If you are building packs today, which cells to choose?

Batteries, Chargers, and Battery Management Systems.
ichiban   10 W

10 W
Posts: 87
Joined: Aug 19 2018 10:41am
Location: Bangkok, Thailand

Re: If you are building packs today, which cells to choose?

Post by ichiban » Nov 07 2018 11:10am

I just got my LG18650 MJ1 delivered today evening. 180pcs to make 2 packs. Randomly measured voltage of 20 cells, they ranged between 3.601-3.605V which is not bad at all. Plan to make the small one under downtube of 14s5p first since this will be my very 1st DIY spot-weld pack. Prior pack is non-soldered / non-weld type. Bluetooth BMS is still on the way from AliExp. I don't see much talks about this MJ1 cells. I think it should be a pretty good balance between high capacity, high rate and lifecycle. Any advice ?

nbdriver   10 mW

10 mW
Posts: 33
Joined: Jul 31 2015 4:38am
Location: Belgium

Re: If you are building packs today, which cells to choose?

Post by nbdriver » Nov 08 2018 3:50am

Nobody as an opinion on GA vs 30Q today for best longevity at 16A load on 10S5P ?

Pajda   100 W

100 W
Posts: 216
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: If you are building packs today, which cells to choose?

Post by Pajda » Nov 09 2018 5:12am

Honk wrote:
Feb 16 2018 11:38am
Is there any news on your 29E cycle test?
Yes, I already finished my standard set of cycle life tests (0.5C-1C and 1C-1C with different DoD) for Samsung 29E. The direct competitors are Panasonic PF and LG MG1 (soon will be replaced by the new LG M29). I do not have yet all tests for MG1 so I can now compare only 29E with PF. PF has slightly better parameters but it is also more expensive (LG is also more expensive). PF has slightly lower IR and significantly better cycle life than 29E when fast charged >0.5C. At standard 0.5C charge rate and 1C 100%DoD discharge the cycle life is almost the same for both cells. The charging current is a huge problem for allmost all high energy Samsung 18650 and 21700 cells.

My conclusion is that Samsung 29E cell has the best ballance of price/power/cycle life ratio in this class, but you must operate it with lower charging current (0.3C is ideal, 0.5C is maximum for good cycle life).

User avatar
spinningmagnets   100 GW

100 GW
Posts: 10726
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: If you are building packs today, which cells to choose?

Post by spinningmagnets » Nov 09 2018 11:41am

Nbdriver, if your max loads are 16A using 5P, then you are pulling 3.2A per cell, and that is well below the max amps limit on the GA (known to work well at 7A) and MUCH lower than the 30Q.

Either one should be fine. I prefer the 30Q even when it is way underspec'd for its capabilities. This is because I may want to use that pack for something else later, and I like a fat safety margin...

ichiban   10 W

10 W
Posts: 87
Joined: Aug 19 2018 10:41am
Location: Bangkok, Thailand

Re: If you are building packs today, which cells to choose?

Post by ichiban » Nov 09 2018 12:08pm

Pajda wrote:
Nov 09 2018 5:12am
Honk wrote:
Feb 16 2018 11:38am
Is there any news on your 29E cycle test?
Yes, I already finished my standard set of cycle life tests (0.5C-1C and 1C-1C with different DoD) for Samsung 29E. The direct competitors are Panasonic PF and LG MG1 (soon will be replaced by the new LG M29). I do not have yet all tests for MG1 so I can now compare only 29E with PF. PF has slightly better parameters but it is also more expensive (LG is also more expensive). PF has slightly lower IR and significantly better cycle life than 29E when fast charged >0.5C. At standard 0.5C charge rate and 1C 100%DoD discharge the cycle life is almost the same for both cells. The charging current is a huge problem for allmost all high energy Samsung 18650 and 21700 cells.

My conclusion is that Samsung 29E cell has the best ballance of price/power/cycle life ratio in this class, but you must operate it with lower charging current (0.3C is ideal, 0.5C is maximum for good cycle life).
How about LG MJ1 ? I don't see much detail on it.

Pajda   100 W

100 W
Posts: 216
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: If you are building packs today, which cells to choose?

Post by Pajda » Nov 09 2018 3:02pm

Unfortunately, DCIR and common discharge tests cannot be used to assessment about cycle life. So you can easily find cell with excellent DCIR, but very poor cycle life. This is the case of Samsung 30Q.

LG MJ1 is using Ni-rich NMC technology, I think it is first gen of NMC 811 cathode from LG. The second generation is LG M36 cell and I think it will very soon replace the MJ1. More precisely, LG will replace soon the whole 18650 and 21700 product line with the new Mxx cells like M26, M29, M36 and M50(21700). First, I was thinking that Mxx product line will be cheaper but with slightly worse cycle life than the old generation. But at the particular example of MJ1 vs M36 from my first results (500 cycles) I can say that there is no significant difference under all of my tests.

