huge lipo fire , not cool

On the surface it appears that these fires are random..
various cell chemistry, RC LiPo , A123,(Heli)even the exploding FLA (Pb) a feww weeks ago.
different cells, different chargers, different applications,..often with no obvious cause.

BUT, i see one common factor.... the Human factor !
All of them were assembled/ configured, connected, and "managed" by a human. !
That has to be a major contributory factor
 
But so much of these things can be prevented cheaply. Eg Cell-log, make sure it isn't set to sleep mode. Get rid of the splash screen. Each cell-log drives a relay, with the charge voltage through a set of NC (normally closed) contacts. If one sub-pack goes wild, then the charge gets cut off to all of them. No muss, no fuss, cheap. Allows delta V detection and wont let you charge if the cells are too low as well (also important). Cell-log will bitch about a cell error (ie. balance wire disconnect) as well, and alarm too.

Sheesh. Throw a fuse in each series connection between the packs, and a 60 degree NO clixon (resettable thermal switch) across the brake line.

Make sure your packs are protected from mechanical damage, and the wiring is routed carefully. Double-up on the heatshrink, and use dual wall (ie. glue impregnated). Heatshrink semi-permanent anderson connectors together so you dont end up with parallel connections un-paralleling.

I'm doing all of this with my new pack build, and its open source. Coroplast around the cells for abrasion resistance, mounted on polyethylete cutting board, with 1" velcro straps. Its going on a downhill bike that gets thrashed half to death, last thing I need is a fire.
 
Redundant reliable electronics is a good plan.

Cell-logs are neither reliable or redundant, nor are they good to leave across a battery due to uneven and high drain they have.

If you want hardware and software redundancy one easy way to go is a Methods board and a balancing charger.

Another way to go is a Methods board and a bulk charger that has a management board added to it that looks at each cell voltage. I designed one but have not built it. There's a thread around here on it somewhere.

The key is a dual redundant system, one hardware and one software system in parallel monitoring the cell voltages during charging. Each independently capable of shutting off the charger.

Combine this with safe electrical connections and mechanical protection and packaging.

Still would be nice to have the packs detach from the bike so they can sit in a safe charging spot away from the bike.
 
Alan B said:
If you want hardware and software redundancy one easy way to go is a Methods board and a balancing charger.
Click to expand...

:D TY for making my 1st DIY power pack a safe and enjoyable one in the next few weeks (probably months) to come...
 
Jeremy Harris said:
I very strongly suspect that pack build quality, battery management strategy and the safety features in the charging system have a massively greater influence on battery pack fire or explosion risk than cell chemistry. I'll go further, by suggesting that a badly built or managed LiFePO4 pack (such as A123 cells) will be significantly more dangerous than a well built and managed LiCoO2 cell pack (say cheap HK LiPo).

For the majority of DIY pack builders it is their skill, care and knowledge that is going to determine battery pack safety, not the chemistry of the cells. In fact I doubt the chemistry of the cells is even really important in practice.
Click to expand...

I really don't want to disagree because I've used all lipo, a123, nmc in the 26650 format and the only ones I've not used are the head ways.

I've accidentally tried to charge a123 26650 cells using the li-io setting on my icharger. And the cells didn't even got warm to the touch. I've accidentally shorted a single cell when it fell off my hand and ended on a u shape steel channel (that one scared the hell out of me) and this cell gave its capacity as if nothing had happened.

So in summary I like/love/use both lifepo4 and lipo packs but at the end of the day the only chemistry I trust to charge indoors is lifepo4. Everything else I charge in my garage.

Note that I am a strong believer of hand checking everything and not trusting bms'es.

That being said I still want to try the NMC pouches from batteryspace (coming soon).
 
migueralliart said:
Jeremy Harris said:
I very strongly suspect that pack build quality, battery management strategy and the safety features in the charging system have a massively greater influence on battery pack fire or explosion risk than cell chemistry. I'll go further, by suggesting that a badly built or managed LiFePO4 pack (such as A123 cells) will be significantly more dangerous than a well built and managed LiCoO2 cell pack (say cheap HK LiPo).

