I’m pretty sure Isis released the layout spec
6 Likes
How did you charge those cells? What is the material you used to spot-weld the batteries? The series connections look way to small and the discoloration indicated that it was cheap nickel plated steel from aliexpress. So either you didn’t insulate correctly on the fold ie. use poly-amid tape and while charging something got warm and expanded causing a short or your charge leads shorted one of those weak series connections your undersized connection got redhot and bent down melting through the batteries insulation, shorting the battery to itself causing a fatal termal runaway in your battery, Once one battery goes they all go.
Next time use 100% pure nickel. 8mmx0.15mm Nickel can handle 6A plated steel can handle 2A. Also Kapton or poly-amid tape is important. Double check your balance leads!
1 Like
no, look at the original pictures, it’s folded down the middle of the two-cell stacks.
this should be stickied for what not to do.
1 Like
No man. You should take another look at the finished pack.
I think your problem was here where I’ve marked when you folded the pack…
Of course, you bridged the polarities so of course, they shorted 
1 Like
Each half should be like this if thats the shape you want
And that nickel should only be used for parallel, never series
2 Likes
He’s right. 8 cell stack on bottom, 7 cell stack on top.
(first, I know so few about batteries that’s why I will rely on an expert to build mine).
@eboosted was right. This thing is a bomb. 'am sorry to say this but the FAA, airlines and transportation authorities have all the right to be concern. Batteries are capable to bring down commercial aircraft easily and produce a major catastrophe. Specially a hand-made custom battery pack. A few years ago one of the most modern airplanes, if no the most, a Boeign 777, was destroyed by a short from a faulty battery inside the avionics compartment below the cabin. A hundred million dollar aircraft destroyed by a simple faulty battery. 
1 Like
where they all at the same voltage to start? Did you check?
if not then this is what did it. one got over charged. then a chain reaction. 
folding could have also shorted like other people suggest. not sure why you would fold them sine they are 2 wide and 2 high either way…
I think for sure it was the folding that shorted the cells and then the charging later while shorted that set it on fire. very unlikely a faulty cell as they are very very rare and if series charging especially with a new pack its very unlikely to be out of balance already and even if it were out of balance and were to send a cell way high I don’t think that would do it. According to manufacturers’ testing I think maybe all the 18650s in whatever chemistry pass a standardized shorting test and overcharge tests and don’t go into thermal runaway and it’s only when many of them are put beside each other that they will build up a temp that’s capable of sending one into runaway. so I read.
Haha. (Almost) All of you are convinced that i folded those like on the edited picture Blasto posted.
Like you see, the bottom two rows have 7 cells and top two has 8. 7 folds to 7, and 8 folds to 8. The center-fold would be very foolish 
And that fold is the easiest way to get the pack in the form it was. Minimum cutting and bending of the nickel stripes.
Actually if you folded the pack like Blasto pointed, it would maybe be slightly possible to avoid the sparks for a long time, if you are in luck 
@bimmer: I have tested that those are pure nickel. Sand it to expose potential steel and keep in salt water for overnight. So the discoloration must be related to welder settings. I happened to be today at my local hackerspace, and we also discussed about this topic. Friend told also that the most plausible thing that happened, was what you said; one of the cells were faulty, expanded, heated and caused an chain reaction. Also the friend told that while it would discolor, it is still not that much heat that it would damage the cell internals. Soldering the cells is way more destructive/dangerous and yet still people get successful packs.
@ducktaperules: Charging the whole pack from one end to other, while monitoring the individual cell voltage.
@scepterr: Shape you pointed was my first idea also. I came up with the different configuration, because it was easier to accomplish, especially with the folding-part. Of course i could have folded every cell-pair individually, but that again, would have been harder. Yea, single nickel stripe is not enough in series if you are going to run serious currents. What i use is around max 10A
@saul: Yes i measured they all were the same voltage. And like i said on this post earlier, they were individually monitored.
@Hummie: It is not the folding. Read the top lines of this post. But a friend said that he has had a few faulty cells also that were wrapping-melting hot. Luckily he did not have the series cells hugging each other. Only the one parallel got bad.
So the most plausible conclusion; A faulty cell. What i learned; Use kapton tape or similar to wrap the cells, and cycle the cell capacity one time before assembling into a pack.
But what you suggest for the spot-welder settings?
About my background; I have hundreds of salvaged laptop-battery cells. I have made many batteries for my e-bike and different things, though i have always used these brackets which protect the cells touching each other: (picture is for quick illustration purpose only)
This time i were over-convinced that the cells are surely good, that only the voltage measurement would do. The old laptop cells have to be measured with capacity on nominal current and also internal resistance.
2 Likes
There would be only 1 fold in the configuration I posted
Those shaped “nickel” rolls, are not pure nickel, It sparks when you weld right?
1 Like
Proper insulation. Yes masking tape is proper abrasive resistance when applied over kapton like fishpaper…if ya don’t believe me disassemble any production LiPo battery.
Edit: general reply…not directed towards who I replied to
2 Likes
improper insulation. kapton insulates heat, the thing that you dont want in you pack. you have the entire pack wrapped in it so no airflow or heat movement is possible
1 Like
Greco… Been building Li-ions for a long time, the ends are insulated, and the packs are never operating at 100% of thier rated capacity. Yea kapton traps heat, but cells don’t get hot when you pull less then half of thier current capabilities 90% of your riding.
My packs run cooler then my ESC’s. Say improper insulation… Or safe insulation and not stressing cells… I’ll choose safe insulation and treating my cells like babies.
1 Like
Why they never leave a few cut openings on the plastic wrap? That should fix the heat trapped inside, no?
@scepterr: To accomplish the shape you see on original post 5th picture how do you do it with one fold?
It doesn’t spark while i weld. Although while welding to nickel-plated steel, you can get spark-free weld if you give enough and even pressure. But i will test the stripe again to make sure. Could be that i haven’t sanded it enough 
@Deckoz: Yea i have the same paper than on production packs. Material is plastic and surrenders under soldering iron. I have used that on every LiPo-packs i have made.
@pixelsilva: Humidity gets in…
1 Like
it will help but not eliminate the problem. try to cover as little surface area as possible when using kapton in packs <6p and heatshrink the whole thing.
1 Like
Just 1 fold between 5th and 6th series at the end
What was going on with voltages? Nothing crazy, everything coming up nice and evenly?
Did it start sparking immediately?
I’m really hoping we develop some kind of theory, for safety’s sake.
Melting insulation shouldn’t even have been an issue, whatever went wrong was already far gone at that point