OpenGD77 Tier 2 now partially working

posted in: DMR, GD-77, Ham radio | 15

Just a quick update on the OpenGD77 Tier 2 functionality.

I’ve been working on the Tier 2 functionality for the last 2 weeks, and I now have Tier2 “Passive” functionality working, but it still has a some bugs.

Tier2 Passive is the technical name for duplex repeater operation, where the radio known as the Mobile Station or MS for short, gets its timeslot synchronisation from the Base Station (or BS for short), which is the repeater or duplex hotspot.

 

 

The main problem is the almost lack of documentation for the HR-C6000 DMR IC that is used in the GD-77. Although we have one data sheet, its the “Simple” version, and is written in Chinese and the diagrams don’t display correctly.
I’m sure that the full documentation pack would not just be this “Simple” data sheet and include application notes and also full documentation of the chip.

But since all I have is the “Simple” datasheet, everything has to be done by trial an error, as I have no way of knowing what the correct way to operate the chip.

For example. Currently in order to differentiate between TS1 and TS2, I seem to need to put the C6000 into a mode where its actually receiving constantly on both TS1 and TS2. I then only decode the audio from the selected Timeslot.
However the C6000 still sends the ID’s and TG’s for both Timeslots to the CPU, and at the moment I can’t see a way to tell which ID is for which Timeslot.
This isn’t a major problem if only one Timeslot is in use, however if both Timeslots on a repeater are in use, the radio displays the ID’s whenever anyone starts transmitting on either Timeslot, even in the middle of an existing transmission that you are listening to.

The other complication is that the MMDVM repeaters and MMDVM duplex hotspots, don’t seem to work in exactly the same way as the Motorola or other commercial repeaters.

I had the firmware working quite well with my duplex hotspot, but when I tried it on a Motorola DR3000 repeater and the audio faded out ever couple of seconds.

Then even when it worked fine on Tx and Rx for Motorola repeaters and duplex hotspots, there was an Rx issue on some MMDVM repeaters.

 

Anyway. I think the firmware is now almost at the point where I can make a Tier2 Beta version publicly available, if people are initially able to put up with things like the ID’s not displaying correctly.

15 Responses

  1. Mike
    |

    Great news Roger,
    many thanks and keep up your excellent community and development work!
    73 Mike

  2. ken
    |

    thank you

  3. Matt Fairbairn (M6SEU)
    |

    Hi Roger.
    Could it be that its in promiscuous mode as my 868 reacts the same if i put it in digimon with both slots enabled ?
    You must have square eyes by now… Magic work mate.
    Many Thanks Matt

  4. Roger Clark
    |

    The OpenGD77 firmware currently only filters on colour code, so is always in a form of Digital Monitor mode.

    I wonder if the IDs appearing for both TS on the Anytone is something they did by design or possibly an unavoidable consequence of listening on both slots.

    The problem I have, is that if I only listen on one slot, I can’t read the TS number from the C6000, and I need to know the TS number so that when the PTT is pressed and the radio needs to transmit, that it transmits on the correct TS as defined by the signal that is being constantly sent from the repeater.

    One potential workaround for this is to only listen on one slot most of the time, but when the PTT is pressed, switch to receiving on both TS and wait for at least 3 DMR frames to arrive, i.e wait until I see the TS number toggling from 1 to 2 and back again, ( or 2 to 1 and back to 2 again), to confirm that the radio is still in sync with the repeater, and then actually start transmitting

    This would result in up to 120mS delay between pressing the PTT and the transmission actually commencing, but it may not be noticeable and this is what the official firmware may be doing

  5. Matt Fairbairn (M6SEU)
    |

    May be right about the delay if you have 2 dmr radios 10 foot apart there is a noticeable delay when you key up until the other reacts and nearly a second on the transmit audio also if you key up and talk like on analog and start talking you usually find the prefix of the callsign is missing. Checked the Anytone and there is a delay when you key before it goes into tx might be longer than 120mS.

  6. Roger Clark
    |

    Matt

    Was that through a repeater or direct ?

    I think some radios buffer up some audio before they start transmitting, because potentially the audio sampling rate of the CPU does not match the DMR data rate being set by the server.

    In the GD-77 the CPU and the C6000 DMR DSP, seem both to run off the same clock frequency (12.288Mhz), so they should not get out of sync with each other.
    However in Tier2 Passive (repeater) operation, the data rate is governed by the repeater, because the C6000 synchronises to the signal it receives from the repeater.

    I don’t know for sure, but one potential reason some radios appear to start garbling the Rx audio after a fixed period of time e.g 2 minutes, could potentially be a buffer under-run, where the CPU is pushing out the decoded audio to its digital to analog converter at a slightly higher rate than the audio data is arriving from the repeater, so that eventually it run out of data in the buffer and strange things will then happen to the audio.

    The AMBE audio codec is very strange, and its hard to know from listening to garbled audio precisely what causes the effect that is being heard

  7. Matt Fairbairn (M6SEU)
    |

    Both direct and through repeater.
    Soon be Xmas Roger. Santa better be good to you for all this graft.
    Good work my man.
    Cheers Matt

  8. Roger Clark
    |

    Santa already ordered a RS-918 Chinese clone of the McHF open source HF QRP SDR transceiver .

    It’s got the magic ingredient of open source firmware 🙂

  9. Matt Fairbairn (M6SEU)
    |

    Just got my Xiegu G90 as well.
    Early Santa from good lady.

  10. Roger Clark
    |

    Xeigu G90 looks like a great radio. Shame there isn’t an open source firmware for it, otherwise I’d have definitely bought it in preference to the RS-918

  11. Jay Hill
    |

    I have 2x GD-77 here. I’m currently using one as a hotspot under Pi-Star, and it works great. Would it ever be feasible that 2x GD77 could be used to make a repeater? I only ask, as I have a license application currently ongoing for a DMR repeater (GB7GS), and it would be kinda nice to have the first OpenGD77 repeater!

    73 de G0FHM

  12. Roger Clark
    |

    It would not be possible to do this, because the crystal oscillators of the radios would need to be synchronised, and it’s not practical to do this via the USB serial connection from PiStar

  13. Lorenzo
    |

    Wow fantastic finally the Function key abandoned in the new modern radios but very useful to quickly enter the functions in the menu, congratulations I am trying the version and despite being still raw it is already usable and very practical good work and thanks Roger
    Lorenzo

  14. Ray C.
    |

    The OpenGD77 firmware is a definite improvement on the UI and tier-2 firmware to work well with my hotspot.

    I also recently picked up a RS-918. It’s a neat toy.

  15. Roger Clark
    |

    Hi Ray

    I’m just getting to grips with my RS-918, and so far have only used it on Rx, and it seems to work quite well.
    I’m now setting up the build environment so I can modify and rebuild the firmware if I want to.

Leave a Reply