Spot Welding Copper Strips to 18650 Battery Cells

Vex
I'm using a ~5000w ups transformer which I rewound. It puts out about 6v and a truckload of amps ( ~1200 as a guesstimate)
The switching mech is a 30amp SSR controlled by a 555tiner circuit. Single pulse, don't know exact pulse lengths sorry but it is in the order of 40-60 ms I'd say. If I go too short I get nothing as the SSR opens and closes to quick for the AC cycle. The pulse curve is different to a CD welder so it's hard to compare.
I use pure copper electrodes. I just use solid core wire about 1.5mm diam. These are clamped in copper welding collets so as about 20mm sticks out.
I started off using a MOT but upgraded to this UPS transformer I got cheap. It's way better than the mot. But not as strong as a DN10.
I don't have the voltage to drive tungsten electrodes. They won't weld sheet for me and stick so hard I destroy whatever it is trying to get them off (think angle grinder)
Hope that helps.
K
 
kdog said:
Vex
I'm using a ~5000w ups transformer which I rewound. It puts out about 6v and a truckload of amps ( ~1200 as a guesstimate)
The switching mech is a 30amp SSR controlled by a 555tiner circuit. Single pulse, don't know exact pulse lengths sorry but it is in the order of 40-60 ms I'd say. If I go too short I get nothing as the SSR opens and closes to quick for the AC cycle. The pulse curve is different to a CD welder so it's hard to compare.
I use pure copper electrodes. I just use solid core wire about 1.5mm diam. These are clamped in copper welding collets so as about 20mm sticks out.
I started off using a MOT but upgraded to this UPS transformer I got cheap. It's way better than the mot. But not as strong as a DN10.
I don't have the voltage to drive tungsten electrodes. They won't weld sheet for me and stick so hard I destroy whatever it is trying to get them off (think angle grinder)
Hope that helps.
K

very good info, thanks. So it is possible to spot weld copper even with longer pulses such as yours, they don't need to be super short. My weld times are typically 1-10ms, my max amperes are around 5000A now, but energy is currently insufficient with 200Ws. I will increase my energy capacity and max amperes, and reduce my total resistance (around 3.6mOhm now).

I was also getting a lot of blowouts with anything other than copper electrodes (molybdenum, tungsten, 30%-70% tungsten copper)

so waiting for another capacitor to up my energy to 400Ws



Q2: why are we not considering aluminum tabs?
 
Yeh my pulse times are longer but my current is lower and more even. Yours is short and really sharp. Not sure if it is better in all cases, but most think short and sharp is good. With my SSR it's a zero point switching so it only turns on when the AC voltage is at zero no matter when the timer starts. So if I go too short the timer goes on&off before the SSR even starts. In theory I could get 10ms but what happens is I sit there tapping the pedal over and over trying to get the timing just right. I guess I'd prob be doing 20ms for .15Ni
Aluminium has an oxide coating that is tough, insulative and has a really high melting point. I tried out of curiosity to weld some... Access Denied. I Prepped it and it was thin, like .1mm. I've heard it can be done but it's a whole heap harder than copper.
In all cases of welding copper or brass. It's a fair bit easier if you are welding onto nickel.
 
Oh and I file the electrode tips if I blow out and give them a quick scuff on some 240 grit every 20-40 welds. Just to keep them shaped nicely. Slightly rounded on the end. As for pressure, my electrodes are on a couple of arms with adj Spring tensioners. I've found this much more reliable than hand held electrodes, but it does get difficult making big packs. I'll have to figure something out there...
 
From Sunstone eng YT vid.
310ws for .125mm copper
510ws for .25 mm copper
But they were welding with an electrode on each side....cheaters!
 
I have found .2 mm to be a ceiling of sorts with the power levels the can will handle. Over this thickness and it gets a bit sketchy with heat and weak welds etc.
with a slot, I get reliable welds with .3mm in nickel/brass but not so much as a tack without the slot (at this thickness)

If using the common plastic cell holders/separators (which I strongly recommend), the standard width of the ribbon is 8mm, in order to fit the 8mm slot.

8mm X 0.15mm = 1.2mm squared = 16 ga (common round cross-section wire)
8mm X 0.20mm = 1.6mm squared = 15 ga
8mm X 0.25mm = 2.0mm squared = 14 ga
8mm X 0.30mm = 2.4mm squared = 13 ga

http://www.engineeringtoolbox.com/awg-wire-gauge-d_731.html
 
so I tweaked my 1F / 20V /200Ws CD welder and I am now able to weld .1mm copper. Sort of. I basically cleaned all contacts and increased copper cross section for bus bars and electrode cables.

- at 20V blowouts are inevitable, welding is really sensitive to pressure, electrode shape. I think welding should be done with lower voltage and longer pulse.

- difficult to find ideal welding energy. Too high - blowout or puncture and evaporate piece of copper or make a small blowout hole - copper patch sticks to the can but there is now a hole in the tab. Too low energy - weld doesn't stick. It is not really repeatable and uniform result.

