Design help?

Hi all, I'm a first time battery builder and was hoping someone who knows more than me (everybody) would like to comment on my design. It's a 28S20P 18650 split into 2 separate packs wired in serial to provide 115V. There's 280 cells (14S20P) folded like a taco in a 10X9X6 inch package times two. I've got a Kweld and honeycomb holders on order and that's about it. I'm stuck on choosing the right nickel strip and bus design that'll be sufficient for 250A max?
View attachment Battery layout.pdf



I'm thinking this will work and .15mm is easy to source but I'm not sure it's thick enough. I know I'll need some kind of copper/nickel bus strip but I haven't figured out how to go about it.


Thanks!

It's not so complicated. At 20P, you are asking for 12.5A per cell peak.

The 2 biggest considerations:

1) where and how you tap at each end for your loads, the B+ and B- for the whole 115V pack.

And

2) how the cells are interconnected in the middle

If in the row between paralleled C1 and paralleled C2 through all serial connections, you have a strip between every C1+ and C2-, then that's the 12.5A per strip. 1/20th of the requirement. This pattern is copied through C13+ and C14-. For all middle rows.

I'm no expert on welding, but that looks like what the strip pictured is designed for. If it's an intermittent 250A load, this is probably OK. If continuous, looks thin, how thick are the strips?

However, the pack taps for C1- and C14+, deserve special attention. Perfect balance would be achieved by tapping each C1- cell with a piece of wire, then bring all 20 wires to a single common B- pack terminal. Duplicate for the B+ side. Each wire is again carrying 12.5A, so #18 or #16 is sufficient. But that's a lot of wires. If you cut down 20 wires to 10, that means the taps are carrying 25A each. Probably want #16 or #14. This will be fine. Every third cell for like 7 wires is OK too at 35A, but now #14 or #12. Beyond that, the nickel strips making the parallel connections for C1- and C14+ could get overloaded, as their loading exceeds the serial connections.

Start carrying the full 250A in a single wire, that's #6 minimum, and only if very transitory. A full time 250A load calls for #1 minimum using 200C silicone wire. All wire sizes mentioned are AWG. #4 Likely a reasonable compromise.

Instead of getting the grid-shaped nickel like the one you posted, consider buying just a wide nickel strip, that way you have a much much bigger cross-section.
0.2mm nickel will probably work but since you have a Kweld, which can handle 0.3mm, I'd go with the latter.
As for the current dissipation, could've not explained it better than BlueSeas.

This is just what I was looking for, thanks! I like the one wide strip idea, according to this chart my 2P strip won't cut it.

Then maybe a long strip of copper connected to each cell as a bus down the side?

Since you've been kind enough to help me out, I've got another one: If I need 12.5A per cell for 250A (3 minutes max) and 2.5A per cell for 50A continuous, what's your favorite 18650 cell for this?

That's an interesting table. Copper definitely wins the war.

My Headway cells came with silver colored bus bars. They are not magnetic, so that eliminates iron and steel. Too heavy for aluminum. Short of sawing one in half, and looking for copper, is there an easy way to tell the difference between 100% nickel or zinc or stainless?

I think they are nickel plated copper, but that may be wishful thinking.

Copper has gotten so pricey that my old boss doing marine work is stripping the copper out of remnants #6 and bigger for $2 a pound. Bought a machine that removes the insulation, since if that's still there it's only .50 cents a pound.

BlueSeas said:

My Headway cells came with silver colored bus bars. They are not magnetic, so that eliminates iron and steel. Too heavy for aluminum. Short of sawing one in half, and looking for copper, is there an easy way to tell the difference between 100% nickel or zinc or stainless?

I think they are nickel plated copper, but that may be wishful thinking.

Click to expand...

Tin plated copper is very common for commercial bus bars (the tin serving as anti-corrosion plating). I'd give it a little kiss with a file on a corner or edge of the bar, and see whether copper doesn't show through.

Pure nickel (always) and stainless (usually) are both magnetic materials, just not as strongly as iron or carbon steel. So you can use a strong magnet to eliminate those options. Zinc is wildly unlikely to be used for these things except as a plating for steel.

You could spot weld nickel for the parallel connection and then solder on copper for the series connection.
However, you need proper soldering experience and a very strong soldering iron to not bring too much heat for the cells, or else you will damage them.
This person did it with great execution, and I'm pretty sure I'll do it too for my next build

Chalo said:

Tin plated copper is very common for commercial bus bars (the tin serving as anti-corrosion plating). I'd give it a little kiss with a file on a corner or edge of the bar, and see whether copper doesn't show through.

Pure nickel (always) and stainless (usually) are both magnetic materials, just not as strongly as iron or carbon steel. So you can use a strong magnet to eliminate those options. Zinc is wildly unlikely to be used for these things except as a plating for steel.

Click to expand...

Good idea...thank you....there's copper under the plating. Good news!

Good video, thanks for that. Now I've got another YouTube rabbit hole to go down. I'm guessing he had to solder the copper because it's difficult to spot weld but I'm going to try not to after watching this:

https://youtu.be/byvr-dwwao4

According to the chart it seems like one wide strip of nickel will do it no sweat, maybe pre-solder split copper bus strip then spot weld?

You may want to take a look at the busbars I designed. I don't like nickel strip for high current applications and would much rather have copper.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=108833

Thanks for that Mr. Sanchez, it looks very similar to what I'm planning and well done. I've yet to find a suitable copper bus like that since I'm using the "diagonal" style holders due to space constraints, do you have a good source?

No, I designed these and had them made. If you are pretty capable on solidworks etc. then you could probably design your own. The copper for mine was around £500 including duty so not exactly cheap, but it barely gets warm under load which was exactly what I wanted.

Shame you aren't going with straight as I still have some busbars left.

Any advice on tools for selecting the best 18650 for my application? I found this spreadsheet which seems like the best thing going so far:

View attachment Cell comparison spreadsheet.xlsm

Any suggestions on US-based suppliers as well?

Thanks!

To be fair, selecting the right cell isn't too difficult. You are fairly limited in choice due to the high current (250a) requirement.

Are you sure you need 250a? If so is that constant or would it be for a very short period of time. Your equipment may be capable of drawing that much but what size is the motor and what is it's constant wattage rating. You may find you don't need as much as 250, you may even find you need more!

The reason I ask is because you could potentially use a samsung 35E cell which is slightly more energy dense for the same cost. However that'll only be pushed to around 10amps per parallel cell (200a constant)

Otherwise you'd be better off with a Samsung 30Q or Sony VTC6. Both of which can pulse to 30a and are happy at 15a which would be enough to fulfil your requirements. They are slightly less energy dense but are solid cells which I tend to prefer. Hope this helps.

My design goal at this point is for 250A max for no more than 2-3 minutes so 12.5A per cell, which I believe is longer than what's considered a "burst"?

The 3500mA LG MJ1 is appealing but at 10A max continuous, I'm guessing it would be over-stressed.

Thanks.

Those really just aren't realistic ambitions if you want any longevity at all

Here's an example of true experts discussing the upper limits of LiPo performance

which wrt to power density are light years beyond cylindricals.

https://www.rcgroups.com/forums/showthread.php?p=46317309

Thanks for that John, but I've written off Lipo for a number of reasons. I'm pretty much locked in to the 18650 format.

I was not in any way suggesting you use LiPo.

My point was, if those C-rates are not practical from LiPo units that are head and shoulders above normal LiPo

that should make you realize you need to go back to the drawing board with cylindricals.

If you can't reduce the continuous power needed - and yes minutes at a time are not "bursts" but continuous

then you'll need to carry higher Ah, more space and weight for the pack.