[Fixed] Occasional transmission on the wrong TimeSlot

[IU4LEG]

Fantastic

[IU4LEG]

Very well,
Samuele and me, have make a qso for about an hour on the ir4uby repeater and on the ir6uda (the same repeaters of yesterday's test) and we can confirm that during all the time of the qso, the radios have NEVER TRANSMIT on the wrong timeslot. We used the OpenGD77-202001010900.sgl firmware.
I believe the problem is solved.
However, If I should notice strange behavior in the next few hours I will communicate it here on the thread.
For the moment, thanks, we will update in the next few hours.

[VK3KYY]

OK.

I agree, I think the fix from Colin in the latest version has probably fixed the problem for 99.9% of the operation

However, now i have captured some data from inside the GD-77 internal communication, I think we may need to modify the firmware to make some more improvements

[IU4LEG]

OK.

I agree, I think the fix from Colin in the latest version has probably fixed the problem for 99.9% of the operation

However, now i have captured some data from inside the GD-77 internal communication, I think we may need to modify the firmware to make some more improvements

Very good Roger

[IZ2EIB]

Hi guys!

I agree, I think the fix from Colin in the latest version has probably fixed the problem for 99.9% of the operation

However, now i have captured some data from inside the GD-77 internal communication, I think we may need to modify the firmware to make some more improvements

Hello Roger.
Me too I also agree.
I took the time to carry out the needed checks and sadly, just as you wrote, the problem is still there:

02012020_FW.png (2.67 KiB) Viewed 4098 times

02012019.jpg (111.18 KiB) Viewed 4098 times

What happened is that from time slot 1 by invoking disconnection (TG4000) the same was performed on time slot 2, which is rather unpleasant, but luckily at that time time slot 2 was free (as I have already had the opportunity to write here on the forum, it seems to me that this thing of engaging the wrong time slot occurs when the one not affected by the transmission is free, when it is busy it does not happen, this from what I could understand with my tests).
I hope that the interesting data between the MK22 MCU and the C6000 DMR chip you monitored between the C6000 IC and the CPU MCU will allow to permanently fix the problem and, why not?, improve the DMR functioning after understanding their meaning.
Thanks for your hard work Roger and let me also thank Colin who spent time and effort to provide a patch, thank you Colin!

73 best regards de Fabio IZ2EIB

[G4EML]

Fabio,

Thanks for the report.

One thing we can not see from your images is what the repeater was doing when you made the call.
It would be very helpful to know if the green light on the radio was on or off before you make a failed call.
This is because the method of finding the correct timeslot is different when the repeater has to be ‘woken up’.

You will only ever see a report of the wrong timeslot when that timeslot is not busy. That is just the way the repeater works. It will never receive a local transmission and a network transmission at the same time.

Some people may think that getting the timeslot correct should be easy, but a transmission from a radio on slot 1 is exactly the same as a transmission on slot2. The only difference is when it is sent. We are working to improve this but it is a complicated problem. Made more difficult by not having good data sheets for the chips involved.

Colin G4EML

[IZ2EIB]

Hi Colin, thank you for your reply!
From my side I am fully aware that this is a very complicated problem that is hard to fix and I am really very grateful to all of you for the effort you are making to try to fix it.
Thanks!
However, about the status of the repeater I cannot be 100% sure but I am pretty sure that in the occurrence the green LED on the GD-77 was off as well as the RSSI indicator.

73 best regards de Fabio IZ2EIB

[IU4LEG]

today,me and Samuele doing other tests on the usual 2 repeaters. We discovered that the problem occurs when the repeater receives a low signal.
let me explain better,
I was transmitting with an external antenna (yagi 6 elements oriented towards the repeater and the problem never happened). Looking at the dashboard, my signal was always stable, around -108db. Samuele was transmitting with the original gd77 and was walking, therefore his signal oscillated from a maximum of -109db to a minimum of -123db. When the signal was idle, it sometimes went to transmit on the wrong timeslot. when samuele's signal became stronger the problem was gone.
I hope I have explained it well
73 de iu4leg

[G4EML]

Hi Fabio,
Thanks for the report. That confirms what we already thought.
It is not the signal received by the repeater that causes it but the low signal received from the repeater by the GD77. When the signal is weak the signal that the repeater sends to identify the timeslot is sometimes decoded Incorrectly due to noise. The changes we are making are to try to improve this decoding.

Colin.

[IU4LEG]

Hi Colin, thanks for the explanation.
73