Release 3rd May

User avatar
kd2lh
Posts: 291
Joined: Mon Dec 02, 2019 2:44 pm

Re: Release 3rd May

Post by kd2lh » Mon May 04, 2020 5:20 pm

This may relate to the implementation of "Automatic Frequency Control" which is used to dynamically shift the center frequency of a receiver to the correct place for 4FSK / TDMA reception.

Most MMDVM Hat style radio interfaces (like the one used in the JumboSpot and by others) don't implement the AFC feature. This is why RXOffset is used to shift a hotspot to try and match the radio(s) used with it. The radio chip is the software defined ADF7021, and it does have an AFC feature available, but not implemented on the boards controlled by the STM32 / Arduino processor.

The handheld radios generally have AFC operating on their Receive side. The use of a GD77 as a hotspot for DMR operation (only) gets the benefit of AFC on it's receive side.

I'm not completely clear on why the JumboSpot doesn't implement AFC, but it could well be that the processor chip is completely occupied dealing with DMR protocol at higher levels. Someone more familiar with the code can comment on that.

SO5AJG
Posts: 60
Joined: Sat Nov 16, 2019 3:44 pm

Re: Release 3rd May

Post by SO5AJG » Mon May 04, 2020 5:41 pm

Thanks for the clarification. Certainly, such a system is very bothersome, especially as many radios use one hotspot.
Janusz - SO5AJG - Warsaw

G4EML
Posts: 218
Joined: Sat Nov 16, 2019 10:01 am

Re: Release 3rd May

Post by G4EML » Mon May 04, 2020 6:49 pm

The ADF7021 AFC only works with constant carrier transmissions. It is not able to cope with TDMA pulsing transmissions as used by DMR. The bursts of noise in the unused timeslot cause it to loose lock.

User avatar
kd2lh
Posts: 291
Joined: Mon Dec 02, 2019 2:44 pm

Re: Release 3rd May

Post by kd2lh » Mon May 04, 2020 8:34 pm

Is AFC functional and in use in MMDVM with the MMDVMHat boards on the other protocols (YSF, etc...)?

Marc

G4EML
Posts: 218
Joined: Sat Nov 16, 2019 10:01 am

Re: Release 3rd May

Post by G4EML » Mon May 04, 2020 9:09 pm

I have no idea. However DStar and YSF are not TDMA modes so it may be.

Even without AFC a constant transmission is easier to demodulate as the detector can filter and average out the signal to determine a good threshold point. DMR is difficult to decode because the noise bursts in the unused timeslot upset any averaging circuit. Likewise any coupling capacitors will distort a TDMA waveform. This is a weak point of the MMDVM approach. DMR Chips such as the HRC-6000 in the GD-77 have specially designed decoders that do a much better job of recovering the data.

IU3MCP Giovanni
Posts: 1
Joined: Mon May 04, 2020 8:48 am

Re: Release 3rd May

Post by IU3MCP Giovanni » Tue May 05, 2020 12:37 pm

Thanks !!!


'73 de IU3MCP Giovanni ;)

User avatar
f6fzo
Posts: 30
Joined: Sat Jan 04, 2020 7:28 am

Re: Release 3rd May

Post by f6fzo » Tue May 05, 2020 2:37 pm

Merci :mrgreen:
F6FZO Serge 43 Nord

SO5AJG
Posts: 60
Joined: Sat Nov 16, 2019 3:44 pm

Re: Release 3rd May

Post by SO5AJG » Tue May 05, 2020 4:28 pm

There are often such situations (about which I wrote earlier) that there is such a trembling sound as in the attached recording. This appears on both GD77 and RD5R. From what I observed, this happens just like I listen to the correspondents for longer, but I did not notice it occur when I am a participant in the conversation.
https://drive.google.com/file/d/1yOstI0 ... sp=sharing
Janusz - SO5AJG - Warsaw

User avatar
EB3AM
Posts: 179
Joined: Fri Jan 24, 2020 1:40 pm
Location: Catalonia, not Spain
Contact:

Re: Release 3rd May

Post by EB3AM » Tue May 05, 2020 6:05 pm

SO5AJG wrote:
Tue May 05, 2020 4:28 pm
There are often such situations (about which I wrote earlier) that there is such a trembling sound as in the attached recording. This appears on both GD77 and RD5R. From what I observed, this happens just like I listen to the correspondents for longer, but I did not notice it occur when I am a participant in the conversation.
https://drive.google.com/file/d/1yOstI0 ... sp=sharing
Yes, we have been observing this phenomenon in audio from time to time for a few days (or months).
If when you listen to this audio, you turn off the phone and turn it on, the audio sounds right again. We have also been able to observe that with another receiver (md380) the audio is not distorted at the same time as the Opengd77 firmware does.
I thought it might be a DMR timeout issue, which is now at 13, and it's an odd number, and I did some testing with that number set to 14, but there's no difference. :mrgreen:
What seems clear is that it is a firmware issue. It is also not very annoying, as it does not last too long and tends to recover well.
Jordi, EB3AM, Talarn JN02ke, Catalonia.

SO5AJG
Posts: 60
Joined: Sat Nov 16, 2019 3:44 pm

Re: Release 3rd May

Post by SO5AJG » Wed May 06, 2020 6:48 am

This topic has been going on for some time. The recording was made by a colleague who recorded my conversation and this effect also appears on his radio GD77 from time to time.
Janusz - SO5AJG - Warsaw

Post Reply