Page 4 of 8
Re: Update 29th Jan
Posted: Fri Jan 31, 2020 4:52 am
by Ik0nwg
excellent job !!!! as always!!! thank you
Sal
Re: Update 29th Jan
Posted: Fri Jan 31, 2020 5:05 am
by VK3KYY
Guys
With the change to the search order. I don't think its working completely correctly.
I need to do some more testing, but sometimes where it should be using the DMR ID data, it seems to be using TA
Re: Update 29th Jan
Posted: Fri Jan 31, 2020 6:47 am
by DO4CZ
VK3KYY wrote: ↑Fri Jan 31, 2020 3:19 am
Also, I've added a 50mW and 5W++ power setting,
Wonderful news!
Thank you very much, Roger!
CWID and Firmware lockup Re: Update 29th Jan
Posted: Fri Jan 31, 2020 2:03 pm
by kd2lh
Following up on the problems I saw, thinking about it it occurred to me that there might be a problem relating to CWID, and mode switching the radio while receiving incoming voice packets from Brandmeister (or other VOIP infrastructures).
The failure I have on video happened during a brief interruption in incoming voice from a talkgroup (ours) and might have coincided with the ID being sent in FM Morse. If the mode switching isn't clean, and there are pending voice packets PiStar is trying to push to the phone, could this be the cause of the lockup?
I don't know the logic, so can only make a guess based on my use with CWID active, where others probably have not set it.
If you're testing the new 1/22 or later firmware, please try turning CWID on (it's in the MMDVMHOST definitions under Expert mode of Configuration). Set "CWID" to one, and set a reasonable (5?) minute interval. Let's see if anyone else experiences the lockups or corruption.
I've been running the .082 1/22/2020 version clean for a day and a half, and will not load Roger's latest 1.29 version.
Re: CWID and Firmware lockup Re: Update 29th Jan
Posted: Fri Jan 31, 2020 3:05 pm
by F1RMB
Hi Marc,
kd2lh wrote: ↑Fri Jan 31, 2020 2:03 pm
Following up on the problems I saw, thinking about it it occurred to me that there might be a problem relating to CWID, and mode switching the radio while receiving incoming voice packets from Brandmeister (or other VOIP infrastructures).
The failure I have on video happened during a brief interruption in incoming voice from a talkgroup (ours) and might have coincided with the ID being sent in FM Morse. If the mode switching isn't clean, and there are pending voice packets PiStar is trying to push to the phone, could this be the cause of the lockup?
I don't know the logic, so can only make a guess based on my use with CWID active, where others probably have not set it.
As it is done in the actual code, if some frames are sent by MMDVMHost while CW is keying, they are just buffered, and handled once the CW keying stops.
About the lock up, yesterday I stumbled on a possible problem with MMDVMHost communication (in the HS -> MMDVMHost direction), that one could be the problem here.
kd2lh wrote: ↑Fri Jan 31, 2020 2:03 pm
If you're testing the new 1/22 or later firmware, please try turning CWID on (it's in the MMDVMHOST definitions under Expert mode of Configuration). Set "CWID" to one, and set a reasonable (5?) minute interval. Let's see if anyone else experiences the lockups or corruption.
I've been running the .082 1/22/2020 version clean for a day and a half, and will not load Roger's latest 1.29 version.
I'm always running the latest version, with CW beaconing enabled, and period is set to 10min. I was also running in at 1min, just to stress the system, as it will key the beacon
after x minute(s) of silence.
You mentionned a video, did I missed the URL ?
Cheers.
---
Daniel
Re: Update 29th Jan
Posted: Fri Jan 31, 2020 9:45 pm
by kd2lh
I have not edited and posted the video yet, but can do. Will work it tomorrow.
The 1/31 version ran from 8:57 this AM to about 3:50 this afternoon and rebooted, locking up the USB interface.
A power cycle of the Pi Zero W restarts the connection to the GD-77 in hotspot mode.
I did this, putting things back into hotspot mode, but...
Within minutes of restarting the Pi Zero, it locked up again. I'll power cycle the GD-77 and reboot the Pi next.
I tried turning off CWID and it doesn't appear to be helping.
It's been dropping out pretty consistently this afternoon. Two more lockups of PiStar with the GD77 rebooting to channel mode.
These are older GD-77 radios with a manufacture date of 2017. The new one I just got dated 2019 has a DSP version HRC6000 V1.2 and I wonder if this newer version DSP would make a difference?
- - -
Here's a link to a screen capture of PiStar desktop, the live log and a video of the failure. Ignore the display of the PiStar desktop on the video (the device's browser had stopped updating). The still screen capture of the desktop is from a PC that was updating properly up to the failure.
You'll see some voice activity about 30 seconds into the video, then the last incoming Brandmeister group voip conversation drops out mid transmission and GD77 reboots.
https://drive.google.com/open?id=1Klp_a ... ziw7EN-KwK
Marc
Re: Update 29th Jan
Posted: Sat Feb 01, 2020 12:34 am
by EB3AM
I use this version for 48 hours and seems to work fine.
Screen ok, modulation ok...
No bug detected...
Re: Update 29th Jan
Posted: Sat Feb 01, 2020 6:21 am
by F1RMB
Hi Marc,
kd2lh wrote: ↑Fri Jan 31, 2020 9:45 pm
I have not edited and posted the video yet, but can do. Will work it tomorrow.
The 1/31 version ran from 8:57 this AM to about 3:50 this afternoon and rebooted, locking up the USB interface.
A power cycle of the Pi Zero W restarts the connection to the GD-77 in hotspot mode.
I did this, putting things back into hotspot mode, but...
Within minutes of restarting the Pi Zero, it locked up again. I'll power cycle the GD-77 and reboot the Pi next.
I tried turning off CWID and it doesn't appear to be helping.
It's been dropping out pretty consistently this afternoon. Two more lockups of PiStar with the GD77 rebooting to channel mode.
These are older GD-77 radios with a manufacture date of 2017. The new one I just got dated 2019 has a DSP version HRC6000 V1.2 and I wonder if this newer version DSP would make a difference?
- - -
Here's a link to a screen capture of PiStar desktop, the live log and a video of the failure. Ignore the display of the PiStar desktop on the video (the device's browser had stopped updating). The still screen capture of the desktop is from a PC that was updating properly up to the failure.
You'll see some voice activity about 30 seconds into the video, then the last incoming Brandmeister group voip conversation drops out mid transmission and GD77 reboots.
https://drive.google.com/open?id=1Klp_a ... ziw7EN-KwK
Marc
Thanks for the video, I will watch it in a moment.
Here is a modified version of the firmware, if you could test it.
Beware, it doesn't include the latest changes made in the last few days (extra power settings and so on).
Cheers.
---
Daniel
Re: Update 29th Jan
Posted: Sat Feb 01, 2020 6:57 am
by F1RMB
Marc,
Something came to my mind about the reboots.
How many DMRIDs do you have in your GD?
I mean, looking up through could take a lot of time, in some conditions, and if it takes too long, the watchdog will kicks in.
As a test, can you flash an empty DMRIDs list, just to be sure.
Cheers.
---
Daniel
Re: Update 29th Jan
Posted: Sat Feb 01, 2020 9:11 am
by YT5HOK
Daniel,
I had no problems whatsoever with your latest FW (posted in that other thread) used as a hotspot, not even dropouts in regular use, mostly RX on local TG220 and global TG91.
Also I have 10-11k contacts in DMR ID list. It's not full, but it's alot for this device.
I'll continue testing this FW posted here, as I understand it's successor to the last one you posted in other thread, that I'm testing (with HS v.0.0.87), right?