[jacquesm]

Could have been dropped in a bad way. This is the reason why Lithium Ion cells go bad, external damage. There are numerous examples of people falling with their e-bikes, not even visible damage on the outside of the bike or the pack other than maybe a scratched up handlebar and then 12 hours later the whole thing suddenly goes up in flames.

Here is one example caught on video:

https://www.youtube.com/watch?v=bbWmCiMJdT8

[andi999]

Company training reminded us, that a leaking Lithium ion cell produces hydrofluoric acid, which if you touch it you need to see a doctor.

[jacquesm]

Yes, it is an inherently unsafe battery chemistry. And they're everywhere. I really hope we find a better replacement soon, there are already some promising contenders but none that meet-or-exceed on all specs.

[userbinator]

That training was probably designed more to instill paranoia and cover legal arse, because my initial reaction upon reading that was "What the fuck!? No. No. No. Not even close."

Burning fluoropolymers will produce HF. Lion electrolyte is, besides being rather flammable, nowhere near as dangerous. It's a mild solvent. A leaking cell, if it's not burning, will not expose you to HF. In fact I'd say the risk of chemical burns from a leaking alkaline battery is higher.

[rbrtdrmpc-]

As i said the battery has a lot of space, is glued, and sealed shut within a “box”, you should be able to throw them out of the window with no problem at all, I’ve dropped mine a lot and they are already 5/6 yo

[jacquesm]

Yes, I would expect them to be designed for that, especially given the proximity to the body. I've reviewed a design for a LiPo powered wearable once and the list of recommendations was surprisingly long when it was all done. One thing that I find curious about this instance is how it failed, I'm very surprised that what appears to have burned through is the inner shell, rather than say a burst wall in the outer shell.

That said: headphones are subject to all kinds of abuse and that's why I'm particularly interested in what went wrong here, either, as you say, Bose has an inbound quality control issue or there was possibly some thing that this pair of headphones was subjected to that pushed it outside of its design parameters (electrically or physically). It's hard to say for sure without a much more detailed autopsy of the device (fortunately not of the user...).

I have a Sony rough equivalent of this device, and if I would ever drop it on a hard surface I'd stop using it simply because I would never trust a Lithium Ion battery after it has been subjected to a shock load. But I'm a bit paranoid about them, the more you are familiar with some tech the less comfortable you feel while using it.

As for your edit: thermistors can and do fail, but the battery protection circuitry is hip to that and should simply stop charging the battery if the thermistor ever open circuits or short circuits. Theoretically I guess it could fail with a resistance that indicates a safe operating temperature but I have never seen that happen (and I've worked on tons of Li-Ion packs, enough to see a couple of failed thermistors, all of them open circuits).

[0_____0]

Speaking of shock-loads to Li-Ion batteries...

I was once responsible for figuring out how to track the flight path of an object being towed through the air on a rope. This ended up being a GPS transponder with a ISM band packet radio and a Li-ion pouch cell. Packaging this was difficult, as the device would occasionally be slammed into the ground at at least terminal velocity.

One of my first prototypes was essentially all of the above stuffed into a bit of ABS pipe, with some foam inside for cushioning. We flew the prototype once, slammed it into the ground, and it promptly stopped working. Upon cracking the enclosure open, I found that the battery lead wires had broken free from the PCB under the shock load. As I extracted the pouch itself, I found it had formed itself into an arc segment, with r=1.5", same as the ID of the ABS pipe.

[jacquesm]

That sounds like a pretty lucky escape. A little bit less lucky and you'd have had nothing left but some black and unidentifiable bits.

[rbrtdrmpc-]

All valid points, about the thermistor i was more worried about it operating out of spec, it should have a specific thermal/resistance curve and maybe it is not following it anymore

[jacquesm]

Such a soft failure would be concerning. One BMS that I am intimately familiar with has two thermistors, one stuck to the batteries using a plug and a short cable, another surface mounted on the BMS board, if the delta between the two gets too large it shuts down any charge cycle in progress and bricks itself. Pretty harsh but given the alternatives it is the only safe thing to do.

[rbrtdrmpc-]

Oh that’s a nice redundancy, glad to know, I’ll probably check for those when i open battery powered stuff from now on