some dPMR questions

[krbonne]

Hi all,


First, I am new to the radioreference forum. I see there are three forums on digital voice, so I hope I chose the correct one


I have been doing some reading on DMR and dPMR and I have a couple of questions concerning dPMR:

- Looking on the well known shopping websites, there seams to be quite a few (mostly Chinese) radios out there that are advertised as being dPMR:
However, dPMR does allow quite a lot of freedom: tier 1 (TS 102 490) vs. tier 2 (TS 102 658), three different
modes in tier 2, 4 different voice codecs, choice of VOC bits.

Most of the shopping-websites provide very little technical information on there

Is there a centralised reference site (a wiki) for all the different dPMR radios out there and exactly what specifications they adhere to?

- What exactly is the technical difference between tier1 and the tier2/mode 1 (peer-to-peer)? (appart from the frequency-range og 466.1-466.2 Mhz and power-levels).

If you would manage to get tier1 radio and a tier2/mode1 radio on the same frequency, would they be able to communicatie with eachother?

(Yes, I did start browsing throu the ETSI specs).


- What is the exact difference between tier2/mode1 and tier2/mode2. I know the websites say "mode1 = peer-to-peer, mode2 = repeater", but what exactly is the difference on RF stream-level?

- I know there is DSD that should be able to decoce NEXEDGE (which is dPMR), but is there an full open-source stack of dPMR?

It would be nice to have the basic stack to write our own dPMR repeater for ham-radio use. Are there already projects for this?
Anybody interested in doing this?

- As dPMR has a VOC-option "11" (manufacter specific), it might be interesting if we cannot use this to create a version of dPMR for ham-radio using the codec2 voice-codec.
Anybody interested in this?


- I have been looking at the si4668 chip from silicon labs. This chip supports a number of modulation scemes, one of them being 4GFSK.
(see https://www.silabs.com/products/wireless/EZRadioPRO/Pages/Si446x.aspx)

Does anybody know what is exactly the difference / link between 4GFSK and C4FM?

Could this chip be used to build a low-power C4FM receiver/transmitter, something equivalent to the low-power D-STAR hotspot nodes?



Many thanks in advance.

73
kristoff - ON1ARF

 

[krbonne]

channel code

Hi,

A small follow-up in my own message.

Concering question 2: compatibility between tier 1 (dPMR446) and tier2/mode 1.
I have started reading the ETSI specs about this.

One of the things that I did notice is the "channel code". That is a code (based on a value between 0 and 63) part of the header that identifiies a frequency.
TS102 659 (tier 2 dPMR) section 6.1.5.1. specifies that the channel code is calculated based on the frequency.
TS 102 490 (dPMR446) section 6.1.5. provides a list of channel-codes for the 16 channels between 446.1 and 446.2 Mhz.

If I apply the formula of TS102659 to the frequencies in the table of dPMR446, I get other values. E.g. the first channel in that list (446.103125 Mhz) results in channel code 16, while the table of TS102 480 defines channel code 1 for that frequency.


I also found that a stream received with a different channel-code then expected should be concidered to be interference and be ignored (see section 10.1 of TS102490 and 12.1.1 of TS 102 658).


So, can I conclude that tier1 and tier2 are by nature not compatbile, even if they define a stream-format that is almost identifical?


73
kristoff - ON1ARF

 

[nd5y]

Welcome to the dPMR Association, Narrowband 6.25Khz FDMA Digital

 

[krbonne]

Tom,

I once asked them for more information on an open-source implementation of dPMR for ham-radio use, and the reply I got was "do not use dPMR for ham-radio, that is what D-STAR is for".

Concidering the fact that one of the cofounders of the dPMR MOU forum is icom, that is probably quite an answer that was to be expected.

Anycase, I do want to look at the possibility to implement an open-source implementation of dPMR for at least tier1 and modes 1 and 2 of tier2. The first step would be to look at DSD and see if we cannot build a high-quality dPMR stream-analyser based on that.
This would allow us to get an idea of the actual signals generated by all these Chinese dPMR manufacteres and verify how well they follow the official ETSI specification


So I am looking for people who are willing to help out:
- to test audio-streams of all the dPMR radios out there
- coders of the dPMR stack
- to implement a good C4FM modem for source-card interface / microcontroller (somebody with good DSP knowledge)
- to help convert the protocol stack into a usefull API for the upper-level applications
- to help document everything

Who is interested? This is your change to really change ham-radio and get your name of the hall of fame.


