Gimbal getting out of control during survey flights

[vr-pilot]

Hello everybody!

Since using the current E90 gimbal firmware 2.40.3 (January 2019) I have the problem that the gimbal might get out of control during survey flights. When I use any earlier gimbal firmware version the problem does not occur. But the big disadvantage of using the older firmwares is that RC connection losses (5.8 GHz video and telemetry) remain permanent until landing (manually via 2.4 GHz or by RTL) and restarting the UAS. With the current FW version 2.40.3 RC connection losses are fully recovered midair quite fast.

So I have to decide between unusable survey photos from stable RC connection flights that nevertheless have to be interrupted by stopovers to reset the loose gimbal on ground (incl. UAS restart), and good survey fotos from flights with possibly no video transmission and telemetry data (e.g. not knowing the battery level, flight plan progress etc.).

Did anyone experience this, too? Are there known ways to fix this problem?

Here are some sample fotos of survey flights that had to be interrupted due to a loose gimbal looking "somewhere":

 

[Steve Carr]

Have you tried to use the gimbal update firmware? E90 Fix for Gimbal Vibration

 

[vr-pilot]

Have you tried to use the gimbal update firmware? E90 Fix for Gimbal Vibration

Hi and thanks for your answer!
I've downloaded the file under your link even before I started this new thread. ;-) I thought this patch file would be for "vibration fx" only, but I will give it certainly a try, maybe already tomorrow.

Do you have any information about the "RC-does-not-reconnect-after-connection-loss-problem" I always experience when using any older version than the current 2.40.3?
In other words: is the reconnection behaviour of the FW patch file as good as the FW 2.40.3?

 

[Steve Carr]

In other words: is the reconnection behaviour of the FW patch file as good as the FW 2.40.3?

I don't have an answer for that. I will ask around, however. Perhaps someone has more experience with it.
@arruntus seems to know this system better than most.

 

[arruntus]

I would try to put in both the camera's firmware and the ST16S firmware since they are the two pieces involved in that error.

 

[vr-pilot]

I would try to put in both the camera's firmware and the ST16S firmware since they are the two pieces involved in that error.

The above mentioned patchfile did a good job concerning the reconnection robustness during a test flight yesterday.
The gimbal "lookaway"-problem (see photos at the top of this thread) did not occur, though it was not very windy during the test. Maybe the patch file (displaying as Gimbal Version 2.61.0) is a fix for that, but I could only confirm that after having flown in some stronger winds...
So thank you all very much for the help. I will report here when a high wind test is done.
I have seen the gimbal "lookaway"-problem on additional two out of eight H520s with E90, so that error might be "relatively common".

off topic errors (independent of the patch file):
- In "Outdoor" display mode (white scheme) the zoom factor is not displayed (empty field right to "123" below the camera angle band). "Indoor" mode (black scheme) works fine.
- there are absolutely no visual or aural warnings anymore during any kind of connection losses (RC 2.4 and video/telemetry 5.8 GHz).

 

[arruntus]

@vr-pilot it is highly recommended to clean the cache of the transmitter, of the software, to avoid that values of previous versions influence new versions. This can cause such errors.

We should take the following as a rule: when upgrading we should clean the cache and the application data so that everything is reset for the new version.

In extreme case, make a complete cleaning, that is to restore to factory the ST16S and to reinstall everything from 0.

 

[vr-pilot]

Thank you for the ST16S recommendations. I frequently delete pictures and video files on the remote control to clean up memory space.
I will try to make a complete cleaning as you suggest, maybe from 0. Maybe this will help reducing conflicts the next weeks when I start using the E10T on the system.

Yesterday the gimbal FW patch also worked flawlessly during a relatively windy day (no "lookaway" errors as shown in 1st post).
There comes a further question in my mind:
On the download site "H520-Software Releases for Yuneec Partners" the gimbal FW version numbers and file names are all identical (except for the E30z):
Gimbal
Region Version Download
E50 all v2.40.3 gimbal.yuneec
E90 all v2.40.3 gimbal.yuneec
ET all v2.40.3 gimbal.yuneec
E10T all v2.40.3 gimbal.yuneec
E10Tv all v2.40.3 gimbal.yuneec
E30z all v2.62.2 gimbal.yuneec
Does it mean that I can use the forum's FW Gimbal patch FW 2.61.0 also for my upcoming E10T? Is it similar or comparable to the E30z FW v2.62.2?
My guess is that these FWs above look identical but each one is a "unique" (pun not intended) version?

 

[vr-pilot]

Hi folks,
with this film strip above I just wanted to say "Thank you!" again for the help with the gimbal patch firmware for the E90. As you can see the gimbal is working with a kind of "resilience-sequence": once the look-away-problem occurs it resets itself midair to normal operation... (this works without having to land and doing a reset on ground)
BIG THX! Great improvement for me!

 

[DoomMeister]

In answer to your last two posts, it is good to hear the gimbal firmware solved tour problems at least in part as you no longer have to land and reset.

The earlier post brings up the FW naming likenesses for the different cameras. I am quite certain you are correct in assuming there are differences. The gimbal firmwares address issues with balance for each different camera and while similar are not balanced the same between models.