- 0.1mm copper is mechanically too weak. It's like paper. Tears easily. Easy to puncture it with a weld. I think to use copper we would need thicker. 0.2mm or 0.15mm. I would not dare to use battery pack with .1mm copper.
 
I strongly recommend against spot welding copper to 18650's. The weld will never be reliable or strong enough.
The only proper way to do it is pulse arc like the Orion.
There is a guy named Fritz somewhere who was designing a cheaper pulse arc.
Stick with nickel and solder copper to the nickel if you want more current capability.
 
Fritz is still working on the design. Super smart guy. That is why I built his 800 WS CD welder he designed.

Tom

I edited because I was wrong with my info.
 
update: I rebuilt my CD welder and increased max capacity to 2F @ 20V, which is 400Ws.

I am able to weld 0.2mm copper at around 12 to 13V with full discharge. The higher the voltage the higher the chance of blowout and sparks. Welding copper also puts much more stress on MOSFETS and electrodes, I had to redesign MOSFET drive circuitry, and buy an oscilloscope in the troubleshooting process. All was fine with 0.2mm nickel, but 0.2mm copper requires a few times more energy and when blowout occurs there are some significant transients happening which can eat up MOSFETs in no time.

Also while testing welding copper on an old cell, something got shorted inside the cell when a weld was made, causing it to instantly heat up to some 50 degress. It was an old cell half empty.

Welding energy required to weld 0.2mm copper can be decreased if entire strip is cut in half longitudinaly, so you end up with basically two parallel strips which are welded by electrodes being placed on each strip separately. This increases the weld quality significantly. However then you end up with a very narrow strip and if the electrode is close enough to the edge, a blowout is likely.

The point being: I am pursuing this more out of scientific curiosity, but welding a battery pack of some 100+ cells requiring few hundred welds - I think welding with copper 0.2mm strips would be too risky. 0.1mm is mechanically too weak. I will try also 0.15mm within a few weeks.
 
My sunstone engineering CD welder can not even come close to welding 0.15mm slotted copper @ 240ws. I agree I don't think its worth trying to weld copper with electrode based CD welders. Pulse arc is the only way to go, or your going to need so much power it becomes dangerous.
 
redilast said:
My sunstone engineering CD welder can not even come close to welding 0.15mm slotted copper @ 240ws. I agree I don't think its worth trying to weld copper with electrode based CD welders. Pulse arc is the only way to go, or your going to need so much power it becomes dangerous.

well I am at least glad to hear that the welder I built stands up to commercial solution :)

There is little info on pulse arc welding online, if anybody has info on how that welding machine works, please share info or links.
 
update: I've run hundreds of test welds with 0.2mm copper and fine tuned welding energies, timing, and electrodes. 0.2mm opper can be welded but nickel must be at least .3mm to take the heat without blowing up and electrodes need to be dulled to create ideal contact point (too small: blowup, too big:no weld). Even then, the weld energy must be fine tuned to be enough for welding copper but not enough for blowing up nickel. It's a small margin. I added energy calculations to my welder to track that.

However, 0.2mm copper is not mechanically strong enough. I think even thicker copper will not be mechanically strong enough, as material it is less flexible and with repeated vibration/wiggling it breaks, even though the weld was perfect. So for some applications I would use copper, but not for ebike which will take gazzilion of vibrations and hits over time.
 
eTrike said:
Great info, thanks for sharing! Correct on mechanical issues with copper-- more isolation is required. Fwiw I've used a copper welded pack that sees a lot of movement and vibration (no suspension) for over 5 years. As long as the welds aren't used for mechanical integrity (as they shouldn't be on any battery pack anyhow), and the pack is designed well, I don't see a problem.

Good point. If the pack has external casing which provides structural integrity, and if padded inside, it could be good enough.

Did you build that pack yourself? What copper thickness was used?
 
After opening 23 BL1840 Makita packs (the newer pack version), I can tell you they use 0.15 mm thick nickel-plated COPPER and spotweld it on the VTC4 cells.

I noticed while scavenging cell.... that orange glare while cutting with the tabs with the dremel... I you sand the nickel on the surface, you clearly see the pure copper under the plating.

So Makita uses copper in their rugged hevy-duty tool packs meant to substain constuction field related abuse. Of course, the cell holder they use are rock solid, but strangely, the nickel-plated copper aloso seems quite solide.... maybe the plating helps make the tabs be more rigid.
 
Pics from the residual Makita Toolpack tabs (after using demel tool like so - see 6:15 : https://www.youtube.com/watch?feature=player_embedded&v=Oz6Uis05ewA )

See what I mean : Is that copper or what ??
LPF.jpgIMG_1072.JPGIMG_1069.JPGHPF.jpg

Seems like Makita has figured it all... Spotwelding copper. They probably use the Orion Pulse 250 EV spotwelder or something... Or maybe the nickel coating changes the game and makes it all possible.

