"Black box" Raspberry PI based DMR decoder?

[rumcajs_tr]

Hello, i am looking for solution for adding a "black box" digital voice decoder based on Raspberry PI (or similar micro computer) as an addition to older analog receiver with discriminator or I/O output (IC-R8500 in my case). AOR makes similar box ARD300 but with display and for enormous price.
My idea is to use a Raspeberry PI (model 4 could be powerfull enough) and use the audio IN for connection to receiver's output, then process the signal in the blackbox by using DSD+ or some linux clone and then output the decoded audio to the audio OUT of the PI. So basically an inteligent black box connected between radio discriminator and a speaker with USB power supply.
Does anybody have some knowledne about similar project or can give advice how to proceed?
Thanks,
73, Ondrej

 

[wgbecks]

My suggestion would be to use op25 for decoding (listening) to P25 Conventional, P25 P1 & P2 Trunked, DMR, YSF, and DSTAR. In most cased you'll only need one SDR. No need for external connections to a radio discriminator or having to tune that radio to follow trunked systems.

 

[rumcajs_tr]

Hello, thank you for suggestion with the op25. I have multiple SDRs (airspy r2, SdrPlay Duo,...). However, I wan to make use of my desktop IC-R8500 communications receiver, which is supperior regarding RF reception (sensitivity, dynamic range, etc...) - much better than SDRs. There are many weak signals, which I would like to decode but I am not able to pick them up on the SDRs.
I can use PC for audio decoding, however I don't want to leave PC running so I am looking for some black box solution. A friend of mine uses ARD300 from AOR, which us excelent, but very expensive.
Does the op25 can be udes with audio input instead of data stream from USB stick?

EDIT: Another idea that comes in my mind is to use one of the AMBE3000 chips (installed on USB stick or PI Hat) to decode the audio stream ... however, I don't have the knowledge to write such software to use the chips.

EDIT2: Also, does the op25 decode DMR or only P25? I am interested in DMR only.

 

[wgbecks]

First let me say that I have had excellent results with the Airspy R2 as an SDR with OP25 but to answer your question, I believe that you can still use raw audio input from a discriminator although I've not tried that since rx.py has replaced scope.py. It's one thing is you're trying to decode weak signal HF stuff but then you wouldn't be using op25 or similar digital decoders. So unless you are located in a an environment with numerous strong RF interference that give the Airspy R2 a try.

Now there are two components (decoder wise) in op25. First is rx.py that decodes P25 Phase I and Phase II transmissions and well as conventional P25 non-trunked systems. Then there is multi_rx.py that will decode DMR among a few other digital modes. AFSIK multi_rx.py does not have any options for raw discriminator audio but you could post that question on the SDR forum to confirm.

 

[rumcajs_tr]

Thank a lot for useful info. I will do a fresh PI install and try to get op25 working. A also think that I should first test the audio levels from discriminator output on windows pc to see if the decoder is getting useful signal before messing with linux. The IC-R8500 has the discriminator output out of the box with just jumper setting and no modification/soldering needed so I wanted to give it a try. I will do some tests and report back later.

 

[kruser]

I will do some tests and report back later.

I'd be interested in your results as well.
I used an R8500 as a discriminator audio source before but for some reason, I think I had issues getting mine to pass a clean signal to the external decoder. I know I never put the 8500 on a scope so I'm only going from memory and maybe it was fine but an issue with the decoder software back then. It's been a long time since I'd messed with it and software decoding has improved greatly since then.

Your idea sounds like a fun project and one that I'd like to see also. Mainly so I can do it without an actual computer running.
Something low power like a PI sounds like a perfect solution if it has the power to handle the decoding functions. I'd mainly be interested in DMR as well but other modes would be welcome.
Hopefully multi_rx.py can take a raw audio signal for its input and this was just not mentioned.

I've messed with various SDR sticks with mixed results. My RF environment is very tough with a lot of high power VHF paging signals and a nearly fully saturated 800 MHz band with hardly any separation left between channels. Many of my Icom receivers will handle this environment very well but not all have an easy option to output raw discriminator audio. I'd love to use the R9000 for this but finding a suitable tap has not been found for that model. I'm using an R2500 right now with pretty decent results into DSD+ Fastlane and a sound cards line input but that's not really the low power solution I'm after.
Using something like the PI could also be more of a portable application if the right receiver can be found. I have an Icom R30 but never looked into finding out if it can be easily tapped. If so, this could be a good combo for DMR being as Icom has not shown any desire to support DMR in any of its newer digital capable models.

