High AMP battery build

[martinmertik]

Hi, I’m building high power Honda CRF 250R dirtbike. I will use Zero 75-7 motor and 20s 27p Sony VTC6 battery pack. I would like pull 500+ amps from the battery with minimal voltage sag and overheating.

So my plan is to use 0.3mm thick nickel sheet. One nickel piece for whole 27 cells inside parallel group spot welded to cell terminals. I would like to attach copper sheet with the same size and shape over nickel sheet. What copper thickness would you guys recommend? Should I maybe go with thinner nickel, maybe 0.2mm, and even thicker copper sheet?

I know that attaching copper with nickel is paint in the ass, so I thought about drilling random holes in copper and then soldering it to the nickel plate. What do you guys think of that?

I’m open for suggestions.

 

[amberwolf]

Hi, I’m building high power Honda CRF 250R dirtbike. I will use Zero 75-7 motor and 20s 27p Sony VTC6 battery pack. I would like pull 500+ amps from the battery with minimal voltage sag and overheating.

At nearly 20A per cell, that's going to have some sag already just in the cells, down to around 3.5v per cell, when full. That's quite a bit more than if you only pull <10A per cell, which will sag to around 3.7v per cell at the same SoC.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=99016


Let's see if my brain can do math today (meaning, this could all be wrong; you should check it to be sure): Assuming the same Ri as tested in that thread, around 14mohm per cell, then with around 0.2v drop in the cell at 20A, the power loss in each cell is almost 3 watts (2.83w). That's a fair bit of heat. In one parallel group, that's 27 x 2.83, or about 76 watts. Times 20 series groups is then about 1520w of heat inside the pack from cell heating alone. Personally, I wouldn't want to heat my pack up like that.



Regarding attaching the copper to the nickel, isn't there at least one thread just all about that, that might have info you could use? It's probably either by or posted in by Spinningmagnets, if that helps.

 

[John in CR]

...and all that sag Amberwolf is showing you is coming just from the internal resistance of the cell, so it's all turning to heat. Do the math and heat problems are guaranteed at 500A even with perfect connections to and between the cells.

If ridden fairly easily with short bursts of high current the pack may perform perfectly well, but get in some mud or up a long climb and high current draw becomes much more continuous.

 

[martinmertik]

I exaggerated about 500+ amps getting pulled from the battery. This would be the worst case scenario only possible with different controller.

Absolute peak is going to be more likely at around 400-440 A which is roughly 15A a cell which is not that bad.

I will definitely check out threads about joining both copper and nickel. What nickel and copper thickness would you guys recommend? Thanks both of you for you answers.

 

[amberwolf]

Absolute peak is going to be more likely at around 400-440 A which is roughly 15A a cell which is not that bad.

That's still a bunch of watts of heat in the cells alone, even though it is only about half the sag of 20A vs 10A.

BTW, what I was calculating up above for power loss and heating is JUST the DIFFERENCE in heat between 10A and 20A--it is not the total heating that happens inside the pack (you'd ahve to calculate based on the total votlage sag at any particular SoC, as well as the cell Ri at that SoC).

But I might've made a mistake; I think I forgot to calculate in the 27 parallel cells, and used just the resistance of a single series of cells, which is 27 times higher than what the pack would actually have.

So calculating using:
-400A current
-cell resistance 0.014ohm x 20 in series is 0.28ohm, / 27 in parallel is 0.0104ohm pack resistance.
-actual voltage sag (not just difference in sag between discharging from 10A to 20A/cell) at nearly full is about 0.5v+ per cell, x 20 in series, is about 10v sag at pack level.

I think that comes to around 1600 watts of heat generated inside the cells themselves, at 400A.

Using 500A, I think it comes to around 2600 watts total pack heat. (which makes some sense, if it's mroe than twice the voltage sagging).

I could have all the math wrong, so you should check this yourself to be sure I'm doing it right.

 

[DogDipstick]

Alot of plug welds to solder the copper to the Ni sheet properly to transfer that much power reliably... I would think you might be better off modifying some other way.

How fast are we thinking this machine will go?

Plus the draw is alot on any cell? Why run near the ragged edge.

I agree with those experts above... You NEED a lower IR cell.... for that in the package.

 

[Upperfoot]

I would have to recommend the Samsung 20S or Samsung 30T or 40T if you want to tackle this, they punch a lot harder than the VTC6