When short-circuited a cell heats up first in a few spots, potentially raising the temperature high enough to force the cell into thermal runaway, and causing a fire and possibly bursting the cell. Stopping the short-circuit before this can still cause localized damage from heating but there’s really no way to know without some testing.
Yes, there are definitely some possible safety issues if you plan to use cells that were short-circuited. Since we cannot quantify the internal damage though, if any, to those cells it is your call whether to use them or not. Anyone who says you can definitely use them is just guessing.
A test you can do to check for really badly damaged cells is to charge them all to the same voltage and then leave them alone for two hours or so. Carefully record the voltage of each and then leave them alone for at least a week. Then recheck their voltages. Any cells that have dropped down in voltage a lot more than the others should not be used. Recycle them.
There’s no way I can know the internal condition of those cells so if I am asked I am forced to err on the side of caution and to recommend not using any of those cells.
If the cells are not usable for any application, like esk8, then they are not usable for all applications. Short-circuit damage doesn’t care how you use the cells, some risk is there for any use.
But it’s up to you, based on your preferences, priorities, and your tolerance for risk.
I plan on building a 12s8p pack for my mountainbord. My enclosure can fit 8 cells wide, but only 11 in length. So my plan is to have the last p-group lie flat on the other 11s8p cells which will be standing vertically.
Any precautions I should take when building a pack like that? Should I connect the last p-group with a folded nickel connection, or use standard wiring? I think I’ll use a piece of plastic, fishpaper and kapton to insulate the group lying on the other ones. BMS will lie on top too.
Should I solder the balance wires underneath the nickel to keep the bottom and top flat?
Hey guys, apologies if this has already been discussed in this thread. I noticed a cool post over at endless sphere regarding nickel/copper sand which. "Copper/nickel sandwich" buses for series connections - Endless Sphere
Are any of you guys doing this? I’m going to give it a go on my next build as it looks like a massive time saver only needing to solder the battery + & -leads. Guys seem to be getting the best results with nickel plated steel only at the connection point. Looks like a good option for a high discharge pack.
Seems fine to put the 12th pack on top if you add adequate insulation.
Making that join with nickel would probably prove to be a little awkward if it is to be how I imagine it to be… might be easier to make a wire connection there.
Run nickel tabs onto the sides of the pack for balance wires.
Can confirm. Chi ships thousands of batteries, built with the “fuck it” approach to craftsmanship, and so far every example of a fire has been removed from social media.
I agree with Al that doing folded nickel for that series connection would probably be a real pain in the ass, but I think it would be possible. If you decide to go that way, I recommend securing that p-group to the rest of the brick in a rigid way, so it cant bounce around and break the welds or fatigue the nickel. I would use a bunch of silicone to lock it in place, but there’s plenty of other options.
I recommend a good abrasion resistant plastic, or FR4/G10 would be even better. Either way, if everything is rigidly secured then it shouldn’t matter too much.
I agree with Al here again. Use small nickel strips hanging off the ends of each p-group for the balance wires to connect to. That way all your balance wires can just run down the side of the brick.
I don’t think this method has caught on here yet because it is significantly more steps required to build in the method in which we all agreed is the safest for esk8. So unless there is something like a 15% or more increase inefficiency, it’s unlikely to catch on here.
Scroll through this thread a little bit and you’ll see the most prominent battery building method here would make this technique hard to implement.
On e-bikes it’s popular because those batteries are often more shaped like a brick. Ours are flat and flexible. By distance, most of our current pathways are 2x 12awg wire. This is already pretty low resistance-wise. We also don’t use strips of nickel that are 8 or 10mm wide. We use 30mm wide strips and use them for short distances that need to be very low profile.
Also, 95% of people building esk8 batteries are not pushing current that would overheat their connections.
TLDR: cost benefit analysis says keep doing what we doing
I’ve used this method now on my rebuild of someone else’s battery build. I am not an expert. I would emphatically say any of the guys on here have more experience than I with regards to battery builds.
I liked that the pack is even slimmer profile in the end, and like you mentioned its “easier” to build, but you need better equipment to do it. My kWeld needed a solidly connected graphene battery to maximize A per weld, and it generates alot of heat, so you can’t weld non-stop. You also need to address the “sharpness” of your tips quite often. Granted, removing them and sticking them into a drill and holding them against a sheet of wet/dry 1000 grit doesn’t take a ton of time, but its time.
There is notably the sacrifice of flexibility. You can’t expect your pack to flex over and over using a copper sheet as the medium and not expect it to break. I basically took out all flexibility from my pack, and the build its on is really stiff already. The battery will be sitting in the enclosure now, and not attached to the board.
I am not advocating for the change of method, but I am raising my hand in a direct response to your question.
Some p groups are standard double stack, but most of them are in triple stack bricks. These are individually wrapped in fish paper, glued together, and then wrapped up in fiberglass tape. The series connections between each half of the brick are folded 0.2 mm nickel, and there’s 2 more layers of fish inside there for good measure.
There’s EVA foam around most of the battery and packing foam in the larger voids, none of the batteries touch the fiberglass. The series connections are triple 16awg (70A max current) and single 8awg. The BMS has B- connected to main discharge negative and C- sent out to a separate 16awg line. P- is unused, I’m like 90% sure this is how you do charge only BMS.
All of the wires have some extra waviness around the flex joints, and I can flex the enclosure into itself before the wires get loaded up in tension. The wires are immobilized with a combination of Gorilla Glue and fiberglass tape. Fuse, loopkey, charge port etc. will be in a separate enclosure from this.
I realize all of this invites questions like “why the hell is everything so weird”, etc. Long story. Any apparent safety issues though?
