identify exact DMR frequency

[idarlund]

Hi,

In Norway we have an operator called Sikringsradioen (Sikringsradioen: VHF-radioene med lengst rekkevidde og egen nødknapp) which operates a PMR network with base-stations in the VHF frequency range. The analog frequencies are well documented; http://frekvenser.no/Jaktradio.html (the channels with "Sikringsradio" in the "Navn" field).

The digital frequencies are however not documented. These channels are also only radio to radio frequencies without repeaters (at the moment). I have a radio with these frequencies programmed, but since the device is locked I'm unable to read what frequencies they are sending on. So I used an rtl-sdr with sdrsharp and sdrpp to figure out the frequencies it uses. This is done only by eyeballing the sdr interface and are not 100% accurate. Here's a screenshot:



And here's the list of frequencies I've made:
01: 143.4315
02: 143.4190
03: 138.9315
04: 138.9190
05: 143.5064
06: 143.4945
07: 139.0065
08: 138.9940
09: 143.2064
10: 143.1939
11: 138.7060
12: 138.6940

I've tested that they actually work with an other programmable radio (with group call id 100000, color 1 and TS 1), but as I mentioned these frequencies are not perfect which might have an effect when starting to use these over longer distances.

How can I figure out the exact frequency a DMR radio are sending on?

 

[Ubbe]

Monitor another frequency that are known and enter its frequency. I'm not familiar with SDR++ but go to settings and calibrate the ppm offset until you have your mark in the middle, the same way you did with the 138Mhz frequency.

You get the wrong frequencies as the receiver you use have not been frequency calibrated in the program. All my RTL-SDR dongles seem to need -1 or -2 ppm adjustment.

All your frequencies have to be reduced by 0.0065MHz and that equals -5ppm

01: 143.4250
02: 143.4125
03: 138.9250
04: 138.9125
05: 143.5000
06: 143.4875
07: 139.0000
08: 138.9875
09: 143.2000
10: 143.1875
11: 138.7000
12: 138.6875

/Ubbe

 

[idarlund]

Monitor another frequency that are known and enter its frequency. I'm not familiar with SDR++ but go to settings and calibrate the ppm offset until you have your mark in the middle, the same way you did with the 138Mhz frequency.

First of all; thanks for replying! I will definitely check out a known frequency and try to calibrate my SDR!

You get the wrong frequencies as the receiver you use have not been frequency calibrated in the program. All my RTL-SDR dongles seem to need -1 or -2 ppm adjustment.

All your frequencies have to be reduced by 0.0065MHz and that equals -5ppm

How do you know that my readings are 0.0065MHz off? Do you know the actual frequencies I'm trying to figure out?

01: 143.4250
02: 143.4125
03: 138.9250
04: 138.9125
05: 143.5000
06: 143.4875
07: 139.0000
08: 138.9875
09: 143.2000
10: 143.1875
11: 138.7000
12: 138.6875

Are these the actual "Sikringsradio" DMR frequencies or did you just calculate them based on my findings in SDR++?

How can I figure out the exact frequency a DMR radio are sending on?

Either way, I believe my original question still stands. Is there a proper way to determine exactly what frequency a device is sending on or is "eyeballing" the marker the only way?

 

[Ubbe]

All your measured frequencies needs to be reduced by between 6-7KHz to make them match a 12,5KHz stepsize. But it could also be your radio that are off in frequency calibration.

I usually see those 5ppm errors with Nooelec and low class dongles.

The proper way to check your radio's frequency are to use a good quality frequency counter. Most of them, at a €200-€300 second hand cost, have a 1ppm tolerance. That's parts per million. Mega are also million. So at 1MHz it could be 1Hz off. At 100MHz it could be 100Hz off. But those are usually not sensitive enough to measure a transmitter far away and will also be interfered from other signals in the air.

You can calibrate a receiver to a known accurate frequency, a transmitter that has GPS controlled frequency standard. I believe most DMR transmitters are used in that way. My AOR 8200 are accurate to 1ppm and if I set it to operate in AFC mode the display will show the exact frequency of a transmitter. If you only wish to check what frequency a transmitter use that are supposed to follow a bandplan, and most transmitters has to, then a scanner can be used and then compare with the stipulated bandplan it has to follow to get the closest matching frequency.

/Ubbe

 

[maus92]

The frequencies that Ubbe derived look good to me, but I don't know the regulations in Norway wrt channel spacing. I do know the ITU spectrum regs to some extent, and in Norway, 138.0-143.6 MHz is a band allocated for Land Mobile use. I'm sure there is a spectrum management / licensing agency in Norway, so hopefully they publish the exact frequencies that currently licensed for comparison, like what the FCC does here in the US (for non-federal entities.)

 

[maus92]

Adding:

This appears to be a site of interest:

Frekvensportalen

Scroll down to the 138 MHz entry, then follow the Permissions link on the page:

 

[Ubbe]

Good find. Sikringsradoen's permit says:

138,6250 MHz
138,7000 MHz
138,9250 MHz
138,9750 MHz*
139,0000 MHz
139,0250 MHz
139,0500 MHz
139,0750 MHz
139,1250 MHz
139,1750 MHz
139,2000 MHz
139,2250 MHz
139,2500 MHz
139,2750 MHz
139,3000 MHz
139,3250 MHz
139,3500 MHz
139,3750 MHz
143,1250 MHz
143,2000 MHz
143,4250 MHz
143,4750 MHz*
143,5000 MHz
143,5250 MHz
143,5500 MHz
143,5750 MHz
143,6250 MHz
143,6750 MHz
143,7000 MHz
143,7250 MHz
143,7500 MHz
143,7750 MHz
143,8000 MHz
143,8250 MHz
143,8500 MHz
143,8750 MHz
458,1500 MHz
468,1500 MHz *)

