21700 5000mAH Battery recommendation?

Recently I built a new flashlight with my driver which I’m using a Molicel INR21700-P45B With 3S1P Configuration for the battery subsystem. These batteries works fine but it only has 49.95Wh which is too less for a 3*XHP70.3 flashlight. So I want to use 3S2P configuration which have 6 batteries to double the total capacity but I don’t have an idea about which battery is suitable for this.
The light will consume 20Amps of current at turbo,About 10 Amps for each battery. And my plan is using 5000mAH 21700 cells(but not SAMSUNG INR21700-50S due to shorter life and expensive)So does anyone has a recommendation for what cell should I use? Thanks

The LG H50, BAK N21700CG-50, and Molicel P45Β are your next best choices after the 50S. All have essentially the same run time at 10A/cell.

Here’s a table I did with the (ongoing) results of my testing. The “E-Scores” are the actual delivered-Wh down to 3.2V at those different discharge current levels: 20700 + 21700 Battery Ratings and Performance Table | E-Cigarette Forum.

Makes comparing different cells a lot easier. Especially since capacity is only a verrrrrry rough way to choose the best cell (for run time).

The other 5000mAh 21700’s just have too high an internal resistance to give you any more run time over the P45B’s you’re using now. Their voltage sag just affects the total run time too much.

4 Thanks

Thanks for your recommendation, seems like I can use LG M50LT for 3S2P in this case? It has nearly same E-Scores than P45B at 10Amps discharge in this table and have higher E-scores when running in 5Amps(which represents lower gear for the light for more frequently used condition due to heat limitations).

My apologies!
Yes, it’s also one of the cells you can use. Its score is essentially the same as the others I mentioned. Anything under about a 5% difference in the E-Score just isn’t noticeable IMO in actual use. Too many other variables affecting run time to sort out that it’s a tiny difference in cell energy.

I would rather buy another 3 Molicells if you already have 3, instead of 6 other cells…

1 Thank

Good point…it would have the same run time as the other cell choices but would also have the same discharge curve as his first pack, helping to reduce the balancing current flow between the two packs as they discharge.

Not a big deal but the Moli’s are great cells and there’s no run time penalty using them.

1 Thank

Currently I have three Molicel for my original pack. And the battery subsystem is constructs as Battery Holders(each holder can hold 3 21700 Cells in series with protection and balance system)So the original pack has one holder and 3 cells.
My plan is add another holder so the pack will have totally 6 cell in 3S2P and all of them will be M50LT with nearly same batch during manufacturing and capacity. This might be better than buying another 3 molicels(might be another batch and have slightly different electrical statistics)?

In my experience any between-batch differences in cells from the major manufacturers (with the same revision number) are tiny and not of any concern.

It has to be that way IMO as the big commercial customers for these cells would not accept any variability.

But if your Molicels have more than, let’s say, a hundred cycles on them there might, maybe, perhaps, possibly could be some small differences between them and a new set. This would result in a bit more current flow between the packs as they discharge, but not a lot and certainly not near levels where heating or safety is an issue IMO.

You can go with six new cells if you’re at all concerned though with the M50LT’s and 45B’s giving you the same run time. The 45B’s might have the longer overall cycle life as they will be running at a much lower power level compared to their rating.

2 Thanks

I would stick with the Molicel P45B, or use the Samsung 50S. I think both of those are better than the LG M50LT.

I have read reports that the M50LT has a weak supporting structure for the positive terminal and that they deform under relatively small pressures. I’m not 100% certain it was this cell – maybe others here can confirm or deny this.

1 Thank

It is indeed a thinner structure with narrower supports. This may or may not be an issue depending on how much pressure his holders exert on the cell. It certainly wouldn’t hurt to use an equally performing cell with a beefier top structure though.

M50LT can: Imgur: The magic of the Internet

2 Thanks