For the majority of DIY pack builders it is their skill, care and knowledge that is going to determine battery pack safety, not the chemistry of the cells. In fact I doubt the chemistry of the cells is even really important in practice.
Click to expand...

I really don't want to disagree because I've used all lipo, a123, nmc in the 26650 format and the only ones I've not used are the head ways.

I've accidentally tried to charge a123 26650 cells using the li-io setting on my icharger. And the cells didn't even got warm to the touch. I've accidentally shorted a single cell when it fell off my hand and ended on a u shape steel channel (that one scared the hell out of me) and this cell gave its capacity as if nothing had happened.

So in summary I like/love/use both lifepo4 and lipo packs but at the end of the day the only chemistry I trust to charge indoors is lifepo4. Everything else I charge in my garage.

Note that I am a strong believer of hand checking everything and not trusting bms'es.

That being said I still want to try the NMC pouches from batteryspace (coming soon).
Click to expand...

They are the most forgiving cell I have ever tried.

I did get a bootleg A123 18650 cell to blow up overcharging it.

It swelled way up and exploded.

Real A123 cells can be hard to find and have safety features the bootleg cells do not ?

Those cells got checked off my list for sure.

It was sure nice making a pack out of the ones I bought from you. Real deal, not bootleg :)
 
Again another bulk charging error here, its so dangerous IMHO unless you are close to the pack and have means to get the pack out of the area quickly should all go wrong, this and the last few fires we have seen on here have been caused by 2 things, bulk charging and complacency :|

If you really must charge your pack in the basement at least monitor it or use an RC charger, simple precautions, even then charge them like they are going to burn up, I have never been a fan of charging on the bike either, much better to get the pack off the bike, you can better keep an eye on it, charge it on a fire proof blanket or steel sheet, you dont loose your bike if it does let go, etc etc.

I also am not a big fan of the folded steel battery case either (sorry Greg :oops: ) I know a few have adopted this method, too many sharp edges, combined with a conductive surface, although in this instance it does seem the outer case helped prevent the spread of the fire.

I am pleased for you that it wasnt worse, I have a pack that is older but very similar in size, it is charged on a fire proof blanket, under a smoke alarm using a cell monitoring balance charger, although the more fires I see on here the more i am inclined to make a little unit out in the yard and charge them out there.

Thanks for sharing this though as more people need to know and as Luke says fires come from all types, dont think you are safe using Lifepo4 either, you may be safer but its still not safe enough to ignore, a warning for everyone....oh and only bulk charge if you monitor and have some means to cut things off if all goes to hell.
 
A123 cells don't mind 4.2V, in early days some folks did that routinely. But a failed charger can go a lot higher, and they burn nicely once started. We need more than a small safety improvement, and A123 may not be a player anymore after the smoke from their crash clears. Chemistry is only part of the problem, anyway.
 
I've tried to make my Lipo pack for the USPD pretty safe.

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=44100

Arduiino based BMS monitors individual cell voltages on charge and discharge.
Separate charge and discharge ports mean no re-arrangement of wiring.
Packs in the battery case are snug to avoid shifting and chafing.
Cells are balanced every time they're charged beyond 4.1V/cell.
Charge is limited to 4.15V/cell.
PTC polyswitches on each balancing and main line.

I still charge in a fire-safe area.
 
In the end its all going to be dangerous, so i chose the most dangerous of the bunch >:) but in all seriousness just build it safe and secure, and above all simple, and you've done everything you can. you blow up you battery, its the cost of doing this. Though i've set up hundreds of harcase 4S batts and never seen a fire i didnt do on purpose.
 
dnmun said:
still no info on how the fire started. just a lot of opinions about how lipo should be managed. could be a short or overcharged, still not known.
Click to expand...
Its just that, unkown. probably never know unless they had some monitor. likely an out of balance cell that got overcharged, only the OP can tell us how the fire started, how would we know?? of course he's just going to get a shitload of comments about how to care for lipo.
 
Yeah, unless he finds an obvious source, like a spot on the box where a wire welded a hole in it, we'll never know if this was a charging error or just a short.

But my speculation still points at simply the pack being old, and one cell just died, then got overcharged rather than a charger problem per se. I got an old dying lipo pack myself, that I watch a lot closer when charging now.

