Hotspot mode - Apparently stuck in TX

Post Reply
User avatar
YT5HOK
Posts: 123
Joined: Sat Nov 16, 2019 11:36 am
Location: Belgrade, KN04FR

Hotspot mode - Apparently stuck in TX

Post by YT5HOK » Sun Nov 17, 2019 7:06 pm

I had this bug 3 times, in 3 days of testing, with 3 different versions (.7, .71, .72), so exactly once a day, once on every version, always after 5-6 hours from start.
Now, description:
I notice the GD77 Hotspot acting like receiving transmission from HT and staying in that state, but no TX was made with HT. Light would turn on, then go off, but screen stayed in "receiving mode", so to call it, with callsign, TG and RX frequency.
Pi-Star dashboard would show transmission being in progress, but nothing was really happening, except for 0.1 real TX, which is observed on second Hotspot (generic Chinese) and BrandMeister.
Now, HT listening to GD77 Hotspot would not RX anything from GD77 and Pi-Star would stay stuck in fake TX, not reacting on incoming TG traffic, until I really did TX from my HT.
After that all would return to normal, but I never managed to test it so long to see if it would get stuck again.
Marko

VK3KYY
Posts: 1656
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia
Contact:

Re: Hotspot mode - Apparently stuck in TX

Post by VK3KYY » Sun Nov 17, 2019 7:56 pm

I agree.

There seems to be a bug where PiStar thinks the radio is transmitting, but it isn’t.

I think this stops PiStar sending it any data, or perhaps prevents Pistar receiving signals from the radio.

I will investigate

User avatar
YT5HOK
Posts: 123
Joined: Sat Nov 16, 2019 11:36 am
Location: Belgrade, KN04FR

Re: Apparently stuck in TX

Post by YT5HOK » Sun Nov 17, 2019 8:46 pm

I have it stuck again. It's still Sunday here, so I am testing it whole day. Left it stuck to test more, cause of following...
HT receiving from GD77 did receive few transmissions on TG, which is strange... Meanwhile, other HT listening to generic Chinese Hotspot is working fine. I'm suspecting interference. Gonna move frequencies of Hotspots further apart, for testing purposes. They shouldn't interfere with each other, cause they are 100kHz apart, but still...
Will stay on this and report findings...
Marko

VK3KYY
Posts: 1656
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia
Contact:

Re:Hotspot mode - Apparently stuck in TX

Post by VK3KYY » Sun Nov 17, 2019 8:48 pm

It happened to me, but I don’t know the cause

Perhaps network breakup

User avatar
YT5HOK
Posts: 123
Joined: Sat Nov 16, 2019 11:36 am
Location: Belgrade, KN04FR

Re:Hotspot mode - Apparently stuck in TX

Post by YT5HOK » Sun Nov 17, 2019 9:26 pm

Happened again. This time I noticed it gets stuck on same (fake) TX as the last one was. I usually make private call to my other HT with different ID, working over other hotspot, not to disturb others on TG.
I'm gonna leave GD77 hotspot working alone, without any HT RXing or TXing over it, to exclude bug in MD-380/390 experimental FW.
Marko

VK3KYY
Posts: 1656
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia
Contact:

Re: Apparently stuck in TX

Post by VK3KYY » Sun Nov 17, 2019 9:28 pm

OK.

If you can find a reliable way to make this happen it would be a great help, as it makes it would allow me to replicate the bug on my radio and then I can try to fix it

User avatar
YT5HOK
Posts: 123
Joined: Sat Nov 16, 2019 11:36 am
Location: Belgrade, KN04FR

Re:Hotspot mode - Apparently stuck in TX

Post by YT5HOK » Sun Nov 17, 2019 9:52 pm

Of course. I'll try. As I mentioned in "new members subforum", I have 10 years of beta testing behind me, although in Android SW, but I hope to be helpful here too.
Looks like if I don't reset GD77, it gets stuck more often. It did again, confirming my theory of being stuck in same (fake) TX as last real one was. One more thing I noticed from these tryouts is that the first real TX after it, to un-stuck it, is made the same as the one it was stuck on, no matter what I set my HT to, probably cause Pi-Star just continues the TX that it thinks is doing. It can be observed on Pi-Star dashboard and BrandMeister. After it is working again, all is working fine.
I'll need to change my polarization from vertical to horizontal soon, as we say it here (most probably all AROs do, too), cause it's working day tomorrow, but I'll continue to investigate this.
Edit: Oh, yeah, I forgot... There was no HT on GD77 Hotspot RX/TX frequency when it got stuck last time, so bug in MD-380/390 FW is scratched from list of possibilities.
Marko

VK3KYY
Posts: 1656
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia
Contact:

Re:Hotspot mode - Apparently stuck in TX

Post by VK3KYY » Sun Nov 17, 2019 10:26 pm

I think the value show in PiStar is from the value it gets back from the Hotspot when it sends the GetStatus command, but I'll need to double check

The only way to really know, may be to enable more logging in PiStar and look at the MMDVMHost log file.

The code that returns whether the hotspot is transmitting is

Code: Select all

	buf[5U]  = (	hotspotState == HOTSPOT_STATE_TX_START_BUFFERING ||
						hotspotState == HOTSPOT_STATE_TRANSMITTING ||
						hotspotState == HOTSPOT_STATE_TX_SHUTDOWN)  ? 0x01U : 0x00U;
So its possible that the hotspot is stuck in one of these states.

It takes a few hundred mS for the Tx to end, after the network data buffer is empty, as the radio has to send the End Of Transmission DMR frames.

During this time, I have some code which can restart the transmission. So potentially I have a bug in that part of the code and perhaps I should let it finish completely and then restart from the beginning

I suspect you will need to somehow simulate a unreliable signal from the network, rather than a weak RF signal

k8upe
Posts: 19
Joined: Sat Nov 16, 2019 2:15 pm

Re:Hotspot mode - Apparently stuck in TX

Post by k8upe » Mon Nov 18, 2019 1:59 pm

I came here to report the bug but see that it already has been reported. This happened to me last night. Don't know if this helps but when it did happen, I quick keyed my radio and it went away. I waited to see if it would happen again and it has not yet this morning. Also running PI-Star

VK3KYY
Posts: 1656
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia
Contact:

Re: Hotspot mode - Apparently stuck in TX

Post by VK3KYY » Mon Nov 18, 2019 7:38 pm

Thanks.

At least it does not seem to actually be transmitting, its just that PiStar thinks it is ;-)

Post Reply