Spot welding Copper / Copper Nickel / Copper Stainless Steel Sandwich

I have not been welding across the center spiral on the EVE 40PL’s. But doing 4 pairs of welds working my way around.

note the two smaller welds on top on the left cell are only from welding the Stainless to the copper with gear 008, before placing the sandwich on the cell and welding sandwich to cell at gear 385 and the same for the center of the cell on the left.

I have not yet tried to weld across slots in both copper and stainless.

But I don’t think there is much concern. I believe the area between welds when flux is used, that it is basically soldered between welds, not welded. I do not believe this could damage the spiral central weld.

I could be wrong.

Recycling

2.8V or so would be great. Low enough to reduce possible issues when shipping but high enough for me to recharge if needed.

2 Likes

I am preparing to weld up a 10 21700 cell Makita battery with BAK45D.

Nelvick has laser cut me some 0.3mm busbars, which have the slit to weld across.

I have welded 0.3mm of copper before, but it was 0.15mm stacked or 0.2 with 0.1mm stacked, under 0.1mm stainless, with flux, and post 143 in this thread, I determined that gear 550 was enough. That was 0.10ms preheating, 02 intermittent doublepulse and 0.1second interval between welding pulses.

With 0.3mm copper under 0.1mm stainless gear 548, 04ms intermittent doublepulse and 0.2 second interval between pulses with flux, it stuck well, and took a lot of force to remove, but then popped off cleanly and left NO copper on the cell.

I bumped gear upto 598 and it did leave two equal dots of copper on the cell and holes in the copper, but not quite even on the stainless. A hole through the stainless on one side only.

This is with the slit which should lessen the power required.

But my impression is welding dual layers 0.15mm of copper requires less energy than a solid 0.3mm layer, all other factors nearly equal. Makes sense when one considers the resistance between the two layers of 0.15mm copper vs a single thickness of 0.3mm.

With 04ms intermittent( between preheating pulse and first weld pulse), and 0.2 seconds between welding pulses, my impression is I need to step up the gear at least 15 to 20 to 615 territory

There is just the slightest signs of the flux tinning the copper adjacent to the welds, So I believe this power level is just barely reaching the minimum required.


I just returned it to 2ms intermittent and 0.1 second, and the weld was weaker.

Rolling it off still deformed the can though( samsung 50E).

It Still took a lot of force to remove but no copper left on the cell. My impression is 4ms intermittent and 0.2 seconds between welding pulses yields stronger welds, if strength is determined by copper left behind on the cell. The force required to remove being able to deform the can might indicate the power is enough.

Dremelling a slit into the 0.3mm copper is delaying testing.

I am skipping the slit and just using more power for subsequent tests in this post.

I am up at gear 660, .10ms preheat, 04ms intermittent, doublepulse 0.2second interval, and feel it needs a smidge more.

….-.

..Ok gear 680 seems to be a good minimum with no Slit. the entire area between the welds is tinned. Distorted the F out of the can rolling it off. The way the electrodes seemed to sink into the strips just felt right, and I would have been surprised to find anything but two equal solid welds.

So gear 598 minimum with slit( prefer 615) , and gear 680 with no slit in 0.3mm copper under 0.1mm stainless seems to yield good result on a samsung 50E.

Preheating 0.10ms, 04ms intermittent, double pulse, 0.2 second interval between welding pulses.

I think the welds which took considerable force to pop off, but left no copper behind, would likely still be strong enough for most applications, but not Esk8.

On the Makita with 0.3mm slotted copper, I am going to use 3 pairs of welds, and start at gear 615 on the BAK45D

I am thinking, I don’t like the slots/slits. I think on the cells with a center spiral or the faint dot that the BAk45d has, I’d prefer a small round hole just for alignment of the strip and to help avoid the 6mm center diameter. I’d rather not have to weld across the slit.

I can see the slit/slot being useful when the welder is short on power, and if one has the laser cutter, but dremelling a slot even with a new diamond cut off wheel is a Pain in the keester.

Time to go for a roll.

2 Likes

The busbars which came with my 10 21700 cell Makita kit are just 0.2mm nickel plated steel.
They fail the spark test.

I got some 0.15mm nickel to be sure, which did not spark at all.

They just feel, and sound like steel when dropped on a surface, and feel way more magnetic than pure nickel.

