Unitrunker decode issues on Harris systems

[tbiggums]

I'm trying to use Unitrunker 1.0.26.16 with some Harris P25 systems, and no matter what I do, I can't get the decode health better than 65% or so. The control channels are nice and strong at my location, and the one shown in the Uniscope screenshots is from a transmitter site literally three blocks away from me, so it shouldn't be a signal strength issue:

I've tried Unitrunker on two different Harris P25 control channels in the area, and I never get much above 65% on either. The Harris CC's around here are not part of simulcast systems--they're single transmitter sites.

I've tried both a discriminator tapped PRO-2035 and the PSR-800's built in "IF" output from the headphone jack. I get nearly identical results with either radio.

If I tune to a Motorola P25 control channel, or an EDACS one, I get almost 100% decode health. It's only on these Harris P25 systems that I can't get it above 65%.

I've also tried a second computer (one running XP, the other Windows 7), and the results are the same.

Pro96Com does seem to decode these Harris systems 100% (with a PRO-2096), but I'd prefer to use Unitrunker.

I've also attached a log file from the close-by Harris site.

Anyone have any suggestions for improving the decode rate?

 

[SCPD]

First off - thanks for posting the Uniscope images! Much easier to comment when I can see what's happening.

There's a small amount of fuzz in the signal. This is likely what is hurting your decode rate.

From the program, the only thing I can suggest is raising the sample rate from 48000 to 96000.

Outside the program - try different antennas or placement of the radio.

An overly strong signal can cause problems. Try attenuate - if the radio has it.

Oh ... wait. Is this Harris system on UHF by chance?

 

[GTR8000]

It's this system...

Lakewood PD / West Metro Fire / Wheat Ridge Trunking System, Lakewood, Colorado - Scanner Frequencies

I've experienced similar issues with Unitrunker having trouble decoding Harris P25 control channels. Perhaps not coincidentally, several users have reported scanners having trouble with decoding them as well.

 

[SCPD]

Ha!

Yeah, I should have looked at the decode log (thanks for that too).

Motorola 700/800 systems put out 40 messages per second. This Harris system only puts out 27 messages per second. The program assumes 40 messages = 100%. This system (and others like it), 27 is actually 100% so you are already at 100%.

Adjusting the decode rate calculation based on vendor behavior is an old bug that I still have yet to get around to fixing. Understand that the only problem here is the decode rate is under-reported.

Thanks for the reminder - and the screenshots + log.

 

[tbiggums]

Cool, thanks... So I should just assume the 65% health is 100%, then? I figured the decode health was based on the amount of forward error correction used.

As long as I'm not likely to miss affiliations or anything, I'm glad Unitrunker will work on this system. I mostly want to run this on an older Pentium 4 3GHz machine, which runs about 35% CPU usage using Pro96Com, but barely busts a sweat running Unitrunker.

I tried a bunch of series resistor and voltage divider combinations on the PRO-2035 discriminator tap, and as long as the signal wasn't clipping or way too low, it didn't seem to make much difference. The same was true with adjusting the volume on the PSR-800's IF tap.

 

[Forts]

A big EDACS system close to me recently switched to a Harris P25 system too... and I was wondering why I was struggling getting a strong signal (even though UT itself was working great)... So I just checked my decode rate... 65%