Wire bonding for high powered packs?

Can a wire bonding technique (cell fuses) work with high current packs?
Sure it works for high current when you have thousands of cells in parallel but in our high powered ebikes that have 10-20A current draw per 18650/21700 cell, is it smart to use thin wires?
First, voltage sag, can it work without adding any?
Second thing, if the wire is thick enough for 10-20A, will it really be useful as a high current fuse if a cell dies inside a parallel group?

Anyone?

rg12 said:

Can a wire bonding technique (cell fuses) work with high current packs?

Click to expand...

Yes. When there are large P groups.

rg12 said:

Sure it works for high current when you have thousands of cells in parallel

Click to expand...

Correct.

rg12 said:

but in our high powered ebikes that have 10-20A current draw per 18650/21700 cell, is it smart to use thin wires?

Click to expand...

No.

rg12 said:

First, voltage sag, can it work without adding any?

Click to expand...

No. V=IR. There will always be sag. You decide how much is acceptable.

rg12 said:

Second thing, if the wire is thick enough for 10-20A, will it really be useful as a high current fuse if a cell dies inside a parallel group?

Click to expand...

Not unless the P group is large.

Cell level fusing works best when there is a high ratio of P group current to cell current. I.e. Lots of cells in a group, and each cell normally outputs low individual current. The total group current is high. Then each thin wire doesn't create a large voltage drop in normal use, but if a dying cell drops voltage low enough that all the other cells combined can flow current back into that dying cell that can blow the fuse.

Small P groups of high current cells doesn't have the same effect.

serious_sam said:

Not unless the P group is large.

Cell level fusing works best when there is a high ratio of P group current to cell current. I.e. Lots of cells in a group, and each cell normally outputs low individual current. The total group current is high. Then each thin wire doesn't create a large voltage drop in normal use, but if a dying cell drops voltage low enough that all the other cells combined can flow current back into that dying cell that can blow the fuse.

Small P groups of high current cells doesn't have the same effect.

Click to expand...

That's what I'm asking, why not use a thicker wire and then when a cell dies and lots of amps trying to be pulled from say an 8P pack with 1 dead cell within them, then a much higher current will go through that dead cell in comparison to a pack with lots of P cells and a thinner fuse wire.

Also, can you please explain to me why and how is current flowing through/into a dead cell in a P group?
Some dead cells are just shorted so they drag the whole group down and some have almost zero voltage/zero voltage...

rg12 said:

Also, can you please explain to me why and how is current flowing through/into a dead cell in a P group?
Some dead cells are just shorted so they drag the whole group down and some have almost zero voltage/zero voltage...

Click to expand...

Current flows from high voltage to low voltage. A dead shorted cell is 0v. So this:

serious_sam said:

if a dying cell drops voltage low enough that all the other cells combined can flow current back into that dying cell that can blow the fuse.

Click to expand...

Lot's of false information here!
Fuse wire is usable in low P high(ish) current per cell application too!
I have a 6S4P pack where the cell connections are 0.3mm dia. copper wires. I regularly discharge the pack with 40A (10A/cell) without any issue. The fuse wire resistance is ~5% of the cell's internal resistance only, so the extra voltage drop is negligible. During testing, a dead short blown four parallel fuses in fraction of a second.
I don't see any reason, this could not be scaled up to 20A/cell, as high energy cells having less IR (than my Pana 16850PFs).
The higher short circuit current could blown a thicker fuse wire without any issue.

T.

just because you can does not mean you should.

your desk is not a tesla factory. you dont have robotic ultrasonic welders that cost half a mil a piece.
forget about wirebonding. just grab your 0.15 or 0.2 nickel strip and just use that. much more secure and gives you less problems.

dont reinvent what works, you are not elon musk.

takyka said:

Lot's of false information here!
Fuse wire is usable in low P high(ish) current per cell application too!
I have a 6S4P pack where the cell connections are 0.3mm dia. copper wires. I regularly discharge the pack with 40A (10A/cell) without any issue. The fuse wire resistance is ~5% of the cell's internal resistance only, so the extra voltage drop is negligible. During testing, a dead short blown four parallel fuses in fraction of a second.
I don't see any reason, this could not be scaled up to 20A/cell, as high energy cells having less IR (than my Pana 16850PFs).
The higher short circuit current could blown a thicker fuse wire without any issue.

Click to expand...

10A/cell is not high current for 18650, so your user case is not applicable.

- A 10mm length of Ø0.3mm copper wire has resistance of ~2.38mOhm.
- At 10A each wire converts 0.24W into heat (@141A/mm^2).
- At 20A each wire converts 0.95W into heat (@283A/mm^2).
- Anything over 40A/mm^2 would be considered "hot" for copper.

Regardless, you've missed the point of cell fusing entirely anyway. Cells don't just instantly fail in short-circuit. It can happen from extreme electrical or physical abuse, but then you probably have bigger problems to sort out. Your short circuit "test" isn't really a real world situation for 99.9% of user cases. Almost all pack fires that I have heard about are caused during charging.

Bad cells usually just degrade over time (i.e. self discharge, whether due to dentritic crystal growth or other reasons), and the sagging voltage puts stress on the surrounding cells, which silently kills the pack. In large P groups, the cell fuses can blow at relatively low delta V (nowhere near a dead short). This situation is ideal for large sealed packs with hundreds of P cells (think large EVs), because it automatically isolates bad cells before too much damage is done, and allows the pack to remain servicable without intervention.

For relatively low P count packs in ebikes, fuse wires make far less sense.

If you're worried about a cell spontaneously going dead short, you've got much bigger issues, and a cell fuse isn't going to stop the fireball.

And did you solder the copper to the cell ?

1. "10A/cell is not high current for 18650, so your user case is not applicable."
-It is your opinion and has nothing to do with the topic. First post is about 10-20A load/cell. So, my 10A example is applicable.
2. "At 10A each wire converts 0.24W into heat"
- Meanwhile, each cell in my example, converts at least 3,4W into heat. In ideal conditions! For me, the extra loss is marginal, minimum acceptable.
3. "At 20A each wire converts 0.95W into heat"
-So, the fuse wire sized for 10A would blow. Did you try to support your opinion with some fake calculations, or just simply did not realize, bigger currents would require thicker wires?

"And did you solder the copper to the cell ?"
- Yes, of course. And please don't start the "soldering kills the cells" debate, as long as you never soldered a single cell the proper way.

T.

short answer: no, you dont use wires for cell connections ever. you are not one of elon musks million dollar robots.

takyka said:

2. "At 10A each wire converts 0.24W into heat"
- Meanwhile, each cell in my example, converts at least 3,4W into heat. In ideal conditions! For me, the extra loss is marginal, minimum acceptable.

Click to expand...

Everyone reading this can see that those 2 things are totally different. Losses are a secondary consideration after temperature rise. But according to you, a fuse wire and a cell have comparable heat capacity. Or maybe you missed the point again.

takyka said:

Lot's of false information here!

Click to expand...

takyka said:

Did you try to support your opinion with some fake calculations

Click to expand...

That's twice now you've claimed that I'm posting false/fake information (even though the maths is easily verified as correct, and my opinions are at a minimum logical and follow common convention), but all you have is poorly informed opinion. Fock off troll. I'm done with you.

Here's the one I use. Around $75k and works very well.


https://www.youtube.com/watch?v=2l4rB4igFG4

st35326 said:

Here's the one I use. Around $75k and works very well.
https://www.youtube.com/watch?v=2l4rB4igFG4

Click to expand...

That's very awesome. I like how you can connect to the cell rim and avoid making connections on the other side the pack.