OP25 New User - Hitting a Setup Roadblock

[KC1UA]

Hello,

I'm finally getting around to trying out OP25. As I am running Windows 10 here, I have followed the instructions as outlined in this article in the RR Wiki:

OP25 - The RadioReference Wiki

wiki.radioreference.com

I started last evening and everything went exactly as outlined until I got to the CREATE .TSV FILES section. While I believe I understand the editing concept of both the Osomcom and Boatbod setups regarding copying trunk.tsv, renaming it, and editing it, is it necessary for me to use Libre Calc to do so? I know it is referenced but double clicking on the file as instructed brings up a text editor which obviously works. I am not seeing a "red down arrow button in the upper left" to save it as the tutorial indicates. Maybe this is due to the fact I'm not using Libre Calc. At the time of creating this post I haven't tried Libre yet. I know Notepad in Windows can edit .csv files without mucking up their syntax, but not sure if it's the same with Linux. I could be way off....

In any event, after that I am continuing to follow the instructions which tell me to " cd op25/op25/gr-op25_repeater/apps ". I am using an Airspy Mini in a single dongle configuration. I am activating it within VMWare prior to trying to run OP25. With that in mind I am using the following command line to run rx.py:

./rx.py --args 'airspy' -S 3000000 -n -N 'LNA:15,MIX:10,IF:10' -T nantucket.tsv -P symbol,constellation -2 -U -l http:0.0.0.0:8080 2> stderr.2 -X

This is of course the example command line with two changes on my part, bumping the LNA up to15, and changing trunk.tsv to nantucket.tsv. When I hit enter per the instructions I pretty much immediately return to the $. That tells me it did not run from my limited Linux experience, and attempting to run Firefox and connect via the IP address doesn't work.

This is occurring in both the Boatbod and Osmocom setups and without a doubt it is something that this newbie is doing wrong.

Probably unrelated, but the one P25 system that is consistently within range is sadly on the fringe. I don't know if that has anything to do with it or not, but I find that doubtful. I have a ton of Con+ and Cap+ DMR systems and it's my understanding that OP25 will decode them but the tutorial doesn't cover that aspect of the software and I'll need to look further. Also it was mentioned in another thread that OP25 may monitor analog Motorola systems? I'm also wondering if it will monitor conventional P25.

Admittedly I need to do some searching for answers to some of these questions, but if anyone has a handhold for my above issue I'd appreciate it!

Thanks in advance!

 

[KC1UA]

Through some trial and error I got the Boatbod version to the point where I think it's running, as I can connect to the display via 0.0.0.0:8080. As luck would have it this is one of those evenings where I'm not receiving my target system very well if at all. Ack. It appears I had some Osmocom commands in the command line that were causing me grief.

I located the documentation within the software for DMR setup. Yikes? I need to learn more about the .json files and how to call them from the command line. To complicate things I am also about 75% witless when it comes to Linux...so this is a challenge!

Back to it...

 

[boatbod]

Boatbod op25 will decode DMR Connect+ and BS Repeater, but not Capacity+ or Simplex. Honestly there are better sdr alternatives than op25 for DMR reception. OP25's strong point is P25 trunking.

 

[KC1UA]

Boatbod op25 will decode DMR Connect+ and BS Repeater, but not Capacity+ or Simplex. Honestly there are better sdr alternatives than op25 for DMR reception. OP25's strong point is P25 trunking.

Understood, and thanks. My issue as mentioned is poor reception of the only P25 system within range. I thought with trying DMR I could at least see some data and get an idea of how the software works. Another option is conventional P25 but I also have yet to figure that out as well.

I definitely want to learn the software but unfortunately between being Linux challenged and P25 system challenged I think I have a bit of an uphill battle. Our statewide system is still probably a few years out from getting to my part of Massachusetts. I'll hope for better conditions and I'll take another shot at it.

Thanks.

 

[KC1UA]

The Nantucket P25 system is being received a bit better tonight. Running it with DSD+ FL I am seeing the normal activity I would expect from monitoring a P25 system. I am receiving it "well enough" on two different antennas and have OP25 running on the better of the two. I set VMWare up on another PC this afternoon and I'm using a RTL stick on it, one of the rtl-sdr.com models. As with the original setup I can't get Osmocom to work at all; it just returns to the command line, but it appears I am getting Boatbod to run. However, I have some questions.

1. I am running Boatbod with this command line:

./rx.py --args 'rtl' -S 1000000 -n --gains 'lna:32' -T nantucket.tsv -2 -U -l http:0.0.0.0:8080 2> stderr.2

When I hit enter the cursor advances to a new line and sits. The tutorial alludes to several lines of the app's response to doing so. I see nothing. I can connect via http://0.0.0.0:8080 and I see a fluctuating frequency error. Is seeing nothing in the terminal window after running this command line normal?

