Telemetry freezing

[thetechnobear]

Ive reported this on the DJI forum, and will raise a ticket with DJI, but a bit of an fyi ...

basically, it looks like the googles lost the telemetry link, or the drone stopped sending data - mid flight (!)
Im hoping the aircraft logs will reveal more details.

the main concern here is loss of battery indicator... since there appeared to be no indication of drone wanting to return to home. even though battery was getting low.
not only did the goggles not show a RTH procedure - but the drone itself did not initiate it. (since Id have felt this in the controller/goggles)

concern here is the drone would have drained battery (and perhaps done a force landing)

(note: there is a possibility, that the proximity of the drone to home, meant RTH was not initiated)

Ive just landed after a nasty bug occured in the googles. I was flying around in M mode, when I suddenly noticed by altitude and speed were not updaing in the goggles. controller was working fine, and video signal was fine - RC/HD and 4 bars, and I was flying pretty close given there was a malfunction, I obviously decided to land... so I switch to N mode to cruise back, the drone correctly switched to N mode.. however, the goggles did not show this .. so it was obvious now the whole telemetry data had frozen. it landed fine, and was showing 40% battery left. (everything in the OSD was still frozen) now on the ground, I could 'reboot' the goggles. as soon as they rebooted, the telemetry corrected itself... showing N mode, altitude etc. however, the worrying part was ... the battery was at 20%, not 40% as shown on the goggles. as I say, the drone and controller, and video were fine - it was just the telemery data that froze. edit: observations - just checked flight in dji fly app... and also shows data stopped around the time I saw the OSD stop. - I did not get any battery alerts or notifications about RTH due to battery (which is a bit concerning) Im downloading the aircraft logs... though not sure how to inspect this

 

[Kilrah]

I've had my goggles freeze once and it was due to the SD card I had in them diying mid-flight.

 

[thetechnobear]

was this just the telemetry that froze? (i.e. video was still fine?)

I've a suspicion this is the same issue - I had noticed a few minutes earlier the sdcard in the googles had reported being full - so I turned all recording off.
but at that point osd was still updating...

but for sure, I saw this as an 'unhappy' coincidence.

did you report to DJI? they'll probably want to see the aircraft/googles logs... so they can potentially fix in a future firmware update.
(also the more of us that report it, the higher the chance they'll look at it)

 

[A.O.]

I've had my goggles freeze once and it was due to the SD card I had in them diying mid-flight.

And these guys require 2 SD cards yes? One for the bird and one for the goggles?

 

[Kilrah]

Yes video was still fine.
I didn't report since it was obvious it wasn't a bug, just caused by the dead SD card. Goggles would struggle to boot anytime that card was in afterwards.

And these guys require 2 SD cards yes? One for the bird and one for the goggles?

Not required, but yes you can have one in each.

 

[thetechnobear]

Yes video was still fine.
I didn't report since it was obvious it wasn't a bug, just caused by the dead SD card. Goggles would struggle to boot anytime that card was in afterwards.


Not required, but yes you can have one in each.

thats not true... the sdcard is not needed in the googles at all.
(the firmware is store in rom in the goggles, and the telemetry logs will be stored in nvram of the mcu)


so it should have no issue with telemetry even if the sdcard fails - but of course the firmware has to have the code to handle this!

 

[Kilrah]

It's not needed, but if it's there they'll try to use it.

The card when put in a reader would also kind of hang a PC, so no surprise it did it with the goggles too. It went read-only which can confuse devices since they try to write to it, it "succeeds", but the actual contents haven't changed and are no more as expected.

 

[thetechnobear]

sorry, but as a professional software engineer that writes for embedded controllers... i can assure you this is a bug, not something that 'just happens'

yes... it will throw an error when it attempts to write to the 'sdcard', but this can always be caught and then handled gracefully such that other operations are completely unaffected.
esp. when we write real time applications we never block the underlying processing.

I don't know what other diy/open source solutions are like in the fpv field, but really a company like dji should have the resources to deal with these kind of issues - but of course they do neeed end-users to report them, as they may not be aware of such bugs.

 

[Kilrah]

I've had the telemetry freeze again and could link it more specifically to the moment a recording is stopped, when the goggles' card is full (or faulty as in my initial post). Was compiling a list of bugs and improvements to send to DJI so added this and sent it out.

 

[A.O.]

Ok, exactly what do you mean when you say "the telemetry froze"?

 

[Kilrah]

The OSD values in the goggles stop updating.

 

[thetechnobear]

yeah, on further investigation - the flight log was also incomplete.

------

I raised the case with DJI, and provided them with the logs.
dev have confirmed the issue, and said it will be fixed in next firmware update.
(as expected, they did not provide details, just apologised and said it will be fixed - no date supplied)

 

[JoostGT3]

I had exactly the same issue last weekend. No card in the goggles, but there was one in the bird, and it was recording, card was not full. No auto rth kicked in, even though battery was way below 25% when I landed manually. Will be watching this thread.

 

[H3rtz]

Happened to me 4 times in a row today

BUT

I had a full SD card in the googles. Maybe that caused the failure.

 

[huskydriver]

This has happened to me today as well. All screen info froze, and no low battery warning or other failsafe. I flew it home and out in a new battery and no ossues. Is there a fix to this? This can potentially be a dangerous situation. Thanks.

 

[Kilrah]

Making sure your goggles SD isn't full, or not installing one at all would be all I could think of.

 

[bam]

Interesting to see this as well as the " non return to home" I recently managed to fly my dji fpv and I honestly thought someone was F'ing with me. Now it reassures me that I in fact had this experience, but doesnt say much for the product... i was actually ready to smash it to bits when it repeatedly tried to flip on its head during takoff, and only after Two trips out that it never left the ground because I couldnt get the fly app to accept my blood offering to unlock zones that Ive flown and un locked several times before?

 

[Kilrah]

RTH still works just fine, OP said they were in M mode and there is no low battery RTH in M mode.

 

[rktman]

RTH still works just fine, OP said they were in M mode and there is no low battery RTH in M mode.

What?!?? DJI is really this stupid? @Kilrah is never wrong, but nonetheless my brain is strongly resisting believing this...

 

[Kilrah]

Why is that stupid? The whole point of manual is, well - you're the one in control. It'll warn you of low battery but not initiate a low battery RTH for you. Only time a stop is forced is in case of loss of video/control, and auto RTH in case of loss of control only.