@andyp1per . Thanks for copter build to try. All loaded. Awaiting weather. I also reduced some logging load items. Do you think its possible for a UART/CPU issue to factory reset (for want of a better word) the RX?
Im still not sure if I have the same issue. Are you also loosing bind? Do you get RC back again on same flight? With mine a power off/on wont fix it. It needs a full re-bind process.
Today I did some tests, back to basics. Powered up the CRSF TX unit outside of the RC box. Also powered up the RX from 5v supply, not connected to FC. They stayed connected for about 5 hours without issue. Then put all bits together as if for flight with andy build apj file.
It sat there for a further 4 hours with yaapu telemetry running without issue. Not much of a test but why not as weather not great.
Im curious about the binding system. Can someone explain. This is what happened today.
The TX & RX were bound in a model in my X20. I removed them both and powered up on the bench. They required re-binding. They stayed connected for many hours as explained above. I put them back in copter and X20 and once again not bound.
Where is the binding info stored?
After lots of successful RC link tests yesterday, today no bind on RX. It forgot bind again. (this happens on 3 different RXs).
After rebinding I go to fly and then this -
The RX quality is up around 100% and it remains bound however the FC is having issues again with the RC. I hadn’t had this problem for a week or two. (just others)
I recycle the power a few times and then everything looks ok. Had a 40 minute flight with no issues, however I have no confidence in the system at all.
This is with my firmware?
Can you show your TBS RX and TX settings?
Yes with your firmware.
TX
FW V6.19
Hardware V1.03
Bootloader V2.05
RC input CRSF V2
Multi bind disabled
Freq 868
Max power 250mw
Dyn power off
Power sw off
Encryption off
Op power normal
RX Nano V3
Mode 8 chan
Telem on
RF profile Dynamic
Failsafe cut
RC by mavlink On
RX bat sensor off
Output1 CRSF TX
Output2 CRSF RX
Output3 CH3
Output4 Smart Audio
Channel map all 1-1 etc
FW 6.19
Hardware 1.01
Bootloader 2.04
Turn off RC by mavlink
I also noticed this was on just now. Was certainly off most of the time. Could be changing during the the lost bind issues.
Something that could be related. - The CRSF Lua for Ethos seems very flakey. When going into config on radio it will take many minutes to populate the menu. Sometimes will not fill anything at all. Its the latest lua I believe, about 3 months old.
I can 100% confirm that after a loss of bind that RC by mavlink turns back on.
Just had second flight today. After about 5 mins in STAB mode at about 100 mts got radio failsafe & RTL. This time total loss of connection and re-bind required. After bind RC by mavlink was back on.
I think there are 2 seperate things going on. A CRSF problem and a FC connection. Apart from this morning, I was getting much less FC connection issues. The loss of bind is my major problem.
Loss of bind is nothing to do with AP. Sounds like a hardware problem.
Will your firmware fix be in 4.5? I have a cube orange (AC4.4.4-bdshot) with crossfire (6.30) that is giving me failsafe issues. Weather will keep me from testing for a while so I’m not looking for a build right now. I’m planning to roll the crossfire system back to 6.19 before I do any more flying.
Yes, plan is to get this into 4.5
So you tried V6.31 beta? Brave man.
Fine line between brave and stupid. I’m probably leaning to the latter in most cases.
I’ve usually had good luck with crossfire betas so I didn’t get too worried when I upgraded. Thankfully I only upgraded one setup (one tx/rx), so reverting won’t be a huge PIA.
Sounds familiar, I think 2 or so years ago I was struggling with this for a while, on OpenTX and with the then-current TBS Agent, so I don’t think it’s related to Ethos or the current Agent version. Sometimes it took ages to set the VTX power manually by SA due to that problem. If I remember correctly it was fixed by an AP update.
While touching wood, I might say that my failsafe problems are more clear now.
I once again tried a different nano pro RX and so far I haven’t had ANY loss of bind type problems. I have had quite a few temporary RC failsafes where the TX & RX remain connected with high signal quality. I dont believe Ive had a failsafe in stabilize mode, only loiter and auto. Every time I have been able to interrupt the RTL and continue.
Probably I wasent always paying attention what type of failure I had so I thought I had already ruled out the RX for lost bind issues. Its been quite hard to untangle all the different problems that all end up giving a Radio Failsafe. Bad RX and bad TX and on-going FC type issues.
@andyp1per . Is there any point posting the last failsafe logs that seem not to be radio issues. As I said I already reduced some logging loads
With my fix you should not now be getting failsafes. If you are then a log would be helpful
Log. Actually the RSSI isint looking great for CRSF at short range.
HT POWER LINES!
Interesting test today. I turned the CRSF TX power to minimum (10mw) and set a quick mission about 300mts away expecting to get a RC failsafe at some point. It stayed connected but the lowest link quality is near home.
Ive often googled about HT power lines interference with RC but it seems the jury is still out on that one, especially about the 868/915 band. There are some 3 phase 11Kv lines shown in red. I think its quite conclusive that they are causing a problem.
I have defiantly had a bad nano RX loosing link and a bad TX that had died completely but I guess these cables were not helping diagnosis.
Unfortunately still getting random short RC failsafes that cant really be blamed on interference.
Did a long repeating mission and twice got RTL (last one was intentional).
From the logs is looks like the RSSI and LQ just freeze long enough for RTL to be triggered. Not sure if they actually freeze or its the FC link that pauses. Perhaps someone can take a look please. Is it pure chance both RC freezes are 6 seconds?
This is with a Nano Rx Pro?
Yes. Did you see my edit? Both RC losses are 6 seconds. Coincidence?