dsd and icom r-8500

[psointu]

i've had very good luck using dsd v1.6, using mbelib 1.2.5, to decode a local 2m ham repeater ... my setup was a radio shack pro-2006 with the discriminator tap described at Discriminator Output PRO-2006 / COM-205 .. the only exception was i used a 10k ohm resistor not a 12k; the cap was 2.2n as described ... patched into the mic input of an old acer aspire running ubuntu 10.04 with an embedded realtek alc268 chipset based sound card that doesn't offer a line input - only mic in ... as mentioned this has worked flawlessly decoding dmr.

i recently obtained an icom r-8500 and i "created" a discriminator tap by following the instructions at Discriminator Output IC-R8500 ... using the same exact laptop with the r-8500 tuned to the same 2m repeater, i'm struggling to get a decode of it's dmr signal.

i used a breadboard and some spare resistors and caps of varying specification to see if i could stumble upon some combination that would help along with adjusting the input level of the mic input to no avail.

clearly there's something electrically different about the signal offered by the r-8500's disc tap as compared to the rs pro-2006.

anyone tried using an r-8500 with dsd that might offer some insight? otherwise, anyone have any advice on possible solutions?

 

[slicerwizard]

anyone have any advice on possible solutions?

Upload a sample of DMR raw audio. zippyshare.com is an easy to use site.

 

[psointu]

thanks for the reply ... i'm not quite sure how to create a sample of dmr raw audio ... i'm guessing that dsd's -w option is not correct ... would it's -d option to create mbe data files do this?

 

[psointu]

i thought about the concept of capturing raw audio some ... i think this requires recording the signal at the mic input ... so i used the alsa arecord binary invoked with "arecord -d60 -vv -fdat <filename>.wav" and captured the following :

https://dl.dropboxusercontent.com/u/48265006/audio/rspro2006.wav
https://dl.dropboxusercontent.com/u/48265006/audio/icomr8500-12.0khz.wav
https://dl.dropboxusercontent.com/u/48265006/audio/icomr8500-5.5khz.wav

the first is from the rs pro-2006's disc tap ... again that decode is perfect ... i recall reading somewhere that with dmr a reported from dsd inlvl in the single digits should work well and with the pro-2006 i get error lines like these with a perfect decode:

Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=====
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=


the second and third are from the icom r-8500's disc output ... i recalled reading somewhere that kicking the fm filter down from 12.0KHz to 5.5KHz solved a decode issue for another poster in the forums ... i *think* that was a p25 decode issue but i figured it couldn't hurt to try so i captured it for both fm filter settings.
getting any audio out of dsd with the icom requires pumping either the mic input level or the mic boost level ... in either case, the lowest reported inlvl i can get from dsd is in the 9-10% range but the audio is gibberish and the error lines look like this:

Sync: -DMR mod: C4FM inlvl: 9% [SLOT0] slot1 VOICE e:====R====R=======R====R====R====RM====R=======R====R=======R=========R
Sync: -DMR mod: C4FM inlvl: 9% [SLOT0] slot1 VOICE e:=========R==========R========R====R=======R=====R=======R======R=======R======R
Sync: -DMR mod: C4FM inlvl: 10% [SLOT0] slot1 VOICE e:====R=====RM====R=========R=======R====R======R====R====R====RM====R========R
Sync: -DMR mod: C4FM inlvl: 10% [SLOT0] slot1 VOICE e:=======R====R==========R====R=====R=====RM=======R====R====R=====RM=====R====R====R=====RM
Sync: -DMR mod: C4FM inlvl: 10% [SLOT0] SLOT1 VOICE e:========R====R=======R====R====R=====RM====R===========R====R========R========R=====R
Sync: -DMR mod: C4FM inlvl: 10% [SLOT0] slot1 VOICE e:====R=====RM====R======R====R====R====RM====R=======R====R====R====RM====R==========R

 

[racingfan360]

psointu - so I've tried your samples with DSD+, which I find decodes better than DSD in almost all circumstances. While sample 1 is perfect the other two are terrible....you can just about make out some voice, lots of errors shown but it does decode as DMR DCC7.

I don't quite understand why you refer to a breadboard/resistors/caps etc though: It will be needed for the 2006, but are you using it for the icom too? I don't have an icom but it looks like to add a discrim tap its a jumper setting and nothing more?

