DSD - Decoding DMR almost 100%. Can't get p25 to work.

[reformedMhacker]

I've been playing with DSD from the beginning, but have just had time recently to really work with it.
I'm decoding Trbo and Nexedge almost 100% of the time, but can't get p25 to decode worth anything.
I've tried just a discriminator tap. I've tried with a 10k resistor, 2 10k's in series, a 100k, all these with
and without the cap. I've tried Line-ins and mic inputs. I've tried on linux and the windows port.
I've tried the input levels pretty much everywhere. (I even tried the windows port on two pc's.)

I know its not an encryption issue because I even tried decoding a p25 portable on simplex (no crypto
option in the radio) to no avail. It's still garbled (even attenuated the input to the receiver to be sure
I didn't overload the front end of the radio).

It seems like everyone else has the opposite problem, they get p25 just fine
but find DMR to be more temperamental.
Does anyone have any idea what I may be doing wrong?

 

[Boatanchor]

Try forcing DSD to run in GFSK mode by adding '-mg' to the command line in the batch file.

I know P25 phase 1 is supposed to be C4FM, but in my experience (and many others for that matter) the DSD C4FM implementation doesn't work very well at all for some reason.

Maybe it is CPU/System dependent, but I have confirmed that GFSK demodulation works infinitely better on several of my computers ranging from single core 1.2Ghz laptops, up to my dual core 3Ghz desktop.

I went from garbled, barely intelligible audio, with a high BER (error bars to infinity) in the default C4FM mode, to almost perfect Phase 1 audio, with virtually no error rate, simply by switching to GFSK.

Let us know how you go

 

[reformedMhacker]

Still nothing. I've tried -mg, and even tried just for the heck of it -mc and -mq. I even tried -f1 so it would
know to only try for p25.

 

[Boatanchor]

Bugger!!

Sorry, I'm all out of ideas

 

[moonbounce]

With decoding Apco 25, ( using SDR Sharp/DSD ) I get a 100 % decode in the 800 Mhz band and about 80% decode in the VHF band ( 142 Mhz ) so I am thinking it has something to do with the actual band you are on, not the CPU/RAM combination that one has in the computer.

I think the CPU/RAM combination will affect whether you can run DSD and or SDR Sharp together. I had an old laptop that wouldn't run dsd, but it would run Windows XP and anything else I could throw at it.

Moonbounce

 

[ilgrant]

What are you using as your receiver radio?

 

[moonbounce]

I was using an old scanner and a discriminator tap out to line in on the computer and DSD 1.6. Now I use an RTL dongle R820T, SDR Sharp and a VB cable, a much better setup in my opinion, Gets rid of the whole discriminator tap thing.

Moonbounce
.

 

[woody_46]

I have found it takes a lot of experimenting to get it right. The radio I thought would be the best (Icom R8500) was the hardest to get working and only worked with version 1.40. My Optocom was better with the new version while an older tapped Pro 2045 works the best of all. That was with trying them on 2 different dual core computers, one with a Realtek soundcard and the other with an AC 97 soundcard. Input levels are really important and I have little difference .between the 2 soundcards, both work.

Oh I should add that VHF will be more problematic due to channel spacing, overload from FM broadcast, pagers, etc.

RMH you never mentioned what radio you are using..

 

[ilgrant]

I was using an old scanner and a discriminator tap out to line in on the computer and DSD 1.6. Now I use an RTL dongle R820T, SDR Sharp and a VB cable, a much better setup in my opinion, Gets rid of the whole discriminator tap thing.

Moonbounce
.

Thanks for the info, Moonbounce. My fault...I was actually wondering what the original poster was using for a receiver.

 

[reformedMhacker]

Sorry for the delay, I've been at work for 48+ hrs straight. I'm using an old PRO-2037 for my receiver.

 

[woody_46]

I'm not an expert on this, but yes that is odd. Is it a VHF or UHF system? How's the signal strength, etc. Maybe a link to the system might help. If it's decoding the other modes it should do the P25 too.