Why should not 12s2p 21700? (12s2p 21700 vs 12s3p 18650 vs same but 10s)

Hi there

My lipos (12s1p 5800mah) are 5 yo now, so its time to change them. I’m seriously considering Molicel P42A 12s2p as replacement (or 10s2p as im not sure i need such high speeds as i currently have but due to sag even on 12s ill have less top speed).

I understand that my replacement has less power, more sag and probably less range, but i don’t know how much, so advise needed from the ones who had such battery or just experienced guys)

Should i go with it, or i need to step up to 12s3p 18650 (or 10s3p)?

  • 3p for 21700 is too much for my budget and dimensions and weight
  • I don’t care about range btw(but care about power). i would easily trade rage to lightweight But i don’t want lipos anymore.
  • current motor and 12s battery gives top speed of ~50km/h. for me 45 is enough, so a some sag is fine
  • 1.9 - 2.2 KW is sweet spot for me (with 47km/h top speed),
  • Ill run discharge on batteries of about 50a max.
  • smaller battery dimensions is better for me

So, any advice please?

1 Like

Looking at the actual energy delivered (Wh) by the Molicel P26A and P42A at 60A (no 50A data) you get about…

18.6Wh with single 3P group of Molicel P26A 18650 at 60A.
17.4Wh with single 2P group of Molicel P42A 21700 at 60A.

That’s fairly close but the 3P Molicel P26A will run for a bit longer, hold its voltage up a little better, and perhaps a bit cooler since the heat is spread out over more cells.

Looking at the DC internal resistance (DC IR) numbers you get…
6.2mOhm for 3P of Molicel P26A.
9.0mOhm for 2P of Molicel P42A.

You’ll get a bit more power, less voltage sag, from a 3P Molicel P26A setup versus a 2P Molicel P42A setup. This assumes equally good welding and wiring for both packs.

The 3P Molicel P26A pack seems to be the slightly better choice. This is not taking cost, shape, or weight into account…just performance,

Love the new tables! Makes choosing the better cell so much easier.

https://www.e-cigarette-forum.com/blog-entry/18650-battery-ratings-table.7447/

https://www.e-cigarette-forum.com/blog-entry/20700-21700-battery-ratings.7827/

18 Likes

Thanks a lot for advice. Love your tests btw - best source of info on batteries

5 Likes

Thank you for your kind words!

2 Likes

I can only give you 10s data which might not even help but i would go with 3p.

I have a 10s2p p42a pack and at 50A it sags about 6ish volt.

But mooch basically said everything already ^^

3 Likes

Since you dont care about losing the top speed,
I would go 10s3p 18650 as opposed to 12s2p 21700 because you’ll get less sag. (And you also mentioned you ride hard)

12s3p 18650 is ideal if you can afford/can fit 6 extra cells

1 Like

Maybe mooch can correct me but napkin math:

17.4x12 = 208.8 real world Wh
3.7x5.8x12 = 257.52 absolute best case “rated” Wh

A unity murdered my lipos before I had a chance to fully live with them for too long but I’d be surprised if you got longer range from the lipo than a 12S2P P42a, and even more so for the ~223 real Wh 12S3P P26a

4 Likes

Thanks a lot for your help, i have done some approximate calculations to confirm all above using @Battery_Mooch tests (thanks for them)

I have compared 2p configuration for 21700 molicel and 3p configuration 18650 molicel (number in series does not matter for sag)

for 60a load (it persists on source and is dividable on 3 and 2):
2p (30a on cell) 21700 will sag to:
starts at 3.57

and
3p (20a on cell) 18650 will sag to:
starts at 3.65

3p had more wh at 60a (186 vs 174 = 6.4%)

So 2p will hit 3.5v over 1ah of discharge (73% of battery) and 3p will hit 3.5v over 0.65ah of discharge (72.9% of battery).

same for both batteries at 50% charge - they will both hit 3.3v. 3p will do it later tho, due to greater capacity (6.4%)

So sag difference is for first 25% is more noticeable (3p will hold better almost by 0.1v and then difference will decay over first 25% but accounting to greater capacity of 3p of 6.4% 3p will be always ahead by small margin)

source:

2 Likes

Nicely done!
Love seeing the testing data being used.

4 Likes