73
kristoff - ON1ARF

 

[Quasar_A]

Hi,
One of the things that I did notice is the "channel code". That is a code (based on a value between 0 and 63) part of the header that identifiies a frequency.
TS102 659 (tier 2 dPMR) section 6.1.5.1. specifies that the channel code is calculated based on the frequency.
TS 102 490 (dPMR446) section 6.1.5. provides a list of channel-codes for the 16 channels between 446.1 and 446.2 Mhz.

Hello, yes this is one difference, second is "communication format" field. For dPMR446 it is only have two value: "Call ALL" and Peer-to-Peer, other values is reserved.

Modulation, scrambling, FEC, frame format (Superfrmae, Header, End frame) is same.

Chinese dPMR radios uses TS102 658. For dPMR protocol processing it is uses SCT3252 IC.

On my work, currently I am developing a digital controller for APCO P25 Phase 1 (Conv) and dPMR radios (I am embedded programmer). APCO P25 works good, dPMR too, but I have one problem with dPMR, I can't find description for chinese dPMR voice payload (FEC, and voice codec description), and as result, my controller works in dPMR mode without FEC correction in voice payload, only in headers described in TS102 658. Controller receives data, decodes dPMR fields (only dPMR), correct all errors on dPMR fields, sends it to IP for multisite operation, retransmits it locally, for APCO it correct errors not only for APCO frame fields but in voice payload to.

This would allow us to get an idea of the actual signals generated by all these Chinese dPMR manufacteres and verify how well they follow the official ETSI specification

Generally follows (Nanfone, Kydera and others on SCT chips), but sometimes it is have bugs. Ex. in Nanfone NF-9909 I found problem in symbol deviation, at frequencies higher than 440 MHz it had deviation below the standard, at frequencies below 440 MHz symbol deviation had Ok.

 

[krbonne]

Hi Quasar,


First of all. Sorry for the late reply. (it's been a busy couple of days, so I prefered to actually do some coding instead of talk about it )

Thank you for your very interesting reply. Nice to be able to talk to somebody who is actually doing the work for a dPMR radio-stack

Hello, yes this is one difference, second is "communication format" field. For dPMR446 it is only have two value: "Call ALL" and Peer-to-Peer, other values is reserved.

Makes sence!

For dPMR protocol processing it is uses SCT3252 IC.

Looks like a very versatile chip.

On my work, currently I am developing a digital controller for APCO P25 Phase 1 (Conv) and dPMR radios (I am embedded programmer). APCO P25 works good, dPMR too, but I have one problem with dPMR, I can't find description for chinese dPMR voice payload (FEC, and voice codec description), and as result, my controller works in dPMR mode without FEC correction in voice payload, only in headers described in TS102 658. Controller receives data, decodes dPMR fields (only dPMR), correct all errors on dPMR fields, sends it to IP for multisite operation, retransmits it locally, for APCO it correct errors not only for APCO frame fields but in voice payload to.

OK, makes sence.

I guess that the main issue is the scrambler. Concerning interleaving and FEC are there not that much options on how to do this so wtiht a trial and error, that might be found.


Well, you can do correction and regeneration of dPMR if it uses plain AMBE2+, so perhaps that would be a driver for these companies: using a competing codec, the repeaters do error-correction (i.e. provide a better service) so it's in your advantage to provide at least some information (or a closed-source library) so this can also be done for your codec.

Then again, I doubt that this will make any difference.


Kristoff - ON1ARF

 

[princo]

Kristoff,

I have been working on a similar project myself. First off, dPMR is in theory, a great standard. However, it has been quite adulterated because of the freedom to use a non-conventional (i.e. unfamiliar) vocoder. Every dPMR radio I have seen for purchase thus far uses a very obscure vocoder that auctus.cn calls "Tsinghua" Chinese vendors on aliexpress and other sites, have been calling this "ambe+2c" they are in no way compatible with the actual ambe from dvsi. I have posts on this forum detailing it further.

As for the transceiver, the si4463 or si4464 might be able to work. However, the adf-7021 might be better suited. (The adf-7021 is already used in the d-star hotspots) A module with the adf-7021 is already available on aliexpress and is only around $20 U.S. I plan to buy one soon. I have a cc1101 and si4463 on the way already. I have read some anecdotal evidence that the cc1101 was able to decode parts of p25 frames.

