Spot Welding Copper Strips to 18650 Battery Cells

[garolittle]

Good information. Really appreciate the ideas. I have a “family full of lawyers” so I will be researching it carefully. Thank you.

 

[nuxland]

Why’s there such a huge gap in welders where we can get a welder for 230$ (Kweld)that will do .1mm copper but need to spend way more to get something to do .4? Cant they just make a welder with more/bigger bits on the board?

That is actually a great question. I’ve often wondered the same thing. You would think there would be a $2k welder that would allow for enough power for at least .25 mm thick pure copper.

At what voltage you new welder is working? I'm assuming more than 12V and less than 2000Amps

 

[garolittle]

[

At what voltage you new welder is working? I'm assuming more than 12V and less than 2000Amps

Usually weld at 66J - 72J at about 24 milliseconds. Not sure what the settings would be for copper thinner than .40mm but will try soon.

 

[Frank]

Why might this welder work on .4 mm copper at those settings and the kWeld won't?

I'm not home right now but have welded thinner copper using the "sandwich" technique at those energy levels (not sure of the pulse time) but I'm pretty sure it wouldn't work for .4 mm thick stuff.

 

[jonescg]

Why might this welder work on .4 mm copper at those settings and the kWeld won't?

I'm not home right now but have welded thinner copper using the "sandwich" technique at those energy levels (not sure of the pulse time) but I'm pretty sure it wouldn't work for .4 mm thick stuff.

I'm assuming it's the duration of the pulse that matters here. Huge currents for a few milliseconds versus the kWeld doing lower currents for longer? Long enough for the heat to conduct through the copper and away from the weld site?

Also, this generates a very hot plasma arc, and it's the arc which does the melting.

 

[spinningmagnets]

Yeah, it's the plasma. I can't see any way to make a focused plasma jet for a few precisely timed milliseconds for under a few thousand dollars...yet.

 

[flippy]

Yeah, it's the plasma. I can't see any way to make a focused plasma jet for a few precisely timed milliseconds for under a few thousand dollars...yet.

its pretty easy, just grab the leads from your home fuse box and shove them together, instant plasma arc.

 

[garolittle]

“Also, this generates a very hot plasma arc, and it's the arc which does the melting.”

I think this is the key statement.

 

[garolittle]

We appreciate everyone’s ideas so far. It appears we now have consistent weld settings for the cathode (positive) side of the 18650 cells dialed in correctly. This is a practice video has increased speed to show the consistency of the welds. Keep in mind that is .40mm (.016”) thick pure copper. Working on the Pulse Arc welder settings for the anode (negative) side now. Thanks again for all the ideas. Click the link below. Vextrek

https://youtu.be/fsPnn5zntZY

 

[pbert]

I have a question, which i myself think is not so bright, but what about just simply taping the copper connections in place. I mean simply wrapping Kapton tape around the entire battery, holding the connectors in place ? they shouldn't move after they are taped.

I am currently building a 2P14S pack with 40T cells that can output 60 amps continuous and wanted to use copper.

Any reasons not to do this ?

 

[spinningmagnets]

There are some posts about putting a button of foam on top, and then compressing it a precise amount. The best foam is closed-cell, and Poron was identified as a viable choice.

 

[john61ct]

they shouldn't move after they are taped.
...
Any reasons not to do this ?

You **really** do not want **any** possibility of the batteries shifting relative to their connections.

Stationary use cases are different, but bikes are inflicting massive continuous shock & vibration



Sent from my moto g power (XT2041DL) using Tapatalk

 

[pbert]

You **really** do not want **any** possibility of the batteries shifting relative to their connections.

Stationary use cases are different, but bikes are inflicting massive continuous shock & vibration

Yes makes sense. This is what i was thinking would be the risk also... I just thought if it was taped together well enough that is shouldnt move but i guess the risk would always be present and its not something you want to be thinking about when flying through the woods. Shouldn't be worried about the battery trying to kill you, that's what the trees are there for!

Thanks.

 

[pbert]

Here is an update as to how i decided to build with copper connections. I have a malectrics spot welder and with that it is possible to spot weld nickel to copper strips. Here are the welding combinations that are possible with the malectrics:

- nickel to battery = yes
- copper to battery = no
- nickel strip to copper, welding on nickel side = yes
- nickel strip to copper, welding on copper side = no
- copper on battery with nickel on top (sandwich) = no
- copper to battery with infinite slits = kind of but really not strongly attached

So with the above in mind i was trying to figure the best way to create a secure weld on the battery and also securely attach the copper. My previous post about taping is out of the question.

I ended up welding lengths of nickel for the parallel which extend out a bit. Then place the copper connection on top of the nickel strips. Fold over the nickel lengths and weld those to the copper. Seems to be the best way to get a secure weld to the battery and the copper.