Nye frekvenser tildelt Sikringsradioen AS ved denne tillatelsen

/Ubbe

 

[idarlund]

Yeah, I found those frequencies after sending NKOM an email. I got directed to this PDF:

https://frekvens.nkom.no/files/frekvensportalen/1004134%20-%201805583-2%20Spektrumstillatelse%20nr%202924639_23_1.pdf

These are the frequencies that "Sikringsradioen AS" are allowed to use and if you look at 5.1 the service and technology are neutral; meaning they are allowed to use any type of PMR tech. After going down this rabbit hole I scetched up an excel sheet to try to figure out the frequencies. Now it looks like this:



As you can see I plotted and color coded the frequencies suggested by Ubbe which (of course) matched up. But here's the catch... If I program an actual radio to use these frequencies, the first radio aint getting anything.

Also in my email conversation with NKOM, they where actually surprised that Sikringsradioen AS had started using DMR as they've normally used the analog tech. They respondeds with "Eg vil tru at forhandlaren har programmert DMR ved ein feil. Eg har ikkje høyrt at Sikringsradioen har byrja med DMR på VHF. " which translates to something like "I would think that the dealer has programmed the DMR channels by a mistake. I have not heard that Sikringsradion has stared using DMR on the VHF band."

Also; I checked my SDR and it seems acurate. When testing on known frequencies it seems quite spot on. But then again.. this is only by eyeballing the graph. Would be great to be able to count or get the SDR to tell me what the center frequency is.

I believe to be able to figure this out exactly I would have to program the other radio I have with .001-5 mhz in steps and try to figure out the edge cases it works, write it down and then see what the center of those two edge cases are.

Any other suggestions?

 

[Ubbe]

01: 143.4315
02: 143.4190

I've tested that they actually work with an other programmable radio (with group call id 100000, color 1 and TS 1), but as I mentioned these frequencies are not perfect which might have an effect when starting to use these over longer distances.

As you can see I plotted and color coded the frequencies suggested by Ubbe which (of course) matched up. But here's the catch... If I program an actual radio to use these frequencies, the first radio aint getting anything.

Then your SDR receiver are actually correct, so round off to nearest 6,25 raster. 143.43125 and 143.41875 and so on and it will start to work.
It's impossible to be more accurate when looking at a spectrum display, but try and set the SDR++ to have one or two PPM more negative value in the PPM correction.

According to the analog channel names (from your radio?) they are planning to use repeaters that transmit and receive using a 5MHz split. I would think that the channels named INN are what the repeater are using as a receive frequency and can be used as a simplex channel. The DMR channels are 12,5KHz wide and frequencies are selected to not interfere with the analog ones.

01: 143.43125
02: 143.41875
03: 138.93125
04: 138.91875
05: 143.50625
06: 143.49375
07: 139.00625
08: 138.99375
09: 143.20625
10: 143.19375
11: 138.70625
12: 138.69375


/Ubbe

 

[idarlund]

Then your SDR receiver are actually correct, so round off to nearest 6,25 raster. 143.43125 and 143.41875 and so on and it will start to work.

Yeah, I've got several SDRs. This screenshot is from an hackrf if I remember correctly

It's impossible to be more accurate when looking at a spectrum display, but try and set the SDR++ to have one or two PPM more negative value in the PPM correction.

Okay. I will check it out. Do you know if there's any software to frequency "count" by using an SDR? Or is this only possible with specialized equipment you linked to earlier?

According to the analog channel names (from your radio?) they are planning to use repeaters that transmit and receive using a 5MHz split. I would think that the channels named INN are what the repeater are using as a receive frequency and can be used as a simplex channel. The DMR channels are 12,5KHz wide and frequencies are selected to not interfere with the analog ones.

Yes, they are from my pre-programmed radio. And these frequencies are allready well documented. And you are correct that some of these are using base stations. These are also well documented on Våre basestasjoner - Sikringsradioen AS
The only reason I also documented them in the same excel sheet was just to see in the table where they were according to the DMR channels.

01: 143.43125
02: 143.41875
03: 138.93125
04: 138.91875
05: 143.50625
06: 143.49375
07: 139.00625
08: 138.99375
09: 143.20625
10: 143.19375
11: 138.70625
12: 138.69375

Thanks! I will test them out

 

[idarlund]

01: 143.43125
02: 143.41875
03: 138.93125
04: 138.91875
05: 143.50625
06: 143.49375
07: 139.00625
08: 138.99375
09: 143.20625
10: 143.19375
11: 138.70625
12: 138.69375

Thanks! I will test them out

Tested these freqs and they all worked. Hopefully these are the exact center frequencies

Thanks a lot for your help @Ubbe !

 

[Ubbe]

Do you know if there's any software to frequency "count" by using an SDR? Or is this only possible with specialized equipment you linked to earlier?

If you can find a AFC function in a program and you calibrate your receiver then that would probably work.

In an analog FM discriminator you have a DC voltage that show how much off frequency you are but in SDR receivers you don't have that type of discriminator and it has be done in a different way and probably using your eye on a spectrum analyser window will anyhow be more accurate.

If you have a scanner with a FM discriminator that can be programmed to all possible step sizes like 3.125KHz and 5KHz then you could connect that DC voltage thru a buffer circuit to an analog dial and calibrate to show the exact frequency when the needle are at the center. Bill Cheek had a schematic for it to add it to the RadioShack scanners Pro2005/2006 and I also had that function in an old Bang&Olufsen FM broadcast receiver.

/Ubbe

 

[idarlund]

I showed "sikringsradioen" this thread and they confirmed that the frequencies was correct.