Jim

 

[psointu]

hi jim ... after my initial results using the r8500 were not successful, i employed the breadboard along with resistors and caps to see if i could get a signal to the laptop's mic inout from the r8500's disc output that dsd would be able to successfully decode ... that did not help, so i went back to a simple patch cable.

the second and third samples i created, using the r8500, were done using an off-the-shelf rca-to-mono-phone patch between the radio and laptop.

listening to the .wav files with an audio player, you can hear that the audio from the icom is lower ... again, it seems that there's something electrically different between the discriminator output from these two radios as i can't get good decodes from the r8500 no matter what tweaking i do to the capture and/or mic boost levels using the alsamixer binary.

 

[slicerwizard]

thanks for the reply ... i'm not quite sure how to create a sample of dmr raw audio ... i'm guessing that dsd's -w option is not correct ... would it's -d option to create mbe data files do this?

I use DSD+ and just press 'R'. Don't know how well it works on other operating systems, but you're just making your life more difficult by using DSD.

i recall reading somewhere that with dmr a reported from dsd inlvl in the single digits should work well and with the pro-2006 i get error lines like these with a perfect decode:

Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=====
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:==
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=
Sync: -DMR mod: GFSK inlvl: 6% [SLOT0] slot1 VOICE e:=

As already noted, DSD+ decodes that file perfectly.

the second and third are from the icom r-8500's disc output ... i recalled reading somewhere that kicking the fm filter down from 12.0KHz to 5.5KHz solved a decode issue for another poster in the forums ... i *think* that was a p25 decode issue but i figured it couldn't hurt to try so i captured it for both fm filter settings.

The 5.5 kHz filtering is distorting the audio waveform. I wouldn't recommend it.

getting any audio out of dsd with the icom requires pumping either the mic input level or the mic boost level ... in either case, the lowest reported inlvl i can get from dsd is in the 9-10% range but the audio is gibberish and the error lines look like this:

Sync: -DMR mod: C4FM inlvl: 9% [SLOT0] slot1 VOICE e:====R====R=======R====R====R====RM====R=======R====R=======R=========R

Those Icom files have a nasty DC bias (the audio isn't centered around the zero line). Also, lower frequencies are attenuated, which is killing the decoding; you need to deliver flat audio to the program. And, as you've noted, the audio is very quiet. These are all signs of a bad tap job.

hi jim ... after my initial results using the r8500 were not successful, i employed the breadboard along with resistors and caps to see if i could get a signal to the laptop's mic inout from the r8500's disc output that dsd would be able to successfully decode ... that did not help, so i went back to a simple patch cable.

I'd say you need to go back to the breadboard. You should be shooting for something like this:

tap point --- 10uf --- 27k ---+---> shielded cable to PC mic input
                              |
                             4.7k
                              |
ground -----------------------+--->

the second and third samples i created, using the r8500, were done using an off-the-shelf rca-to-mono-phone patch between the radio and laptop.

Then why are the audio levels so low? What exactly is between the tap point in the radio and the PC's audio jack?

 

[psointu]

thank you for the analysis.

as for why the audio levels are so low? that's what i'm trying to sort out

as for what's in between the r8500's disc output and the laptop's mic in : Page Not Found

i'll try working with the breadboard as you recommended.

 

[psointu]

not sure why that link went bad when i copy/pasted it ... let's try it again:

RCA to 3.5mm Audio Cable

 

[slicerwizard]

Ok, I see we're dealing with the R8500's built in flat audio output; 200 mV should drive a Realtek mic input, but you have low audio...

And a DC offset. And distortion. Oy.

At a minimum, I'd try a DC blocker (10 uF cap); I'd also try providing a DC path to ground on the PC side of the cap, i.e. a 5-10k resistor.


If I ran into this problem, I'd just plug in one of my 99 cent USB audio dongles and use it. They look just like this guy:

5.1 Sound Track USB 2.0 Audio Controller - Free Shipping - DealExtreme

Don't know if that's an option under your OS.

 

[barrysmith1010]

8500 DSD

Hi
I know it's a while ago but did you get anywhere with the 8500 and DSD - I'm having the exact same issues with choppy decode, I've tried different PC's, different levels, caps and resistors all to no avail. Everything working fine with Realtek SDR USB stick but not with the Icom.
Cheers

barry

 

[boochk]

R8500 DMR

Did anyone ever get an R8500 discriminator output to decode DMR? I am using DSD+ 1.101 and even added the circuit described earlier in this post. It decodes a little but mostly errors in DSD. My AOR AR8600 can be used to feed DSD+ with no problems with any digital DMR, P25 and SLERS with no errors and perfect copy.

Thanks,
John