I just got a couple of these in. Will see how they do.
Unless the conductors are too small and heating signficantly, I donāt think any appretiable amount of the sag is from the conductors. Nickelās resistivity is 7x10^-8 ohms*m. For .2 mm nickel, I consider 1 cm of width as good for about 30 amps. That means a cross sectional area of about 2x10^-6 m^2. For a 20 mm long tab, thatās a resistance of about 7x10^-4 ohms. At 30 amps, that is a 0.02 V sag per tab. Iām not too worried about that.
For now, I think Iāll stick with 0.2 mm nickel, and adjust the tab size to suit the current. Iāve worked with 0.2 mm nickel, and I know how to make it work with what Iāve got. It seems like even for these new cells itās possible to make a pack that can use the full potential of these cells with that, but youāre getting close to the edge and have to use basically all the are you can for series connections.
3 Likes
Copper is usually cheaper than nickel, gold is used for corrosion resistance. Silver is used because itās oxide is conductive and generally doesnāt increase in volume so oxidation doesnāt damage connections.
If a 12s pack has .02v drop per tab at 30a and 24 tabs thatās approximately .63w per tab or 15w of heat in your battery - nothing burger
On a bigger 200a battery run at 180a - total common for a high power build for me to run 90 battery amps but your mileage may vary. That .126v drop or about 22w per tab or 528w of heat just from the tabs.
Copper should be somewhere in the range of 4x lower resistance. Not a big deal at 30a but can save a significant amount of heat and wasted energy the higher your draw. Especially important in low p group builds that the new cells are making more viable.
A 12s4p using the same tabs nickel will be further constrained because of the potential smaller cross section of the current path combined with the smaller number of cells to absorb that heat and dissipate it.
Copper is the standard best material for conductors for a reason and trying to use these new affordable higher power welders to develop a better system of battery building is kinda the whole point.
Love the counter argument though, donāt miss read me. Putting out the argument against helps the conversation and definitely shows where you actually get the advantage from a material change.
An easy way to weld copper to cells to just make a standard will help improve all battery builds but low power setups get the least advantage
7 Likes
On a bigger 200a battery run at 180a - total common for a high power build for me to run 90 battery amps but your mileage may vary. That .126v drop or about 22w per tab or 528w of heat just from the tabs.
My calculation was assuming 30 amps per cm of tab width. As long as you keep the tab width to current ration constant, you wonāt get any more sag. You will get proportionally more heat as the current goes up, but the sag will be constant.
If I do 240 amps with a single 80 mm wide tab, or 480 with two 80 mm tabs, one on each side, I should still only have 0.02 V of sag due to each tab. I will have about 5 W per tab for an 80 mm tab at 240 A or 10 W for two 80 mm tabs at 480 A instead of about 0.6 W for a 10 mm tab at 30 A, but still the same sag.
Like I said, itās approaching the limit, but as long as you use a full width tab, I think you can do 60 amps per cell on 0.2 mm nickel without too much trouble. Hell, if you were doing single stack and used tabs on both sides, youād probably be okay up to about 120 amps per cell. Also, for brick packs, that means 60 amps per cell is about the limit on 0.2 mm nickel without coming up with something overly complicated.
How dare you display any irreverence towards 0.2mm pure nickel!
What has been more than good enough for everything up to now will continue to always be so! Period Amen!
What if you can weld 0.25mm copper by itself, no brazing paste, no steel sandwich BS, as easily as you can weld 0.2mm nickel now?
Did I hear someone mutter something about prying their respected trusty proven spotwelder out of their cold dead hands?
I totally get that too. My mechanical and electrical empathy needs a series of head doctors to sort out. Theyād ultimately fail, and then question their desire to figure it out in the first place. But it could make for interesting conversation for the very few who actually cared enough about the subject.
I personally canāt let my cells breathe through a cocktail straw when running a sprint or a marathon, even if technically they can do so, and even when I have no intention of making them run either, near their limits, ever.
I donāt consider putting 0.1mm of stainless steel strip, on top of a pure copper strip, to approach anything near 'āoverly complicatedā.
If I were doing it all day for a living, then Iād figure out a way to make sandwiching even easier than it is.
1 Like
Is 0.25 mm copper actually just as easy to weld with no special prep as 0.2 mm nickel? I thought it was much more difficult. If that is the case Iād be more interested in it just to save money on materials. I was under the impression it didnāt weld nearly as easily, though.