2. I see nothing as far as a plot goes, so I can't tell if OP25 is receiving a signal. I thought I was supposed to see a plot, and if so am I leaving an argument out of that command line?

3. "Something" happened over the hours I was away from it. My nantucket.tsv file only contained the control channel of the system, 857.7125. However, upon returning I see that OP25 has picked up all of the alternate CC's and is now cycling through them, I'm assuming trying to find a strong enough CC to do something with. Is that correct?

4. When I originally ran OP25 with just the control channel info, under frequency error I was getting some relatively low numbers that fluctuated maybe around -15 Hz to -75 Hz or so. Now that it apparently has lost the CC that number is much larger, and to either side of about -1000 Hz. This is likely due to the P25 system being on the fringe?

5. DSD+FL is showing active talkgroups and decoding (poorly) P25 audio from them. OP25 shows no history under System Frequencies. I'm assuming same as above, just a system on the fringe which is complicating things greatly.

I guess I'm making progress, and I do have an Airspy on that same system that I can try; I wanted to see the results with an RTL stick. I have yet to make Osmocom work at all.

Any input on any of the above would be greatly appreciated. Thanks.

 

[KC1UA]

Naturally I stopped and restarted after bumping up the gain a bit, and of course I now have plots working.

But they certainly lend to a signal that is seriously on the fringe and tough to work with.

 

[CanesFan95]

That Wiki page is incomplete. I never did get around to finishing it. Try the .pdf in post #4 in this thread.

 

[KC1UA]

That Wiki page is incomplete. I never did get around to finishing it. Try the .pdf in post #4 in this thread.

Thanks, actually I did locate that and have been using it; probably should have mentioned it.

 

[CanesFan95]

You don't have to use Libre Calc, but I believe that's the default in Ubuntu which has the down red arrow button to save. You can tell OP25 to monitor 1 conventional P25 frequency, I believe with something like:

./rx.py --args 'rtl' -S 1000000 --gains 'lna:28' -2 -n -f 855.3875e6 -U -l http:0.0.0.0:8080 2> stderr.2

- or -

./rx.py --args 'airspy' -S 3000000 -N 'LNA:10,MIX:10,IF:10' -2 -n -f 769.28125e6 -U -l http:0.0.0.0:8080 2> stderr.2

And if I remember right, it is normal for the terminal window to not be showing anything after the command. If you used ./rx.py, then you must do Ctrl + c in the terminal window to quit the program. For /multi_rx.py, you just hit enter to stop it.

 

[boatbod]

With weak signals, an RTL device is generally going to need the LNA gain up around 36-39.
For Airspy you might try "LNA:15,MIX:15,IF:10".

If you have a Linux system with a display & keyboard, it's easier (IMO) to use the "curses" terminal for initial setup because you can see what's going on in the terminal screen without having to pull up a separate browser. Simply omit the "-l http:....." command line option to make that happen. On a headless system you're largely stuck with the web terminal because you won't be able to see the plots otherwise.

If possible grab some screen shots of the plots, particularly #1 (fft), #2 (constellation) and #4 (eye). They'll give the people assisting you a better idea what sort of signal we're working with.

It's been a long time since I tried running Osmocom so I can't help much there, but Boatbod is fully autonomous with tuning adjustment; you just have to get it somewhere close (within a couple kiloherts) of the control channel to get it to lock in. If you're tracking a Motorola system they typically don't rotate control channels very often so you may only need to give it the one CC freq in trunk.tsv, otherwise with Harris you'll need to list them all (comma separated) otherwise you might not find the CC on any given day.

 

[KC1UA]

Trying some conventional P25 with the above command line structure now that I have a moment. I am able to see the FFT plot and signals to either side of where I'm tuned so if someone will be so kind to key the mic on the frequency I'm monitoring I'll know if I have success.

For Boatbod, I am running Ubuntu 20.04 LTS in VMWare so I do have display and keyboard access. I'll try to figure out the curses terminal.

I'm using an RTL right now and for the P25 conventional frequency I'm trying to listen to I have the gain at 30 as it's a relatively strong signal at my location. I tried with the Airspy as well; funny thing is that this particular setup for whatever reason does not like my Airspy R2 and basically blows it up to the point that not even a reboot into Windows 10 brings it back; only an unplug/replug does. So I'm sticking with the RTL device for now.

Okay on screenshots, some will follow at some point. Unfortunately I just discovered that OP25 is not decoding the conventional P25 frequency I'm monitoring using Cane's command line as shown in his message. I'm monitoring it with another radio and just heard several transmissions. While I can see the signal in OP25's FFT, and it's strong, no audio decode.

Thanks to both of you for the help, I apparently need it!

