Devo 7E corrupted file system

This last year I have had three times a corrupted file system, always at the flying site after switching off and on the TX. And I did't have a spare TX (which I have got now). Before this year I had never has this problem. I have ruled out some possible causes and want to know other people comments:
1- Operating the TX with low voltage - I use NiMH batteries, and I always get the message "Can't save model" before the battery alarm shows up.
2- Bug in my custom FW derived from master branch
3- Bug on master branch FW - I have never had this problem with 5.0 and 5.x branch versions.
4- Faulty solder on memory IC
5- Faulty memory IC

I bet that the low voltage is the real cause. I have set the battery alarm to 4.3V, and I'll put a 2S LiPo after checking than there is no other issue.

BR

In many years of use I never had this issue during normal operations.

However only when you connect your TX to computer USB port you have to be carefull, and not just unplugging USB. Make sure you always click
"safely disconnect USB device" in you windows first, before unplugging.

Otherwise it could sometimes happen you damage the filesystem, in that case need to reformat the usb drive and put the files back. (make sure always you keep a back up of your models etc)
The Tx firmware itself is not affected, just files.

Thanks for your answer Fernandez.
Recovering the filesystem is not a problem, I always have a backup.
I don't think that the problem is related to not disconnecting securely the USB when the TX is connected to the PC. Las time I connected was some weeks ago to do a backup of the filesystem. And lately I always securely disconnect the USB. After that, I have been flying some days, switching on and off the TX every battery pack with no issue. Then, suddenly one day the TX doesn't switch on and get stuck is USB mode. This same day I had to change the NiMH batteries, that's why I'm suspecting the low voltage as the main cause of corruption.

could be. I power my 7e with 18650 and use polulo step up down regulator, it keeps always 3v3 whatever the state of the battery lower or higher then 3v3 input from 18650.

[quote="...it keeps always 3v3 whatever the state of the battery lower or higher then 3v3 input from 18650.[/quote]

I have been using 1S lipo for many years without problems. The key is to bypass the diode in series with the battery supply chain. Without doing this will cause a diode forward voltage drop so that the main board is not actually get the full 3.3V power supply.

Since reading many years ago about Devo 7e power circuit failing when run with 2s lipo I have kept to the AA NIMH rechargeables, never had any problems and I just plug the devo 7e into a 12v wall plug the night before to charge. Now and then I charge on the smart charger for good measure.
My low battery warning is set to 4100 and critical battery warning is 3900.

@cmpang
I did also bypassed the reverse protection diode, it will also increase battery life, and not waste energy due to the voltage drop over diode.
Still 3v3 onboard regulator, to do the job properly needs to be powered at higher voltage. 3.9V ? So using 1s direct to me seems a bit of sketchy....

Still I would love to have Vbat calibration offset and scale inside hardware.ini so can have the measurement displayed correctly.

I did a running test right after the mod and confirmed that RF output power maintained all the way until battery dropped to 3.75V and noted the voltage readout at that moment for reference.

yes a calibrated Vbat is the root solution ...

Thanks for all your messages.
Although nobody has linked the corrupted filesystem to the low voltage operation. I'll keep the NiMH batteries with the 4,2V alarm to see if the problems gets fixed.
If I understand well, bypassing the reverse protection diode will increase the voltage without no real harm to the batteries (or other parts of the TX).