Anybody see issues with this or possible downsides that i am missing ?

 

[Frank]

It seems to me that would work but only for a 2P design. What thickness copper is that? I'm surprised the sandwich technique did not work.

 

[pbert]

it's 0.15 nickel and 0.2 copper. Why do you say this would not work in more than 2p build ?

 

[Frank]

The way it is now current flows through each nickel strip to the copper from each end and each cell should contribute equally. If you had for example, a 4P configuration, current from the middle two cells would have to flow past the outer two cells via the nickel strips to reach the copper. It would still work but I don't think it would be balanced. Hopefully others with more experience than me will chime in.

 

[pbert]

From my understanding, current follows the path of least resistance. With this is mind, the copper is in contact with the nickel strip on the battery side as well as the nickel that is wrapped around. Therefore the current path would travel from the battery, then through 0.15 thickness of nickel attached to the battery, then through the copper to the other cells in series, through the 0.15 nickel (and in the case of the negative terminal, the copper is actually in contact with the battery directly) then into the battery. If i understand what you are describing, you are saying that the current would be going the length of the nickel and through the nickel that is wrapped around the copper and then out of those wrapped ends that are welded to the copper. This is not the path of least resistance though and from what i understand i dont think this is the way the current would flow.

For a 4p pack i think that it would be possible to have the same build as i've done for 2p. Just cut a longer length of nickel and wrap it around to hold the copper in place. The current would be evenly distributed.

 

[spinningmagnets]

I think it will work, and work better than just nickel alone.

I would like to suggest that you try an "infinite slot" and a nickel cap over copper, with the copper in direct contact with the cell-end. Also, perhaps use 0.15 copper, instead of 0.20, which would not dissipate the welding heat as fast, making a good weld at a lower energy pulse setting.

 

[pbert]

I have some 0.1 copper sheet. I just tried an infinite slit with it and it welds much better than the 0.2mm copper. The problem is that the battery is putting out 60 amps continuous and so the 0.1mm copper is not enough as far as i know. I could order some 0.15mm and try with that but I would need to wait to receive it ...

 

[Frank]

From my understanding, current follows the path of least resistance. With this is mind, the copper is in contact with the nickel strip on the battery side as well as the nickel that is wrapped around. Therefore the current path would travel from the battery, then through 0.15 thickness of nickel attached to the battery, then through the copper to the other cells in series, through the 0.15 nickel (and in the case of the negative terminal, the copper is actually in contact with the battery directly) then into the battery. If i understand what you are describing, you are saying that the current would be going the length of the nickel and through the nickel that is wrapped around the copper and then out of those wrapped ends that are welded to the copper. This is not the path of least resistance though and from what i understand i dont think this is the way the current would flow.

For a 4p pack i think that it would be possible to have the same build as i've done for 2p. Just cut a longer length of nickel and wrap it around to hold the copper in place. The current would be evenly distributed.

I know what you're saying but I didn't think there was any welded contact between the nickel and copper underneath the copper. If you're relying on non-welded contact it may not be reliable.

I've welded .2 mm copper using infinite split nickel-plated steel (with Kweld) but I suspect the .15 mm would work in your case. With a 2P pack at 60A you'd deplete the cells pretty quickly, wouldn't you so maybe heat buildup wouldn't be that big an issue?

 

[spinningmagnets]

I have some 0.1 copper sheet. I just tried an infinite slit with it and it welds much better than the 0.2mm copper. The problem is that the battery is putting out 60 amps continuous and so the 0.1mm copper is not enough as far as i know. I could order some 0.15mm and try with that but I would need to wait to receive it ...

How many cells in parallel? the 2P shown at 60A would be 30A per cell.

 

[pbert]

Yes you are right that there is no weld between the copper and nickel on the battery side and I agree with you that it is not 100% reliable. To help improve the contact, once the connections are made i would put some foam and a outside hardshell (ive used a flexible cutting board in the past) and then use this (by taping around the outer hard shell) to compress the foam down onto the connections, which would help keep the connections in contact. Not perfect but should do the job.

 

[Frank]

That would certainly help, good idea.

 

[pbert]

I have some 0.1 copper sheet. I just tried an infinite slit with it and it welds much better than the 0.2mm copper. The problem is that the battery is putting out 60 amps continuous and so the 0.1mm copper is not enough as far as i know. I could order some 0.15mm and try with that but I would need to wait to receive it ...

How many cells in parallel? the 2P shown at 60A would be 30A per cell.

It is a 2P14S pack using Samsung 40T cells which can put out 30A continuous. Im going to be using it with a controller that pulls 45 amps max on occasion but most of the time will be used with a TSDZ2 pulling about 12 amps.