Battery Cell Autopsy?

[Bikes4two]

• Thanks to invaluable input from forum members I now build my own battery packs
• My oldest pack, 15 months old (a 10s 2p using LG M50LT, 21700, 5000mAh) has been brilliant and has done maybe 60 full charge/discharge cycles?
• Imagine then my suprise when one day recently that it didn't power up
• A strip down to the BMS to voltage probe the cell groups found one of them reading just 0.25v. It was a straight forward job to replace the pair of cells (like for like of course) and the pack is now back in service
  QUESTION(S)
• Is this type of failure, when failures do occur, common and have I just been unlucky to have a poor cell, or might I expect further failures in the near future?
• I am an undemanding ebike rider riding my 250w TSDZ2 almost exclusively in ECO mode so I can't imaging I've over-stressed the cell's rating (and the pack is fitted with a 30A fuse which has never blown)

As an aside, I know that some folks try to revive dead cells, but to me a voltage of 0.25v seems terminal to me, but as a general question, is there a general cell low voltage point below which it is unwise to try and revive a cell (I seem to recall someone mentioning 2.5v recently)?

Thanks in anticipation.

 

[matthewslack]

I would not assume the cells are the root cause without carrying out a low current charge/discharge test. Say 0.5A, either monitored by data logger, or manually if you don't have one. Just voltage readings every 10 minutes or so, throughout charge and discharge.

If the cells pass that test, then the issue is elsewhere, perhaps BMS, perhaps a tiny bit of moisture temporarily affecting one of the cell group drain circuits.

 

[Nealh]

Drain is likely a BMS issue , most cells are very good today including the value china ones like Bak, DLG, Eve, Sinowattt etc,etc.

Agree at 0.25v they are dead and a bin job.
2.5v is ok though one would have to manually do so at milliamps to 3.0v.
2.5v is the std low discharge battery voltage manufacturers use to obtain max mah rating for cells though this is only at a 0.2a discaharge rate. As most know 2.5 is never an advisable lvc to go to and why most cease between 3v - 3.3v as this is typical lvc for a BMS.

 

[Bikes4two]

• Thanks @matthewslack and @Nealh for your input.
• Interesting that you mention a possible BMS issue Neal - I was on holiday in France at the time of the failure and with no tools for a strip down, I ordered a replacement BMS (a BeMuchSafer one) as a precautionary measure and have in fact fitted it to the battery pack along with the two replacement cells.
• On the day before the failure I'd been riding the bike - temperatures were in the upper 20s - at the end of the ride I'd left the bike in the sun (on an open campsite) - it was the next morning that the battery failed to power on so now I'm wondering if something in the BMS didn't like the heat and had caused the cell-pair to drain out?
• Anyway, for the sake of £16 for two cells and a similar amount for the BMS plus a couple of hours 'happy toiling' with my spot welder, heat shrink etc, I'm back into service - just think how I'd be weeping in my soup if I'd had a Bosch battery et al!

 

[Nealh]

I think you would do more then weep , more like a full on blubb.

 

[Nealh]

Typically it is a bleed resistor that remains open , not shutting off continues to drain the cell group. I suppose standing sun could cause an issue, which BMS brand failed ?
I have had two Daly failures one killed my 48v panny celled battery and the other drained my lg hg2 battery a little but not to an unsafe level.
I have some Daly's still in use but won't buy no more.

Can't really complain yet with Be Much Safer BMS.

 

[Bikes4two]

.......which BMS brand failed ?
......

Can't really complain yet with Be Much Safer BMS.

Be Much Safer BMS - but it is very unclear and unsafe to blame the BMS - I still have the old one which I could take a closer look at maybe?

 

[guerney]

A strip down to the BMS to voltage probe the cell groups found one of them reading just 0.25v. It was a straight forward job to replace the pair of cells (like for like of course) and the pack is now back in service

One of the things I worry about considering making my own battery, is accidentally overheating cells while welding, leading to their early demise. I'd much rather trust a robot.

 

[Nealh]

Making your own gives you more choice of the cell you want to use and for bespoke batteries.
Over heating isn't an issue with a decent arduino spot welder, the contact is minimal in time /miliseconds. Soldering cells is where one will over heat them as the heat applied is far more prolonged.
With cells the only place to be careful with when welding is the central area of the cathode end.

 

[Bikes4two]

Fair enough that you should be alive to the possibilty (of overheating the cells) but as Neal says, this isn't really unlikely if using a spot welder for lithium batteries.

Like Neal I avoid the centre and given that you're joining at least two cells together I do a weld on one cell and then one on another cell and then back to the first and so on, thus giving the heat a bit of time to dissipate.