Can sag compensation kill batteries ?

[low trail]

My battery puffed up midflight and i have no idea why.
I would love somebody more experienced with the logs and quads to have a look and help me figure out the reason.
I can certainly say that i crashed with this battery a few times, but i am pretty sure that the battery was not hot or puffed up in any of my flights before.
The other concern is that i have 3 more of those batteries and i have no idea whether i should get rid of them after this incident.

Other batteries have internal resistance 6-16 mOhm with a difference of +-5 mOhm per pack.

The only difference from the previous flights that i was doing with this battery/setup is that i have changed:
Battery Sag Compensation OFF -> Battery Sag Compensation ON to 100%

Video link:
https://youtu.be/5K1Ar4m8ke8

Battery:
CNHL Black Series 6S 1100 mah

Frame: GEPRC Cinelog 35 Performance
Controller: GEPRC AIO F722-45A

Target: GEPRC_F722_AIO
Motors: Diatone Mamba Toka 2004 1700kv

ESC Firmware: G-H-30 - Bluejay, 0.18, 48kHz
Motor Timing: 22.5 deg

Betaflight:
Betaflight / STM32F7X2 (S7X2) 4.4.0 Feb 1 2023 / 23:18:30 (4605309d8) MSP API: 1.45
config: YES
board: manufacturer_id: GEPR, board_name: GEPRC_F722_AIO

Status:
MCU F722 Clock=216MHz, Vref=3.27V, Core temp=45degC
Stack size: 2048, Stack address: 0x20010000
Configuration: CONFIGURED, size: 3627, max available: 16384
Devices detected: SPI:1, I2C:0
Gyros detected: gyro 1 locked dma
GYRO=ICM42688P, ACC=ICM42688P
OSD: MSP (53 x 20)
BUILD KEY: 9c5716d9ca3904285de8fe36a64013bc (4.4.0)
System Uptime: 310 seconds, Current Time: 2023-05-16T11:58:36.240+00:00
CPU:56%, cycle time: 124, GYRO rate: 8064, RX rate: 15, System rate: 9
Voltage: 401 * 0.01V (1S battery - OK)
I2C Errors: 0
FLASH: JEDEC ID=0x00ef4018 16M
Arming disable flags: RXLOSS CLI MSP

[burnt pollen]

that single cell was probably damaged from prior crashes, and was not charging properly, then using the pack, pulled all the voltage out and it puffed.
LiPo safety is a concern, please safety check LiPo's regualrly. You do not want a house-fire from badly maintained or ignored damage.
you charger has a mode that displays each cell's voltage during charge, if one is not approximately the same as the others, it's bad.
there are plenty you tubes on the subject of LiPo safety.

[low trail]

The weird part that i had 25V battery at the start of the flight, which means that it has charged up fine.
And also that the damaged cell was not on the edge of the pack, meaning it had a very low change of beeing damaged in a previous crashes.

[burnt pollen]

your motors are 6S rated?

[low trail]

There are motor model number in the post.
Diatone Mamba Toka 2004 1700kv
The kv is not too high for 6s.

[burnt pollen]

okay. i can't deduce any other info than what i reported. 🔮

[low trail]

Alright, thanks for your time.

[supple axle]

Cell failures happen - VBatSagCompensation lets you push until the absolute limit of what the battery can do, and if that was one of the good batches of 1100mAh packs, they'll push right to the puff point, and you're just done. if they're from the bad batches, this can happen quicker, but more likely a single cell failure, being masked by sag compensation, that tracks.
I've actually only done this to a few dozen packs in flight, but it just happens for racers

What is slightly more concerning are those spikes on the gyro, which are likely causing bit ESC commits when the IMU is seeing electrical noise, so most likely you actually crashed due to a desync, but at some point the battery got overdrawn or damaged from the crash

[low trail]

Thanks for the vbatsag explanation, never using that thing again.
I do not have good control of the quad upside down and the battery started puffing as a did a flip, then i heard that horrible sound and got distracted for a sec and that ultimately got me to a crash.
So i do not think that was a desync, since i never had a desync on this quad.

[low trail]

Also i need to but a better capacitor on my quad, i think i have a high esr one, which is not a good fit.
Maybe that would help a bit with gyro noise.

[supple axle]

You do want to use it, but you basically have to use a mAh consumed indicator, feed it the right info, and actually come back in at the 'low battery' warning, and actually land when it says 'land now'.

(Or use a lower value, so get some sag, but it's easier to compensate for it since you're only compensating for it a little)

[low trail]

Yep, i have those warnings setup correctly.
I have battery low setup to 3.5v per cell and usually i am landing around that voltage.
I have never seen 'land now' message 🙂
I am trying to care about those batteries as much as possible, always charging at 1 amp (1C), never overdischarge (3.5 - landing), never overcharge (always up to 6S voltage), when coming back from a flights always charge to storage voltage.

[low trail]

Can sag compensation kill batteries ?

[lapis fable]

Low voltage, especially when hot and drawing a lot of current, is what puff batteries.

"land now" comes when battery voltage is below the 'critical' value

VBatSagCompensation does NOT cause battery problems. Once the battery voltage is below 3.5V, it does ABSOLUTELY NOTHING.
It does ensure a faster battery voltage check frequency in the code, so your logs, and the OSD, will show a much more accurate depiction of the actual in-flight voltage.
When batteries get 'old' or 'weak', the weakest cell rapidly deteriorates. It gets hotter, puffs more, and gets lower, than any other.
The most common cause of puffing is transient drops to below 3.1-3.2v under high load.
Monitor closely and you'll learn what to do to avoid that problem.

[low trail]

Any clues on how to determine if one cell is getting `bad' ?
Internal Resistance, Voltage after the flight ?

[lapis fable]

Best way is to rig up some load - some halogen bulbs in parallel or something - and see if each cell drops the same amount with a battery checker or voltmeter. Voltage on each cell under load should be the same. A bad cell drops low under load.