Which wire gauge?

[rg12]

Hey Dudes,

I'm building x2 13s 10.4ah packs with a single BMS.
I want to connect each of the 13 cells in parallel with the other pack so the BMS will monitor the pack as a single pack (which it is what it actually is when it's connected like that).

My question is how thin can I go with wires that connect each of the cells?
Imagine that these are the two packs, each has 4 rows of 13 cells in series (top to bottom) and all 4 rows are paralleled.
The gap between the two packs for each cell will be connect with an external wire.

**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****
**** ****

Each pack will be 10.4ah which paralleled means 20.8ah (Samsung 26F cells which are 2C) and have about 40A max draw, let's say 50A if the controller is 45-50 in those scooters that I'm messing.

What do you think will be the amp "movement" in the wires that I connect to parallel each cell in series?

 

[wesnewell]

You can use small wire to parallel the cells as their won't be that much current going through them. However you need some big wire for the series connections and at the ends where you parallel them into one wire. Btw, an 8p pack of 26f cells has a max discharge amperage rating of 41.6A. And I wouldn't pull that much from them if you want them to last. They aren't the best choice by a long shot.

 

[Sunder]

If the cells are consistent, and you're not doing anything funky with it (e.g. drawing from a part pack for a lower voltage), then 28awg should be fine. Most BMSes have a very low balancing current, in the range of tens to low hundreds of milliamps. Balancers that charge rather than drain to balance can charge at higher rates.

Just make sure all 26 cells are perfectly balanced before you join any of them. If one pack is fully charged at 4.2v per cell for example, and the other pack is still in the 3.6v storage level, then 28awg wire will heat up and melt very quickly.

 

[fechter]

If you have 13 series connections (good), then those wires can be fairly small, like 16-18 ga since the current will be shared. If the series connection is a single wire (not so good), it needs to be like 10-12 ga.

 

[rg12]

Oh so I have to connect the two packs end voltage wires together to make the amp draw not go through the thin wires...
So I take both '+' wires and join them together and also both '-' wires that go to the BMS negative right?

What if one cell dies in one of the packs? will it make the thin wires transfer high amps in an instant?

 

[Sunder]

Oh so I have to connect the two packs end voltage wires together to make the amp draw not go through the thin wires...
So I take both '+' wires and join them together and also both '-' wires that go to the BMS negative right?

What if one cell dies in one of the packs? will it make the thin wires transfer high amps in an instant?

Oh.. missed that. Hell yes. Those high voltage whole pack wires need to carry everything the pack can deliver.

 

[fechter]

I probably misunderstood your diagram. If each pack already has the series connections, then forget what I said.

Taking two independently working packs and paralleling them should be no problem. The wires between packs can be small. The two main pack wires just need to be securely connected so there is little possibility of them coming disconnected. If that connection goes open, the balance wires could overload. If a cell goes bad, you have bigger problems with the ones welded to it.

If something bad fails, small wires will just burn off like a fuse. Sort of like a Tesla pack.

You need to be sure the corresponding cells in both packs are at the same voltage before connecting. They can be off a tiny bit, but should be within 0.01V if possible.

 

[Sunder]

What if one cell dies in one of the packs? will it make the thin wires transfer high amps in an instant?

Is this the part that made you think the series connection hadn't been made?

I read that as "What happens if Cell 3 in Pack 1 is at 4.2v, but Cell 3 in Pack 2 drops to 0v because it's faulty".

The short answer is big problems, lots of heat and smoke and possibly fire. Pack 1 would drive a higher voltage at full rate to Pack 2, using the big paralleling voltage wires, let alone the balancing wires (which would melt like a fuse)

The long answer is that it's unlikely. Short of physical destruction, cells don't drop that suddenly to be an issue - most will take voltage, and drain it off, so that your packs always sag and under perform, and you'll get frustrated and stop using it before anything catastrophic happens.

 

[rg12]

All are connected in series and on each pack, the 4 rows are connected in parallel.
I will connect the positive and negative of both packs in parallel and also parallel each cell with a thin wire between the two packs.
So am I still risking burning one or a few thin wires in case something goes wrong, for example a cell dying?

 

[fechter]

Not much risk if a cell dies slowly. A sudden shorted cell would cause a problem but the other 3 in the group welded to it would be more of a problem.

 

[rg12]

Not much risk if a cell dies slowly. A sudden shorted cell would cause a problem but the other 3 in the group welded to it would be more of a problem.

So you say it isn't a bigger risk than any regular setup, even less risky since a sudden short with thin wires can act as a fuse rather than having all the paralleled cells welded heat up, am I right?

 

[fechter]

Right. I don't think there is much risk of something bad happening.

 

[rg12]

Thanks alot