Also, I just donāt think it matters much as long as youāre within reason. I donāt really think going from 2 W per cell of heat from the tabs to 0.5 W or whatever is going to matter much when drawing over 200 W per cell. If weāre going to worry about that, we should probably never use plastics or composites for enclosures because Iād bet switching to a sheet metal enclosure and putting you battery on it with a thermal pad or thermally conductive paste or epoxy would do a huge amount more or a hundred other things. Even through shrinkwrap, fishpaper, etc, that would probably change things more than a couple watts per cell.
I could be mistaken. I donāt actually know how much heat cells are producing at high loads, or the thermal capacity of a cell, which I could use to estimate that based on how quick they heat up under use. I get the impression itās enough that a couple of watts per cell isnāt much, though.
Edit: Also, for this build, Iām really understressing the cells, so they should be very cool anyway. I actually wanted lower power and higher capacity cells, but there arenāt many good options right now, especially for a good price in the area I was looking. I really wanted a cell best suited to about 15 to 20 amps of draw, but high power cells are still the best there.
I was able to weld 0.25mm worth of copper, using brazing paste, but no steel sandwich, using gear 850( of 999) 0.15ms preheating, double pulse with a 4ms interval between pulses using my AwithZ P20B 14.6 kW welder, which is about 240$
Scroll down to post 5 for photos
I can weld 0.2mm copper without flux too, but I donāt feel 0.2mm copper is strong enough, it tears pretty easily.
Stacked 0.25mm copper however did feel strong enough on its own.
I believe even spot welders with less power can successfully weld pure copper, especially with the use of the Flux/ brazing paste and/or -0.1mm Stainless steel sandwich.
My previous 21$ cheapo welder was able to do 0.1mm copper, 0,1mm Nickel plated steel and I bet 0.15mm copper would have been in play if I had tried 0.1mm stainless steel instead of 0.1 nickel plated.
That requires me to buy a new welder or experiment with a bunch of different combinations which is not straightforward. Also, itās counterintuitive to me that steel would work better than nickel plated steel, or that either would work better than pure nickel in combination with the copper. Is the goal to heat up the other material moreso than the copper?
I donāt think itās a bad idea to look at for someone just starting out, but I donāt want to buy a new welder, or do a whole project trying to figure out what works for this battery, and Iām not planning on doing more than a battery every few years.
2 Likes
Stainless steel has higher electrical resistance than regular steel or nickel plated, so it gets hotter under the electrodes, and allows for lesser welding energy to sandwich weld the copper to the cell, than would an equivalent thickness of NPS or Pure Nickel.
The steel can be only a cap over the welded area, it need not fold over the top like one would with 0.2mm nickel.
It is a strong possibility that your current welder can do 0.1mm copper under a 0.1mm stainless steel cap, and that should be equivalent resistance to ~0.4mm pure nickel.
I wish I had more stainless steel on hand, and different thicknesses of it. I am still eager to experiment.
Your initial post was concerned with what if you ever wanted to push 8p of Bak 45D to their limit, one day. I am responding to that desire.
I am not pushing my 2p of BAK 45D to their limits, but excessive temperature of my old battery, even with 0.1mm copper, scarred me, and it certainly seemed the flat pack was getting hottest near its ends, where the copper strips were, despite my limited max amperage.
I was thinking about how best to extract heat from that pack, and decided the minimizing the heating was better than trying to extract it via convection or conduction.
My initial post was that I want it capable of at least 240 amps, but that is only 30 amps per cell. I do think I could make it capable of 480 amps without too much trouble, but Iām not decided on that. Basically I wanted to sanity check that an 80 mm wide strip of 0.2 mm nickel should be good for about 240 amps. Assuming that is fine, I was just on the fence about whether to do one strip on one side, or two on both so I could use more power if I somehow wanted to down the line.
Also, I wanted to check that it would be a bad idea to try to use the Mboards strips at anywhere near 240 amps.
Stack welding nickel seems to be a method which has been looked down upon, on this forum, as ānot best practiceā, for a while, If one strip both top and bottom is in the cards.
Hope you find the best solution which works for you.
No, Iām either buying Nickel wide enough to do a single layer, or Iām going to do the two halves of each p-group separately, and then run something between them to keep the two halves equal. Right now most of my nickel is 0.2 mm x 30 mm, so I may just use that and connect each half of the p-group with nickel and then connect the two halves with braid.
