OP25 Odd frequency error

[NCAmateurRadioGuy]

This isn't an issue with reception or anything, just wanted to see what might be the cause. I had one NESDR running a single OP25 instance monitoring a VHF P25 system and an 800 MHz P25 system. The frequency error stayed around +50 Hz or so on both systems. I then decided to monitor both systems independently using two NESDRs. I used the same NESDR for the 800 MHz system as was running the previous two system config but now the frequency error reads +550 Hz and if I try to use the fine tune offset, OP25 starts dropping tsbks and eventually tries to find the channel again. I also tried the -X option to see if it was maybe a glitch in the interface or something and that just fails to keep on the control channel.
Again, this isn't an issue, the setup works fine, not missing tsbks and audio is great but just thought it was odd, using the same NESDR and antenna but the frequency error changing +500 Hz.

 

[boatbod]

This isn't an issue with reception or anything, just wanted to see what might be the cause. I had one NESDR running a single OP25 instance monitoring a VHF P25 system and an 800 MHz P25 system. The frequency error stayed around +50 Hz or so on both systems. I then decided to monitor both systems independently using two NESDRs. I used the same NESDR for the 800 MHz system as was running the previous two system config but now the frequency error reads +550 Hz and if I try to use the fine tune offset, OP25 starts dropping tsbks and eventually tries to find the channel again. I also tried the -X option to see if it was maybe a glitch in the interface or something and that just fails to keep on the control channel.
Again, this isn't an issue, the setup works fine, not missing tsbks and audio is great but just thought it was odd, using the same NESDR and antenna but the frequency error changing +500 Hz.

Which version of op25 - osmocom or boatbod?
Which hardware (RPi3, RPi4, desktop, laptop?)
I have seen curious behavior when changing the number of devices plugged in to a system. Have you tried running just one instance of op25 with both dongles plugged in, and then one instance with just one dongle plugged in?

 

[NCAmateurRadioGuy]

This isn't an issue with reception or anything, just wanted to see what might be the cause. I had one NESDR running a single OP25 instance monitoring a VHF P25 system and an 800 MHz P25 system. The frequency error stayed around +50 Hz or so on both systems. I then decided to monitor both systems independently using two NESDRs. I used the same NESDR for the 800 MHz system as was running the previous two system config but now the frequency error reads +550 Hz and if I try to use the fine tune offset, OP25 starts dropping tsbks and eventually tries to find the channel again. I also tried the -X option to see if it was maybe a glitch in the interface or something and that just fails to keep on the control channel.
Again, this isn't an issue, the setup works fine, not missing tsbks and audio is great but just thought it was odd, using the same NESDR and antenna but the frequency error changing +500 Hz.

Also, forgot to mention, when on a voice channel, the frequency error usually falls to less than 50 Hz, often <10 Hz.

Which version of op25 - osmocom or boatbod?
Which hardware (RPi3, RPi4, desktop, laptop?)
I have seen curious behavior when changing the number of devices plugged in to a system. Have you tried running just one instance of op25 with both dongles plugged in, and then one instance with just one dongle plugged in?

boatbod repo, of course haha.
Its running on an aging desktop, Athlon II, if I'm not mistaken.
I haven't done any troubleshooting stuff as it has no negative side effects, strong decode, audio great, etc.
The desktop is run headless and I access the audio and web terminals from my main PC, as you might know from being such a help in the recent past.

 

[boatbod]

Also, forgot to mention, when on a voice channel, the frequency error usually falls to less than 50 Hz, often <10 Hz.


boatbod repo, of course haha.
Its running on an aging desktop, Athlon II, if I'm not mistaken.
I haven't done any troubleshooting stuff as it has no negative side effects, strong decode, audio great, etc.
The desktop is run headless and I access the audio and web terminals from my main PC, as you might know from being such a help in the recent past.

Hmmm, it's certainly not obvious what's going on. I have seen what appear to be spurious frequency errors reported by the demodulator but for the most part not enough to cause the auto-tracking to fail (at least on any of my systems). I'd guess it might be noise causing it? How close are the antennas & dongles to the computer?

 

[NCAmateurRadioGuy]

The dongles are directly connected to the PC, with one antenna (for the VHF system which is maybe a mile, two at most, away) in the room 5 feet away and the other antenna is a discone outside around 20ft up. The 800 MHz tower is further away, but I get a strong enough signal to get great results. Electrical noise hasn't been an issue for me, since these dongles are higher quality with metal cases and I'm not working with really weak signals or anything. The only thing I could assume is since the 800 MHz signal is weaker, the additional NESDR dongle beside it might be what is causing the spurious error.

The spurious error certainly does cause auto-tracking to fail for me, it will run for a few minutes, steadily getting a lower decode rate then drop completely. The only reason I even tried it was just to see if it was an actual error that could be mitigated, but it seems to be just an odd side effect of some sort.

 

[boatbod]

The dongles are directly connected to the PC, with one antenna (for the VHF system which is maybe a mile, two at most, away) in the room 5 feet away and the other antenna is a discone outside around 20ft up. The 800 MHz tower is further away, but I get a strong enough signal to get great results. Electrical noise hasn't been an issue for me, since these dongles are higher quality with metal cases and I'm not working with really weak signals or anything. The only thing I could assume is since the 800 MHz signal is weaker, the additional NESDR dongle beside it might be what is causing the spurious error.

The spurious error certainly does cause auto-tracking to fail for me, it will run for a few minutes, steadily getting a lower decode rate then drop completely. The only reason I even tried it was just to see if it was an actual error that could be mitigated, but it seems to be just an odd side effect of some sort.

What do the two mixer plots look like?

 

[NCAmateurRadioGuy]

The 800 MHz mixer is well centered, a nice peak up to just shy of -40 dB (varies a little depending on conditions due to how directional 800 MHz is) with a mixer balance of between 20 and 50 usually (of course this varies as well). The VHF plot is also well centered, nice peak up to around -30 dB and a nice low balance of between 0 and 30.

 

[boatbod]

The mixer plot is generally a nice smooth signal hump, but sometimes you'll see signs of adjacent channel interference. Here's mine below. Note the small disturbance at +0.21; this sometimes gets bigger, but fortunately doesn't cause too much of a decode issue. I'm pretty sure the four little "horns" on the main signal are the fsk4/qpsk modulation.

 

[NCAmateurRadioGuy]

Mine looks similar with really minor humps, small enough to maybe just be noise but I'm not certain.