As for Codec 2, I suggest we see about creating a standard air interface that we can make as open as possible. Some might be able to argue that the chinese dPMR radios violate U.S. ham rules because the vocoder is almost like a state secret. I should note that I have seen two or three dPMR radios from china that use RALCWI as a vocoder. This is actually quite a bit better because at least RALCWI is available as an ic from CML microcircuits.

What I have been working on is a sort of hotspot that just demodulates data from the digital radios and passes it over the internet (intact) and then retransmits it on the other side. Any help would be appreciated.

Hope this helps some,

John Prince
Prince Labs

 

[krbonne]

Hi John,

I have been working on a similar project myself. First off, dPMR is in theory, a great standard. However, it has been quite adulterated because of the freedom to use a non-conventional (i.e. unfamiliar) vocoder. Every dPMR radio I have seen for purchase thus far uses a very obscure vocoder that auctus.cn calls "Tsinghua" Chinese vendors on aliexpress and other sites, have been calling this "ambe+2c" they are in no way compatible with the actual ambe from dvsi. I have posts on this forum detailing it further.

I had a conversation with a fellow ham here and he had the same remark. Indeed, the free choice of codec is both an opportunity and a problem. For ham-radio, it is probably not a good thing, not from a "being able to talk to eachother" point of view.
On the other hand, look at the "digital mode" area on HF. Almost none of these modes is compatible and it is exactly the lack of "needing to be compatible" that has sparked the "furthering the science" attitude overthere.
If you want a mode and just be able to talk to as much people, get a DSTAR or DMR radio.


For the repeater, my impression is that it does notreally matter.
The first D-STAR repeaters did not interprete the audio neither. (That only changed when the ircddb guys wrote the code for DTMF decoding which also revealed how the AMBE packets where formatted. Voice-announcement also came later).
As far as I see it (do correct me if I am wrong) a dPMR repeater should be able to relay any voice-stream no matter what codec. It is then for the local users of that repeater to agree on what codec to use (i.e. buy a particular radio)


One of the main issues will -I think- be to find a way to DETERMINE the codec of an incoming stream. We do not need to be able to decode it. If we can see what codec it is, a repeater should be able to direct the stream to a particular conference, dedicated for that codec; and from there on, to repeaters also having mainly people using radios with that particular codec.

For that reason, my first project is to try to develop a dPMR analyser so it is possible to create lists with what radios are compatible.

As for the transceiver, the si4463 or si4464 might be able to work. However, the adf-7021 might be better suited. (The adf-7021 is already used in the d-star hotspots) A module with the adf-7021 is already available on aliexpress and is only around $20 U.S. I plan to buy one soon. I have a cc1101 and si4463 on the way already. I have read some anecdotal evidence that the cc1101 was able to decode parts of p25 frames.

Using the si-chips was just an idea I had because I am interested in some other si-chips for another application (meshed network in UHF)

I am currently not 100 % sure to what degree C4FM and 4GFSK are compatible. Based on how I understand the protocols now, they should be close, which could be enough from a "hotspot" point of view.


I will surely have a look at these other chipsets. Great info. Thanks!

As for Codec 2, I suggest we see about creating a standard air interface that we can make as open as possible. Some might be able to argue that the chinese dPMR radios violate U.S. ham rules because the vocoder is almost like a state secret. I should note that I have seen two or three dPMR radios from china that use RALCWI as a vocoder. This is actually quite a bit better because at least RALCWI is available as an ic from CML microcircuits.

I am in contact with David Rowe and some other people on codec2. David is working on a special VHF/UHF modem, but it will require a SDR-type radio. The project to see if we cannot get codec2 into dPMR Is a sideproject as dPMR Is a lot simpeler and therefor a lot easier to implement. Both projects have their place.
After all, the goal is allow people to experiment with and learn more about digital voice and digital communication. DV has been to much about "users".
As I mentioned in another message, I am interested in trying to do Es DX on 29 Mhz using DV. dPMR allows us to experiment with that, e.g. by using intra-frame interleaving (i.e. spread out the voice over more frames) to help deal with fading.


Concerning codec2:
There are now a number of projects for SDR-based ham-radio. Concidering a chip like the STM32F4 (which is already found in radios like the CS700 DMR-radio) can do codec2 encoding/decoding in real time, this does open possibilities.

Anycase, it's a nice expermiment, and that is what the hobby is about, no?


As for the issue of the non-documented codecs and regulation, I leave that to the lawers. I'm just interested in technology and learning.

What I have been working on is a sort of hotspot that just demodulates data from the digital radios and passes it over the internet (intact) and then retransmits it on the other side. Any help would be appreciated.
Hope this helps some,
John Prince
Prince Labs

If the code is open-source, I surely would be interested in having a look. Is it posted somewhere?