 

[rumcajs_tr]

Hi, I will start with testing the discriminator audio on Windows PC to be sure that the decoding is working. After that, I will start to experiment with Raspberry PI and op25. The problem is that I have almost zero experience with linux so the learning curve will be quite long and can only work on it during weekends. So it will take some time

 

[Unitrunker2]

Feed a cheap Realtek SDR stick with actual IF from your IC-R8500. You keep the benefit of the Icom's RF front-end.

 

[KA1RBI]

Some notes on various issues that were mentioned....

1) OP25 multi_rx.py supports the disc tap. Set the "args" parameter in the "device" entry to something like 'audio:X' (where X is the usual ALSA name of the desired audio device, for example: 'audio:default'). The "gain" parameter likewise would be set to something like "audio:10" (where 10 is the gain multiplication factor). You want to use the plot "datascope" and "symbol" options to set the proper gain value...

2) AFAIK the R PI (even the PI4) has audio outputs but no audio input - meaning that without some sort of external sound device there is no place to plug the audio from the disc tap... (I'd love to be shown to be in error on this point).

According to Audio Input – Raspberry Pi Projects

The RPi doesn't have an audio input....

3) Did some testing a while back using the disc tap output of an R-8500 (after setting the jumper as mentioned above) and found that it was unusable for high speed (9,600) data. Perhaps that selected unit was not functioning properly... There was definitely audio there, but impossible to recover a clean signal from it... In sum, my disappointing results were identical to those reported above by kruser... You are quite correct though that using such a radio (if it would work) as the RF front end is going to provide superior results over most SDR sticks...

4) Also as Bill has mentioned when using OP25 to decode DMR you would use the multi_rx app...

73

Max

 

[KA1RBI]

Feed a cheap Realtek SDR stick with actual IF from your IC-R8500. You keep the benefit of the Icom's RF front-end.

This is a possibility, but the trouble you run into even after getting the signal levels of the two devices to be compatible (somehow) you're still left with the problem that the dynamic range of the Icom radio IF channel in general will greatly exceed that of the SDR. You've simply shifted the frequencies down to a lower IF.

There are many weak signals, which I would like to decode but I am not able to pick them up on the SDRs.

Yep - so how would simply shifting a chunk of spectrum downwards into the IF range help in any way to mitigate this?

Max

 

[rumcajs_tr]

Hi, I made the IC-R8500 decoding both DMR and D-STAR today using Windows PC and DSD 2.212 Fastlane program. I just switched the jumper as descibed in manual, connected to AGC cinch and used this cheap ebay cable:

1m 3.5mm Mono Jack Stecker Einzel RCA Phono Anschlusskabel 008246 for sale online | eBay

Find many great new & used options and get the best deals for 1m 3.5mm Mono Jack Stecker Einzel RCA Phono Anschlusskabel 008246 at the best online prices at eBay! Free shipping for many products!

www.ebay.com

The input to the PC is set up as "line-in" and volume is set to 50%.
So the decoding is working, now I have to head to Raspberry. BTW, I have two Chinese cheap external USB sound cards so I will try which one works better.

 

[Unitrunker2]

You've simply shifted the frequencies down to a lower IF.

The idea is that the RF front end of the more expensive radio will produce cleaner IF than the SDR can do on its own. Example: person with successful decode of discriminator audio from an Icom rig tries to duplicate that experience on a Realtek SDR but sees poor results.

The pan-adapter also allows decoding two or more nearby signals which a plain disc. tap can't do.

It's all about experimentation. If already you have both components then cost is more in time than dollars. If you have the money but no time then yes go buy a better SDR!

 

[rumcajs_tr]

OK, just a short update. I gave up with the Raspberry Pi (was not powerfull enough) and went with the Windows solution.
Bought a PiPo X8 W10 on eBay for USD 130 and external sound card Behringer UCA 202 for USD 25 and a RCA cable to my IC-R8500. Working great with DSD+ FastLane.