Not going to bother with salt water rust test.

I tried to weld this on top of 0.3mm slot less copper stepping up the gear to 740, all other settings the same as post above

Not enough. Took considerable force to remove and then it popped off almost cleanly. two copper colored dots barely perceptible with a fingernail.

Gear 780 and one side tore the copper but the other side just a copper colored barely perceptible copper colored dot. So not enough.

The side which tore is not really big and wide enough that I would have much confidence in it, in an esk8 application, even if both welds were of equal size.

The tinning of the copper from the flux is almost a perfect 8, if both top and bottom of the number were equal sized

Stepped up to gear 830 and both welds same size and tore copper equally. Did not tear the NPS equally though. The copper is tinned in what looks almost like a perfect chain link around both welds. Very symmetrical.

I am not welding across a slot like I will be on Nelvick’s 0.3mm and the Kit’s 0.2mm NPS busbars when sandwiched, but it seems like:

AwithZ p20B

gear 830

Preheating 0.10ms

04ms intermittent ( pause between preheating pulse and first weld pulse)

double weld pulse

0.2 second interval between welding pulses, using flux, Is yielding Esk8 acceptable welds

on 0.3mm copper withg 0.2 nickel plated steel on top.

This is on the positive tab/ cathode of a Samsung 50E cell which has been welded to previously and had previous Nickel tabs dremelled off with diamond cut off wheel.

A cell which has had previous welds removed and the nickel plating removed always seems easier to get a weld to stick, as long as both strip and cell are flat.

Thicknesses of previous welds, unless removed so strips can sit flat, changes the dynamics and adds too many variables in my opinion.

Gear 830 is a lot of power, the weld leads really jump, an I keep them close to each other but not attached. I try and hold the weld pen bodies more like an H rather than a V, as I feel I can more evenly exert the same pressure on both electrodes, and have their rounded tips be more the same shape pressing the sandwich to the cell and increase consistency and repeatability.

The reason I am considering using the kits NPS busbars on top of the Copper busbars is for simplicity. I can cut off the balance tab reaching for PCB BMS and easily weld sandwich together, then place them on the 4 flux’d cells and turn gear back up to 830

I do have Slotted 0.1mm stainless slotted circles I can position on each cell too, but these are a little bit big for the cathodes, and the slot is not lining up perfectly, causing the outer edge to catch on the cell holder body.

I’d rather use a lower gear(615) and the 0.1mm stainless opposed to gear 830 and 0.2mm NPS, but the latter will save a lot of time, and I have no other use for the Kit’s busbars.

The 0.2NPS will add a smidge more ampacity but it is 0.3mm copper below it.

If I were an electron I know which path I’d choose.

I wound up dremelling some slots in both 0.2NPS and 0.3 copper, and used gear 830, and welded to the 50C can, and Ieft symmetrical copper dots on the cell, but one side was tinned more so than the other side, but this could have been unequal amounts of flux on either side, since the flux can squeeze out the slot and not spread evenly.

Again, I am becoming less and less of a fan of the slot, for either sandwich alignment purposes, or for lessening welder power required.

Welding onto cells previously welded to, and dremelled free of those welds is also another variable as there is no way the scratches left on the cell and the thickness of remaining nickel outside those dremmeled areas will never be the same.

err on the side of power. it seems can puncturing is unlikely if one is using 10% more power than technically required to achieve minimum esk8 welds.

So it comes down to is this higher power level going to negatively affect cell longevity and performance in any meaningful way?

It will affect welder longevity, but again, in any serious way?

I don’t know.

1 Like

The cool looking capacitor powered welder on batteryhookup can do .3mm copper under .1mm stainless with no prep and no flux. Cheap.

1 Like

Having some very good results so far with 0.1mm nickel over 0.2mm copper using the diy500amp flux. In this case I got good results with the p30c in gear 400, one pulse per weld, and 0.5ms preheat. Another benefit of the .1mm nickel is it looks to be thin enough that you can run a sharpie or other hard plastic tool around it to indent the pattern of the top of the cell into it for easier weld placement.

7 Likes

It can’t get better than that.

2 Likes

what percentage of the battery’s resistance is from the welds? I still am wondering the value of doing a couple more welds on each cell.