 

[boatbod]

Okay on screenshots, some will follow at some point. Unfortunately I just discovered that OP25 is not decoding the conventional P25 frequency I'm monitoring using Cane's command line as shown in his message. I'm monitoring it with another radio and just heard several transmissions. While I can see the signal in OP25's FFT, and it's strong, no audio decode.

Thanks to both of you for the help, I apparently need it!

Turn the logging verbosity up to -v 11 and check your stderr.2 file to see if there's any hints why the P25 conventional is not decoding.
Also those screen shots might provide some insight. Note that P25 Conventional is always fsk4 modulation so you'll probably need to add "-D fsk4" on the command line otherwise it'll try to decode using cqpsk which may or (more likely) may not be possible.

 

[KC1UA]

Hi,

Added -v 11 and -D fsk4 to my command line which now shows as:

./rx.py --args 'rtl' -S 1000000 --gains 'lna:30' -2 -n -v 11 -D fsk4 -f 156.2175e6 -U -l http:0.0.0.0:8080 2> stderr.2

Still no audio, but hopefully below screen shots and file attachment of stderr.2 may lend to my issue.

I've also read somewhere in all of the files that I might have to open port 23456 to listen to audio, and tried ./audio.py -u 23456 out of the apps folder but got back a message saying the port is already in use, so maybe that's automatic?

Spectrum plot with signal present:



How the home screen appears:


The stderr.2 file is zipped and attached. Will add more screenshots of what may be deemed pertinent data. Thanks for hanging in there with me.

 

[KC1UA]

Screenshots of the symbol plot and the datascope attached. I'm guessing by what I'm looking at I'm working with a pretty good signal.

Calling it a night, back at it as I can tomorrow. Thanks again.

 

[boatbod]

Hmmmm. The log contains a whole lot of these every time there is a transmission:

01/29/23 19:42:28.234096 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d1055e1583fb12b6, ec=0
01/29/23 19:42:28.234120 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d1055e1583fb12b6
01/29/23 19:42:28.372260 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d10ab169115c4caa, ec=0
01/29/23 19:42:28.372285 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d10ab169115c4caa
01/29/23 19:42:28.505967 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d1055e1583fb12b6, ec=0
01/29/23 19:42:28.506000 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d1055e1583fb12b6
01/29/23 19:42:28.764385 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d10ab169115c4caa, ec=0
01/29/23 19:42:28.764408 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d10ab169115c4caa
01/29/23 19:42:28.896906 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d1055e1583fb12b6, ec=0
01/29/23 19:42:28.896927 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d1055e1583fb12b6
01/29/23 19:42:29.033538 [0] p25_framer::nid_codeword: duid/parity check fail: nid=d10ab169115c4caa, ec=0
01/29/23 19:42:29.033560 [0] p25_framer::rx_sym() error check failed, frame discarded, nid=d10ab169115c4caa

For whatever reason p25_framer::nid_codework() is failing the error checking even though the signal looks good and the actual error count is zero (ec=0). It'd be really handy if you could capture a raw symbol trace of this behavior so that I can replay it in my development environment. With rx.py the way to do this is add "-r capture.bin" on the command line and then let the app run just long enough to grab a couple of live transmissions. If you let it run a long time the file will get large...

Zip up the capture.bin, send it over and I'll take a look at what's going on.

 

[KC1UA]

Here you go. Again, many thanks for taking the time to have a look.

 

[lwvmobile]

Here you go. Again, many thanks for taking the time to have a look.

FWIW, I just ran that bin file through FME, and it decodes voice loud and clear, but there is no trunking data (TSBKs) so is that a trunking system, or just a conventional P1? Here is my log dump and per call wavs I made off of it.

 

[KC1UA]

Hi, thanks for having a look at it. Interesting...maybe I have an audio issue. I ran the available audio test in Linux and was able to hear audio from both speakers, but nada via OP25 obviously.

The capture is of a conventional system on Marthas Vineyard island.

 

[KC1UA]

I have VMWare set up on two different PC's so I ran my other one. In this case I am using an Airspy. The results are the same and I've attached another capture.bin. I'm guessing the results of running it are the same. I'm also attaching a screenshot of where I am testing audio in Linux. In both cases audio is definitely working at the OS level.

As I have two separate setups exhibiting the same behavior I'm guessing that something became misconfigured along the way; just what remains to be seen.

 

[lwvmobile]

As I have two separate setups exhibiting the same behavior I'm guessing that something became misconfigured along the way; just what remains to be seen.

No idea. Could be a Boatbod OP25 thing, or just a simple misconfiguration issue. Again, I was able to decode the voice on the symbol capture bin file just fine with FME, but would need Boatbod to weigh in again after trying out your bin files to see if its something on his end or on your end. If anything, pretty sure we can rule the signal out as an issue, decoded just fine for me, but had issues with OP25 playing back the bin file, but could also be my bin playback configuration or something got broke along the way. Not sure.