I still don't think the balance wires experienced over current. In the OP's picture, I highlighted in yellow the intact portion of the balance wires, if they experienced over current, the whole lenght will be burned.
View attachment 247628
I think you're correct here.
Ironically, I just did some balance lead failure testing last night. I was curious if fiberglass looming would help mitigate cascading failure from one wire burning another, and from dripping molten copper/molten insulation on anything below. (Good news: it does).
What I also discovered in my experiments was that...
- If a wire is in a circuit that has experienced an overcurrent condition
outside of the wire itself, and it isn't fused, then it will
always (in all 6 tests I ran) burns through its entire length.
- However, wires whose insulation which were nicked - and the short happened on the wire itself - won't
necessarily do this.
- Wires near the failed wire, will ignite themselves, if flames, molten copper, etc. drip onto them... but
never along their entire length unless they're touching along their entire lengths (in my tests).
- As one would expect, failures began at the end of a circuit with the highest potential and progress near-instantaneously; the eventual break in the wire/conductor, would happen roughly 0" to 1.5" from where the smallest cross section of conductor began. (e.g. in a 60" circuit, where 0">20" are 4AWG, 20">44" are 22 AWG, and 44">60" are 4AWG, the break in the wire would always occur between inch 20 and inch 21.5).
Here's some video of a few of the tests (you'll see only half of the wire glowing in some of those videos, that's because the other half is sheathed in fiberglass insulation - it still glowed under that fiberglass, you just can't see it from this perspective):
With (right) and without (left) fiberglass loom:
I think for the OP's failure above, my experiments confirms your own observations:
- The failure did not begin in the balance leads.
- The charred balance leads do indicate a cascading failure of sorts; at least one balance wire was likely in proximity to one of the full-charred wires (or the PCB) where it ignited, and the fire spread but was not sustained because wire insulations tend to be UL 94-V0 plastics that self-extinguish once removed from a flame/heat source.
- The actual failure likely occurred on the PCB but certainly along a circuit supplied by one of the charred wires. It's unlikely the wire was nicked if it's still in one piece through all of its length.
- Therefore, to find the source of the failure, it should be along a trace supplied by the completely charred wire. Specifically; at a location of trace nearest the partially-charred balance leads. Or, if those partially-charred balance leads crossed the completely charred wire somewhere, then the failure on the trace would be at some less specific location.