My 21S4P P42A brick pack at 180A (so 45A per cell) does get to 80C on my raceboard before itās empty, so I canāt just ride that pack without constantly checking the bms app. I switched to lipos just recently but the long term solution will be a big tabless cell pack with welded copper and thick wires and beefy connectors.
Ironically the 20s lipo setup I run now heats up the 10awg wiring quite significantly, much more than the lipos heat up themselves.
1 Like
Yeah, my point is just idk how many watts of heat each cell is producing. I know they get hot if run hard. Iām pretty sure they make more heat than the 2 W Iām estimating from the nickel, but idk if itās in the same ballpark or if itās an order of magnitude more. Iād actually be inclined to guess itās more like an order of magnitude more, but Iām not 100%. That is just from guessing based on power resistors in comparison to battery cells, among other things. To really decide how much value there is in changing conductors, you either need to test them side by side on the same cells, or get some idea about how much heat the cell itself is creating, how much heat it can get rid of at operating temperature, etc. If the cell is making 20 W of heat to get to 80 C then another 2 W is small potatoes. If it is only making 5 W of heat from the cell, those 2 W start mattering a lot more.
1 Like
You can check out my test reports for the DC (ohmic) resistance of each cell model but that only contributes to part of the heating. There are exothermic chemical reactions going on too.
To complicate things, the overall resistance drops as the cell gets hotter due to increased chemical efficiency and ion diffusion rates at higher temps.
I measured 13mOhms DC (ohmic) resistance for a P42A at full charge in my testing. At 45A that would be almost 26W per cell just for the heating from ohmic resistance.
My discharge graph for each test report has the max temp for each discharge current level. After the first few seconds the temp rise is very linear over time and easily estimated at various points along the discharge.
That might help you with the data youāre looking for?
7 Likes
The copper side quest is definitely just that, exploring some new concepts to see if it can be done better.
Imho you are right on with your design and the nickel tabs. One of the reasons I stopped using the mboard precut (bioboards too) is because of the lack of a full width tab which was inadequate for some of my builds
No need to experiment with copper, i think youāre in the right ballpark for your build and definitely post some pics for me to nerd out on. Love that shit
BUT i want to mess with the copper. Ive got some time and supplies in the mail (amazon delivered an empty package (ąøą² _ą² )ąø so im waiting on parts still). Ill be posting up what i find and hopefully the group of us can find a good cheap effective method of using it and finding a way to make it not all fiddly.
Probably the biggest factor for me not using copper sandwiches previously is that my welder does nickel great and all the kinks have been worked out already for using it - and it is more than up to the job in 80% of uses. Only a few edge cases really NEEDED more conductive material and .3mm thickness material works if just a bit more difficult with my previous setup.
5 Likes
Cell group 2 having less capacity , with the BMS bypassed for discharge, would only become dangerous, I think, if the other P groups kept total pack voltage above the programmed vesc cut off voltage, and allowed cell group 2 to be discharged below 2.5v.
The BMS probably would just refuse to allow the battery to be recharged at that point.
So I donāt believe it to be dangerous, unless/ until pgroup 2 is discharged below 2.5v, but that might happen after a few more cycles if that Pgroup has a cell which jumped off a cliff, or perhaps a broken weld.
But wait for more experienced answers.
2 Likes
anyone got any tips on how to create this slits so the weld is better? @SternWake how did you weld the 0.35 copper (with flux) but no slits?
Was thinking about some sort of slit maker like a metal punch in this shape. but havent found a good one yet on aliexpress.
I just upped the power. The slits do force the welding current to travel through cell can, and thus reduce the power required to achieve the same strength weld. Iād probably use it if I had the ability to easily achieve it.
With tabless cells, I kind of want to use a round hole instead of a slot, because one should not weld on the center weld spiral that exists on the EVE 40 and 50PL, the Ampace JP40, and the Reliance rs40 and 50. I avoided a similar size area on my Bak 45Dās too.
Got to make sure the electrode does not slip into the slots or you can blow a hole into the can.
Do a search for āinfinite slotā over on endless sphere.
My first battery build I tried to use infinite slot, but had issues not using enough power. the settings i established on test cells with the nickel plating removed, did not translate to cells with virgin nickel plating. infinite slot was time consuming enough I just used thinner copper and more welder duration, and have said screw the slot ever since, but it definitely has merit.
I have been looking into round hole punches the size of a weld spiral, but should probably learn some CAD and send some files to Nelvick to cut on his laser cutter
1 Like