Highest capacity 18650 cells

ryan1685

100 W
Joined
Apr 3, 2015
Messages
126
Looking for some direction here seeing that there is a lot of fakes on Alibaba. Can someone help me to understand what the highest capacity is for 18650 cells. Looking to replace the Samsung 35E's out of my current battery to something that of higher quality and more energy dense. Trying to increase range in my 2018 radrover without carrying another battery
 
In terms of outright energy density the Panasonic 18650GA cells are up there, but only at very low C-rates. They claim 3400 mAh, but I've tested them at less than 1 C and never got more than 3100 mAh, and plenty of sag. The 30Q cells might be the best compromise.
 
This might not apply, but:

The problem is that for the stuff Ive seen so far, the greater the energy density, the lower the power density. Meaning, you can get more capacity, but you have to draw it out at a lower rate. (higher internal resistance, IIRC)

So to use cells with more capacity (if you find any), you have to have enough of htem in parallel to ensure they can supply the peak and sustained current demands in the same way the original cells could, with as little voltage sag as possible (the same or less than the original cells).

You may have to use more of the higher capacity cells than the higher power cells to support the same load.
 
I found the info here: http://lygte-info.dk/review/batteries2012/Common18650comparator.php to be very helpful. After you've selected the cells you'd like to compare, there is a link to read the full review of the cell. For example, here's a review of the LG MJ1 which I'm using in my packs.
 
ryan1685 said:
Looking for some direction here seeing that there is a lot of fakes on Alibaba. Can someone help me to understand what the highest capacity is for 18650 cells. Looking to replace the Samsung 35E's out of my current battery to something that of higher quality and more energy dense. Trying to increase range in my 2018 radrover without carrying another battery

Samsung 35E is on the top of currently available energy density (together with LG MJ1 and Sanyo GA), but it have relatively low power density which is typical behaviour for high energy cells. A simple rule applies, You can have a high energy density or power density, but not both at once.

So If you have "standard small ebike battery = you can not add more cells " you can paradoxically get a better mileage and lifetime when using more powerful, but less capacity cells like LG HG2(expensive but one of the best balance of energy nad power density) or Samsung 30Q(they have significantly worse cycle life than LG).

A simple example: If you have 250 W motor and 250 Wh battery you will be disapointed with Samsung 35E performance, but when you can have more than 500 Wh battery, Samsung 35E will start working fine for you.
 
If it's a 2018 bike it seems kind of a shame to take out perfectly good newish cells and replace them. An extra battery isn't all that heavy, maybe just buy another one and carry it when you think you're going to need it? I could carry one of my Hailong case batteries in a backpack no problem, I carry three times the weight in groceries in the same backpack while riding almost daily.
 
LG M36 is the highest capacity readily available 18650 on the market. 3600mAh, 10A rated.

Sanyo GA, Samsung 35W, LG MJ1 are all 3500mAh and are close.
 
They wont supply that capacity at 10A. Or live long at that load. 0.2C is fine if you got 40 cells in parralel.

This is why accurate testing needs to be done and see how bad ebike loads are for the lifespan.
 
flippy said:
They wont supply that capacity at 10A. Or live long at that load. 0.2C is fine if you got 40 cells in parralel.
I dont know of any cell that gives its full capacity at max discharge rating.
They are separate performance characteristics.
Capacity is universally tested at 0.2C discharge, and there are defined methods for testing max discharge, life cycle,etc.
 
Personally I think the 1 C discharge test is one of the most useful figures a manufacturer could quote. Testing at some rate which is so far below normal operation for a capacity test is a bit like shaving your pubes to make your Johnson look bigger. If you know you need 15 amps continuous to push you down the road, then the number of cells in parallel is easy to calculate at a notional 1 C.
 
From my point of view the 0.2C nominal capacity test is based on the assumption that with lower discharge rate you just cannot get significantly more energy from the cell. And it should be also noted that there are much more low power applications/electronics than high power ones.
 
Pajda said:
From my point of view the 0.2C nominal capacity test is based on the assumption that with lower discharge rate you just cannot get significantly more energy from the cell. And it should be also noted that there are much more low power applications/electronics than high power ones.

but those applications are not relevant to us. :wink:
 
Fortunately there are several independent groups online who test all (?) available batteries for key parameters at several different discharge levels 2A, 5 A, 8A, 10A, 15A, ......30A even,.. if the cell is capeable.
Each time reporting the capacity, temperature increase, etc.
Sites such as, lygte, dampfakkus.de, Ecig forum, etc
 
Pajda said:
So If you have "standard small ebike battery = you can not add more cells " you can paradoxically get a better mileage and lifetime when using more powerful, but less capacity cells like LG HG2(expensive but one of the best balance of energy nad power density) or Samsung 30Q(they have significantly worse cycle life than LG).

Does HG2 really has so much better cycle life than 30Q? Could you be so kind and share a link to that source?
 
Hillhater said:
Fortunately there are several independent groups online who test all (?) available batteries for key parameters at several different discharge levels 2A, 5 A, 8A, 10A, 15A, ......30A even,.. if the cell is capeable.
Each time reporting the capacity, temperature increase, etc.
Sites such as, lygte, dampfakkus.de, Ecig forum, etc
but not lifecycle testing, and that is for us the most important thing. how hong to they last under sustained 1~2C with 3~5C peaks for example.
 
flippy said:
but not lifecycle testing, and that is for us the most important thing. how hong to they last under sustained 1~2C with 3~5C peaks for example.
Problem is, everybodys useage is different, average load, peak load, % DOD, operating temperature, charge rates , etc etc...all factors that influence the life cycle result.
So in some respects, the manufacturers life cycle data is as good a comparason as any.
Has anyone ever run a quality branded 18650 pack beyond being useable , on cycle life alone. ?
 
i am planning to find that out exactly.

i want to run several cells into the ground at various loads. see my topic here:

https://endless-sphere.com/forums/viewtopic.php?f=14&t=93989
 
madin88 said:
Pajda said:
So If you have "standard small ebike battery = you can not add more cells " you can paradoxically get a better mileage and lifetime when using more powerful, but less capacity cells like LG HG2(expensive but one of the best balance of energy nad power density) or Samsung 30Q(they have significantly worse cycle life than LG).

Does HG2 really has so much better cycle life than 30Q? Could you be so kind and share a link to that source?

Sorry it is based on my personal tests which cannot be shared in public right now. LG HG2 is very good cell but it is very expensive and hard to buy (also there are many copies of this cell with much worse parameters), but problem is that 30Q has very bad cycle life at 0.5C-1C test. This behaviour is common for more Samsung cells and the major problem is by my opinion in the 0.5C charge rate. I suppose that at 0.3C charge rate test, the cycle life of 30Q will be much better.
 
Does HG2 really has so much better cycle life than 30Q?

Regardless as to whether you use HG2 or 30Q, the best thing you can do for long life is...charge to a max of 4.1V per cell, and...4.0V is better (no benefit to charging to 3.9V per cell). Storing long-term at 4.2V per cell is an early death for the pack...
 
The highest "rated" capacity which is already possible to buy is LG M36 or LG INR18650M36 with 3600mAh capacity as redilast posted above in this thread.
 
but that rating it useless if you cant even get 3Ah from it under load. then you can buy a 29E's of PF's for half the price and get the same usable capacity.
 
Sony VTC6 seems to have better cycle life than 30Q at 0.5C-1C cycle life test (Source: my personal tests) but I am now only at 250 cycles with VTC6 sample so I cannot tell you right now if it is significantly better. After comparing the prices the VTC6 is not interesting for me.
 
Back
Top