I am in contact with a couple of hams in the Netherland who are also experimenting with dPMR so if you have code to set up a bridge between two repeaters, that would surely be interesting for them!


73
kristoff - ON1ARF

 

[Quasar_A]

Kristoff,
I have been working on a similar project myself. First off, dPMR is in theory, a great standard. However, it has been quite adulterated because of the freedom to use a non-conventional (i.e. unfamiliar) vocoder. Every dPMR radio I have seen for purchase thus far uses a very obscure vocoder that auctus.cn calls "Tsinghua" Chinese vendors on aliexpress and other sites, have been calling this "ambe+2c" they are in no way compatible with the actual ambe from dvsi. I have posts on this forum detailing it further.

As noted, for repeater, it is not important, what type of codec. Scrambling and FEC type is more important question, and this is question is not covered in standard (it is worth noting in DMR too, but DMR no have problems like dPMR).

As for the transceiver, the si4463 or si4464 might be able to work. However, the adf-7021 might be better suited. (The adf-7021 is already used in the d-star hotspots) A module with the adf-7021 is already available on aliexpress and is only around $20 U.S. I plan to buy one soon. I have a cc1101 and si4463 on the way already.

I think, all solutions on IC like ADF have one problem, *it is impossible to build repeater for mixed operation (Digital, Analog with auto switch).

Motorola have some patents for DMR and dPMR. Does this not create problems in the future, in the propagation of open dPMR in the internet?

https://www.motorolasolutions.com/c...xw/static_files/dpmr_essential_properties.pdf

 

[princo]

Kristoff,

As far as my project goes, I haven't started any code because I have yet to receive any modules to play with. I believe the si4463 show some promise but I could be wrong. If you get the chance, you should read the datasheet and maybe play around with their wireless development suite. The only issue I've seen is that the packet handler is a bit tricky. I was thinking for DMR I could try tricking the packet handler into thinking that the frame sync is a preamble, so it would transmit bursts of data like TDMA calls for. I could be wrong though. But it would certainly be to our advantage if we could use the si4463 since the modules are extremely cheap. I.e. $3-$6 for a si4463 vs. $20-$50 for the adf-7021. And yes, unfortunately, these ic's don't do anything analog, I don't honestly see that as an issue for a few reasons, I'm not a ham so this is purely academic for me, and hams (or other radio users) don't pay for digital radios to use analog all the time. What radios are you working with? I'm currently using two Zastone 9908's for dPMR. Hopefully I'll have a DMR radio to play with soon.

As far as the modulation in Fusion, DMR, dPMR and other 4fsk radios, the only difference I've seen is their deviation. Fusion has the widest inner and outer deviation I've seen yet.

Kristoff, is there a better way to get in touch with you so we could maybe swap notes on projects?

 

[kb9mwr]

Here are some of the ESTI standards:

http://www.qsl.net/kb9mwr/projects/dv/dpmr/dpmr.html

 

[Your_account]

I have some theoretical experience with dpmr!!
The problem is dpmr is a sshhtt.....

 

[Hoseman292]

I am in the D.C. area and interested. So far I have analog programming in my radio. I have not attempted to create a digital profile.
I own a VHF Zastone DP860 DPMR handheld transceiver.
Tim

Tom,

I once asked them for more information on an open-source implementation of dPMR for ham-radio use, and the reply I got was "do not use dPMR for ham-radio, that is what D-STAR is for".

Concidering the fact that one of the cofounders of the dPMR MOU forum is icom, that is probably quite an answer that was to be expected.

Anycase, I do want to look at the possibility to implement an open-source implementation of dPMR for at least tier1 and modes 1 and 2 of tier2. The first step would be to look at DSD and see if we cannot build a high-quality dPMR stream-analyser based on that.
This would allow us to get an idea of the actual signals generated by all these Chinese dPMR manufacteres and verify how well they follow the official ETSI specification


So I am looking for people who are willing to help out:
- to test audio-streams of all the dPMR radios out there
- coders of the dPMR stack
- to implement a good C4FM modem for source-card interface / microcontroller (somebody with good DSP knowledge)
- to help convert the protocol stack into a usefull API for the upper-level applications
- to help document everything

Who is interested? This is your change to really change ham-radio and get your name of the hall of fame.


73
kristoff - ON1ARF

 

[Your_account]

Hey
What i know is PMR in the US not alloew the use some other System?!