This is also part of the reason I choose Makita packs... They've done the hard part for me... Now I can just solder the nickel plated COPPER tabs to a copper bussbar like so : https://endless-sphere.com/forums/viewtopic.php?f=14&t=84791&start=25#p1274105

With a 0.15 mm thick x 8-9 mm wide copper tab, I can expect 18 amps per tab ... but since there are two legs on each cell ... 36 amps with my technique ! more than the cell should deliver anyways.
 
If you find a local antique restorer, they will often have a nickel-plating source that is as close and affordable as is available in your region. I asked my local guy, and he said nickel-plating anything copper is easily done by them. He wouldn't give me his source, but I can drop off and pick up from him, and he will likely mark-up 20%, which is fine by me.
 
thanks for the info Matador

I can weld .2mm and get a really good weld, it's just that I am reluctant to use it in a bike which is vibrating a lot. It would probably hold though. I did a ton of thests and nickel as material endures wiggling so much better. I think it is more elastic than copper.

Can you figure out, by looking at the weld points, was it a resistance weld or pulse arc weld? Resistance will have a small press in mark, dimple. Pulse arc will have a small puddle-like spot where material was liquified and solidified later, but generally in the same level as the rest of the tab, not depressed.
 
vex_zg said:
thanks for the info Matador

I can weld .2mm and get a really good weld, it's just that I am reluctant to use it in a bike which is vibrating a lot. It would probably hold though. I did a ton of thests and nickel as material endures wiggling so much better. I think it is more elastic than copper.

Can you figure out, by looking at the weld points, was it a resistance weld or pulse arc weld? Resistance will have a small press in mark, dimple. Pulse arc will have a small puddle-like spot where material was liquified and solidified later, but generally in the same level as the rest of the tab, not depressed.

I understand, but still seem hard to figure out for me. What do you think ? Here are some pics (sorry, my camera seriously sucks and I had to try it many times) :

View attachment 1take 2.jpg
 
thanks for the pics. I also need a better phone camera :)

this looks like a normal resistance weld, with press-in-marks.

I will continue experimenting with copper welding, perhaps I get an idea how to reduce risk of tearing.
 
vex_zg said:
thanks for the pics. I also need a better phone camera :)

this looks like a normal resistance weld, with press-in-marks.

I will continue experimenting with copper welding, perhaps I get an idea how to reduce risk of tearing.

You're welcome! There's one thing I've been wondering all along while discretely following this thread for quite some time now...

Raising resistance makes the welding easier right ? Could the skin effect, that is the fact that copper is plated with a thin layer of much higher resistance metal (nickel), make the welding process possible ?
In other words, I wonder if welding with nickel plated copper is not much easier (and maybe more sturdy) than just plain bare copper because current trough nickel generates enough heat??

I'd be great if we had spy to tell us how Makita does it !... Imagine, Tesla spent time devlopping their ultrasonic weld procedure so they could use high amps cells with high amp connections.... If we could do the same without having to solder copper with an hot iron, or spotweld multiple layers of nickel strips... That'd be awesome !
 
Matador said:
vex_zg said:
thanks for the pics. I also need a better phone camera :)

this looks like a normal resistance weld, with press-in-marks.

I will continue experimenting with copper welding, perhaps I get an idea how to reduce risk of tearing.

You're welcome! There's one thing I've been wondering all along while discretely following this thread for quite some time now...

Raising resistance makes the welding easier right ? Could the skin effect, that is the fact that copper is plated with a thin layer of much higher resistance metal (nickel), make the welding process possible ?
In other words, I wonder if welding with nickel plated copper is not much easier (and maybe more sturdy) than just plain bare copper because current trough nickel generates enough heat??

I'd be great if we had spy to tell us how Makita does it !... Imagine, Tesla spent time devlopping their ultrasonic weld procedure so they could use high amps cells with high amp connections.... If we could do the same without having to solder copper with an hot iron, or spotweld multiple layers of nickel strips... That'd be awesome !

I also plan to do nickel plating as soon as I can, and will give feedback after that. At the moment - it is possible to weld .2mm copper, just I have doubts (might be overstated) about durability of copper in a bike pack. I hope nickel plating will reduce energy needed for a weld, perhaps also improve mechanical properties of the tab a bit.
 
As seen in the ultimate 18650 repsository thread some are doing copper builds and other struggles. There have also been some testing of brass etc.
Is there an alloy that could bring some of the benefits from copper without the drawbacks?

Have anyone tested welding with cupronickel? I think copper nickel alloy comes in various "blends" and I know it is often used ie by oil & gas industry for offshore applications.

I think the resistance if cupronickel is fairly close to copper, but it will not corrode as pure copper will.
 
Copper nickel is available from china....https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20170420192059&SearchText=Cupronickel
 
macribs said:
Copper nickel is available from china....https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20170420192059&SearchText=Cupronickel

cupronickel seems to have very low conductivity, 5% IACS or something like that.

http://eddy-current.com/conductivity-of-metals-sorted-by-resistivity/

chinese substitute even more copper with Zinc that does not help
 
Back
Top