars.  What you can do is look at the width of the vertical bars to determine the harmonic involved.

3rd order = 150Hz wide
5th order = 250Hz wide
7th order = 350Hz wide

You can take the starting offset of the vertical bars and divide by the harmonic to find the offset of the offending transmitter.

e.g. a signal at offset 300 will generate 900Hz 3rd order (150Hz wide) and 1500Hz 5th order (300Hz wide)
So if you see 150Hz wide vertical bars at 900hz you do 900/3=300 to find the transmitter.
Look them up on QRZ and point them to our paper on how to set digital audio levels.
https://www.dropbox.com/s/hlt50x44qs2ua1f/FT8Noise8.pdf?dl=0

We've been working to clean up signals for a few years now but they show up after every contest now.  I only see 1 or 2 per month now.

Mike W9MDB







On Tuesday, December 6, 2022 at 10:18:20 AM CST, Reino Talarmo <reino.talarmo@...> wrote:





I tried searching the group threads and Google, but haven't found anything similar. Can someone explain to me why the odd tx cycles has some seriously drifting signals? It started out as one or two, but by the time I got some video it was quite a fe. Video was taken about 09:45 UTC, and I was initially thinking it was something atmospheric, but it doesn't seem to happen on the even tx cycles.
https://imgur.com/a/pQr7Sih

Hi Andrew
Do you mean those signals that are e.g. from about 1640 Hz to 2040 Hz i.e. signal width 400 Hz. There are faint traces connecting those vertical lines together i.e. typical GFSK signal with smooth frequency transitions as FT8. Most probably those are 8th harmonics of an audio FT8 signal. From the pattern I see in the first three instances that the same message is sent.
The same signal at about 17th harmonics is present from about 800 hz upwards; the same pattern is present.
Why that high order harmonics are present is most probably due to a serious audio path overdriving. It is difficult to tell on which frequency the actual FT8 signal is, but one at 2450 Hz may the original signal.
What rig you have? Is it possible that the is some overdrive inside your rig or PC? If you see similar situation again, you may try to reduce RFF gain or some audio gain.

73, Reino OH3mA