A good short is specacular with any battery. A single 12v truck battery once melted the insulation off every single wire on my truck. A main wire touched the exhaust manifold, and got all the wires hot enough to fry the insulation, and lots of the wires just dissapeared. That was only 12v too. That was quite a fun fix.
 
Of course it would. Nothing to stop the others from getting overcharged.

This is one of my major concerns with the who bulk charging idea - nothing to stop stuff from getting out of hand.
 
heathyoung said:
Of course it would. Nothing to stop the others from getting overcharged.

This is one of my major concerns with the who bulk charging idea - nothing to stop stuff from getting out of hand.
Click to expand...

Unless you have a cell level monitoring /alarm/ cut off system !
Even something a simple as Cell Logs on the balance taps to check/monitor voltage
 
my greentime charger blew n bang loud sound but no fire. charger blew is safe becos there is a fuses. my new accu6 form hobbyking one of them seem short and charging all cells above 4.25 - 4.29v... luckily i was around i discharge the cells to all 4.2v. all my rc charge are suppose to stop @4.2v but it did fail once. so i will some how for sure take a look what volts is not going beyond 4.2v. at least for a rc charger will tell you all the cells for each volts.

i am not scare for troublesome taking out all the cells out n properly charge them with rc charger.
 
im sorry to say , i was really shocked over this incident ?how come over volts to 4.22 and hit nearly 4.3v from a rc smart charger? looks like a faulty and i re- do a charging again it did stop at 4.2v. so that is why hobbyking likes to say charging those lipo cells mustn't go un notice. this is no joke about it, if one gets a bulk charging , just a matter of time.. if one cells just grow older faster n fail, a fire will arise. charge it always with care
 
kentlim26 said:
im sorry to say , i was really shocked over this incident ?how come over volts to 4.22 and hit nearly 4.3v from a rc smart charger? looks like a faulty and i re- do a charging again it did stop at 4.2v. so that is why hobbyking likes to say charging those lipo cells mustn't go un notice.
Click to expand...

RC chargers may not even always read the voltage correctly, either, for a number of reasons. My Venom charger does nto read either voltage or current accurately; it's close enough for general purposes, but if it happened to be a matter of a tenth of a volt or a half an amp would blow something up because of it, I couldn't trust the Venom to be that close. :(

I havne't tested the Accucel6 for accuracy yet, but I should dig out the Fluke and a couple of other DMMs and do some comparison readings vs what this thing thinks it's doing.

So no matter what charging method you use, you should be watching the stuff 100% of the time if you truly want to A) know what is happening and B) want to be able to do something about it should your charging method fail to stop when it should for whatever reason.
 
I'd second AW's comment, as it's spot on in my experience.

I've had three RC chargers and one of them was way off. I binned it after I found that it was allowing two cells to go up to around 4.5V in reality. I suspect a lot of the RC stuff is built right down to a price and may not be that great, especially some of the cheaper stuff that seems to get retailer branded, rather than manufacturer branded. The higher end RC stuff is probably OK, but it's a bit of a lottery determining what has been designed well and what might be just a cheap clone.

As AW says, it always pays to cross check the readings from these things with a good multimeter, just to be sure they aren't on the blink.
 
lol this is why i use MWs, doesn't shut off so you're the only one to blame if it gets overcharged
 
Jeremy Harris said:
I'd second AW's comment, as it's spot on in my experience.

I've had three RC chargers and one of them was way off. I binned it after I found that it was allowing two cells to go up to around 4.5V in reality. I suspect a lot of the RC stuff is built right down to a price and may not be that great, especially some of the cheaper stuff that seems to get retailer branded, rather than manufacturer branded. The higher end RC stuff is probably OK, but it's a bit of a lottery determining what has been designed well and what might be just a cheap clone.

As AW says, it always pays to cross check the readings from these things with a good multimeter, just to be sure they aren't on the blink.
Click to expand...

I think it mostly comes down to QC. A high end brand name might sell practically the same product as a no name, but the brand name compnay might spend the money to test the product, and discard ones that don't meet tolerance standards before shipping to customers.
 
You must log in or register to reply here.
Back
Top