The batteryhookup welder rocks. crushes the Amazon black box I had. I’m freaked wondering can the probes short to the metal on the bottom of the big caps?

I have it wrapped in a towel after having two other welders short to electronics debris. Never again

1 Like

Cool to hear this may be a viable unit for a reasonable price.

I have ordered a LOT of batteries from BH over the years.

They have a 3dp base for that welder:

1 Like

When fishpaper will just not do, ….. bust out the bath towel.

I apologize for not getting welded samples to mooch yet.

It is on the top of my project list.

My lower back is not cooperating when standing after sitting. So I’ve been avoiding sitting.

2 Likes

I’m up there at “a lot” from batteryhookup. At least 3500 from back when they had essentially new a123s and then the robot batteries.

1 Like

I have some 0.2mm nickel plated copper that Nelvick sent me.

AwZ P20B

I started at gear 500, 0.10ms preheating, 04ms delay, 0.2 seconds delay between double weld pulses.

No Flux, No Sandwich, No Chance.

Gear 600, same.

Gear 700, One solid weld which tore the npcopper exceptionally easily, and one non existent weld

Gear 800, two tiny pin pricks of strip left on cell, but took very little effort to remove. I bet another 80 to 100 gears might be enough, but I didn’t try.

Pure 0.2mm copper, same exact settings, same exact results.

Add flux same settings under 0.2mm Nickel plated copper= 2 solid welds

Add flux under pure 0.2mm copper, again, 2 solid welds.

I can tell no difference between using 0.2mm copper with or without flux no sandwich, and Nickel plated 0.2mm copper both with and without flux, no sandwich.

I expected the Nickel plated copper to weld solidly with less power, but the nickel plating is super thin and easily scraped off.

So functionally, there appears to be not one whiff of difference in the power required to weld them

For reference, I use gear 386 for 0.2mm copper under 0.1mm stainless with flux, and on some cells gear 320 was enough, So SSSandwich and flux, require less than half the power as No flux no sandwich does.

I do not think 0.2mm copper no sandwich even welded solidly with flux at gear 800, is strong enough for Esk8 applications. The copper tears away from cell too easily, but when there is 0.1mm stainless on top it is easily strong enough.

It is as if like without the stainless on top the 0.2mm copper is greatly weakened.

I have chiseled the stainless from atop copper, and then peeling off the copper by itself, it is significantly stronger.

1 Like

Be nice to also see a comparison with the available .3mm copper and nickel hybrid stuff off alibaba, as well as comparisons between different number of welds. Could weld like 10 times and maybe it’s worth it. I imagine the welds are always the bottleneck.. until u get enough down

same here for the 0.2mm nickel plated copper I had bought via ali - I could not weld it standalone without a sandwich

1 Like

@BenjaminF got the loose cells and mis batched box! Thank you!

Going to start prepping for comparison tests for @Battery_Mooch.

Was thinking of doing 6 and 8 examples of each test i wanted all in .3 copper or .25 copper - but all the same. 4 welds. Two tabs per side per cell. So depending on how consistent they feel 3 or 4 cells for each weld process.

  • clean copper
  • With flux
  • With ss sanwich
  • Both ss and flux

Recording the settings for each set so a good reference for how much energy is getting pumped into the cells per each method.

Then i wanted to do nickel at 2,4,6 and 8 welds. weld comparison with nickel as i want max contact area on each and 2 tests per cell or three cells per weld set. So 12cells of weld tests to see If how area of contact and weld count affect the resistance.

Any suggestions for things to test or try let me know. Ben was very generous with extra cells to test with so there’s a few more than needed at present.

4 Likes

I’m glad they arrived safely! Looking forward to your test data, specifically the “without flux” tests per our discussion :face_blowing_a_kiss:

3 Likes

I’ve been lurking on this forum for awhile to pick up hints and tricks to help me with building my own packs for my rc airplanes. Question, where do you guys buy your copper busbar from?

1 Like

Aliexpress

Thanks for reminding me.

I just ordered a meter of 0.25mm thick 100mm wide copper, from ebay.

I cut my own busbars with tin snips.

Perhaps it is cheaper on aliexpress.

Their website annoys me.

1 Like

I want to solder the balance leads to the extended SS-layer of the sandwich.

What is good stainless steel soldering flux, please?