Now back to your question :D LG MJ1 is definitely the king of the cycle life from the holy trio GA ,35E and MJ1. In fact, MJ1 has one of the best cycle life from almost all 18650 cells I have already tested. M36 seems to be at least as good and at the same time cheaper. Currently LG totally dominates in 18650, because Panasonic did not improve its 18650 product line for about last 5 years.

But different situation is in 21700 line, where LG M50 is also very good, (Samsung have the same problems with cycle life as in the 18650 line) but the new Tesla-Panasonic cells NCR21700A from the first tests seems to have excellent performance in DCIR. For example LG M50 and Samsung 50E are 2C rated cell, but NCR21700A is 3C rated (all are 5000mAh / 18.2Wh cells). I am now awaiting samples of this new Tesla cells and will start the cycle life tests soon. If Panasonic managed to improve the cycle life of NCR21700A as well, it will be nice competition here. The second important question is if Panasonic will use this new chemistry in its 18650 product line or it will from now focuses on the 21700 product line only.

ichiban   10 W

10 W
Posts: 87
Joined: Aug 19 2018 10:41am
Location: Bangkok, Thailand

Re: If you are building packs today, which cells to choose?

Post by ichiban » Nov 09 2018 9:04pm

Pajda wrote:
Nov 09 2018 3:02pm
Unfortunately, DCIR and common discharge tests cannot be used to assessment about cycle life. So you can easily find cell with excellent DCIR, but very poor cycle life. This is the case of Samsung 30Q.

LG MJ1 is using Ni-rich NMC technology, I think it is first gen of NMC 811 cathode from LG. The second generation is LG M36 cell and I think it will very soon replace the MJ1. More precisely, LG will replace soon the whole 18650 and 21700 product line with the new Mxx cells like M26, M29, M36 and M50(21700). First, I was thinking that Mxx product line will be cheaper but with slightly worse cycle life than the old generation. But at the particular example of MJ1 vs M36 from my first results (500 cycles) I can say that there is no significant difference under all of my tests.

Now back to your question :D LG MJ1 is definitely the king of the cycle life from the holy trio GA ,35E and MJ1. In fact, MJ1 has one of the best cycle life from almost all 18650 cells I have already tested. M36 seems to be at least as good and at the same time cheaper. Currently LG totally dominates in 18650, because Panasonic did not improve its 18650 product line for about last 5 years.

But different situation is in 21700 line, where LG M50 is also very good, (Samsung have the same problems with cycle life as in the 18650 line) but the new Tesla-Panasonic cells NCR21700A from the first tests seems to have excellent performance in DCIR. For example LG M50 and Samsung 50E are 2C rated cell, but NCR21700A is 3C rated (all are 5000mAh / 18.2Wh cells). I am now awaiting samples of this new Tesla cells and will start the cycle life tests soon. If Panasonic managed to improve the cycle life of NCR21700A as well, it will be nice competition here. The second important question is if Panasonic will use this new chemistry in its 18650 product line or it will from now focuses on the 21700 product line only.
Thank you Pajda.

You definitely know batteries and what you are talking about, precisely.

We need more of people like you who did not simply gather info from reading, but do the testing on their own for real. Greater than that is that you are willing to share what you found out. Please keep on your works and share to the community for the greater good of all.

My salute to you.

Pajda   100 W

100 W
Posts: 216
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: If you are building packs today, which cells to choose?

Post by Pajda » Nov 10 2018 7:31am

Thanks ichiban :bigthumb:

I do not consider myself an all-knowing battery expert. My specialization is battery application "expertise". I have read many books about battery chemistries and have heard to many lectures of battery experts but I never got the answer on a simple question. "Maestro, can you tell me, what is currently the best battery cell on the market for my e-bike/e-car/e-toy..?" Or the promised answer was disgustingly expensive.

Now back to the useful info :D I am awaiting next week the delivery of the Samsung INR21700-50E cell samples from Nkon. I will imediatelly start the cycle life test. Tesla-Panasonic (produced by Sanyo :wink: ) cells are on the way too, but their delivery will took few weeks.

From 21700 cells I have aleady running LG M50, Samsung 48G and for about two week Samsung 40T. LG M50 cycle life should be considered as very good but M36/MJ1 seems to be even a bit better. But if you look at the prices, it is already possible to order M50 cells from 100pcs quantity for the same USD/kWh price as M36 cells. So that you can benefit from the other advantages of the new 21700 format. I must say, that I did not expect that the retail price and overall technical parameters of 21700 will equalize with 18650 so soon.

Post Reply