New 14s battery pack, US18650VTC6 vs NCR18650GA

[Jan-Erik-86]

Thank you very much for the reply Pajda, that explains a lot!

I must admit the VTC6 performed much better then i had thought, and so i am actually leaning a little bit back to that now.
First of all, 70% capacity after 1000 cycles in itself would be enough (would take me 5 years, at least), but i also won't use even close to 100% DoD. How is your test on a lower DoD doing so far?

I know now that i get more "bang for the buck" with a PF/29E cell, as the VTC6 is ~30% more expensive and have a worse cycle life, but with the higher discharge voltage of the VTC6, even though it looses more % of Ah then for example a PF/29E in a cycle test, won't the actual Wh compensate a little for it?

While i want the battery to last some years, i also want to have fun, and it's not fun if i have 30% capacity left on a 3 year old battery, hit the throttle and reach LVC. For me that's the same as having 30% less capacity left, even if i could theoretically pedal home in lowest power setting. Do i risk this being the case with a a PF/29E?

 

[john61ct]

Do you have anything about Calendar life or Shelf life? Because to be honest it's very unlikely I'd do 1000 cycles. However, it's much much more likely I'd want to keep the same battery for 5 years or more.

long as you store at low SoC - but never allowing self-discharge even close to say 3.0V,

and in cool temps, 5 years is nothing, for all LI variants

 

[Jan-Erik-86]

I have measured how many cells i can fit inside the frame when using the Vruzend kit, and i ended up on exactly 112 cells, with only a few mm to spare. So 14s8p it is, and hopefully i can order it today.

So, only questions remaining, with an average load of 12-14A and short peek loads of ~50A:
1) Will the LG M29, Samsung 29E (E6 that would be, nkon don't have enough E7 in stock), or Panasonic PF perform more or less identical, or does one stand out as a better option (ignore the price)?
2) Will a M29/29E/PF cell have significantly increased IR after some years? Do i, in a few years, risk hitting LVC if i have ~30% left and hit the throttle?
3) What's the capacity loss of a M29/29E/PF cell after ~500 and ~1000 cycles?

 

[john61ct]

I recommend building three packs at the same time with different cells.

Run rigorous tests, keep good notes, and let us know in 3/5/7 years how it's going.

Then, by the time the first pack reaches EoL you will know exactly which cell out of those three, to use for its replacement.

And so will we

 

[Jan-Erik-86]

Hahaha, sounds like a great idea, I'll get right on it! :lol:

Nah, really, I should just say thank you all for all the great help with this, and I'm sure I'll be happy with my new PF pack. Just placed the order now.

 

[cwah]

Hahaha, sounds like a great idea, I'll get right on it! :lol:

Nah, really, I should just say thank you all for all the great help with this, and I'm sure I'll be happy with my new PF pack. Just placed the order now.

Why not the mh1? I'm thinking to order these. Either mh1 or m36

 

[Jan-Erik-86]

I did consider it, but from what i understood "anything higher capacity will die off quicker"...?

 

[docware]

"anything higher capacity will die off quicker" ..... that is only flippy ´s opinion, as he prefers 29E above all. Read again Pajda´s reports. He has biggest collection of really tested cells, bigger than anybody else here.

 

[Jan-Erik-86]

Ok, so you guys recommend i cancel the PF order and instead order an LG cell? So, the M29, MH1, MJ1 or M36?

 

[docware]

I would recommend you LG MJ1 or M36. Or Samsung 21700 50E. Definitely not PF.
Or maybe SONY VTC6 as in my opinion VTC6 could be more durable at lower temperatures.
By the way, you will never use the cells 0,5C - 1C 100 % DOD.

 

[Jan-Erik-86]

Ok, thank you. I've sent an email to nkon now asking if it's possible to cancel the order so i can place a new one on different cells.
I think i'll go with M36 then, and see if it's possible to have a more heated storage area for the bike in the winter time, and just be a bit gentle on the throttle should it be very cold.

Edit: Since i don't have a spot welder I'd like to use a kit like Vruzend, so i think going with 18650 instead of 21700 is the best option? Or are there similar kits for 21700 as well?

 

[flippy]

"anything higher capacity will die off quicker" ..... that is only flippy ´s opinion, as he prefers 29E above all. Read again Pajda´s reports. He has biggest collection of really tested cells, bigger than anybody else here.

sorry, but its not a opinion, its a measured fact. anyone that has done lifecycle testing or can read a datasheet can verify this.
and by comparison: a 2500mAh will generally outlive a 2900mAh. but the price per Wh is worse so 2900 is the sweet spot right now. when that changes i will change my recommendation.

Edit: Since i don't have a spot welder I'd like to use a kit like Vruzend, so i think going with 18650 instead of 21700 is the best option? Or are there similar kits for 21700 as well?

please dont, friction fit kits are a disaster in real life.

 

[john61ct]

Vruzend no longer relies on "friction fit", as of v2.1

https://youtu.be/rylbFnTgFI8

 

[cwah]

"anything higher capacity will die off quicker" ..... that is only flippy ´s opinion, as he prefers 29E above all. Read again Pajda´s reports. He has biggest collection of really tested cells, bigger than anybody else here.

sorry, but its not a opinion, its a measured fact. anyone that has done lifecycle testing or can read a datasheet can verify this.
and by comparison: a 2500mAh will generally outlive a 2900mAh. but the price per Wh is worse so 2900 is the sweet spot right now. when that changes i will change my recommendation.

Edit: Since i don't have a spot welder I'd like to use a kit like Vruzend, so i think going with 18650 instead of 21700 is the best option? Or are there similar kits for 21700 as well?

please dont, friction fit kits are a disaster in real life.

Did you compare on identical discharge rate? Obviously a 3600mah cell if you discharge test at 1C it wouldn't compete against a 2500mah at 1C.

But if you were to do 0.5C vs 1C, how would it compare?

 

[docware]

Since i don't have a spot welder I'd like to use a kit like Vruzend, so i think going with 18650 instead of 21700 is the best option? Or are there similar kits for 21700 as well?

Find somebody with spot welder, for 18650 also.

 

[Jan-Erik-86]

I'll be using the 20A rated Vruzend v2.1, yes, not the 3,5A rated v1.6.
While i know there will be some added resistance and increased size compared to spot welding, I'd like to at least give it a try due to it's simplicity. Worst case i can always take it apart and spot weld later...

 

[flippy]

Did you compare on identical discharge rate? Obviously a 3600mah cell if you discharge test at 1C it wouldn't compete against a 2500mah at 1C.
But if you were to do 0.5C vs 1C, how would it compare?

just checking spec sheets you get less cycles from higher capacity cells, its still physics and chemistry.

the lower cycles life does not compensate with the higher capacity. especially considering cost. its cheaper to build a bigger battery with lower capacity cells if you want better lifespan. this is partially also the reason why tesla still uses a 3350mAh cell in the model S instead of 3500+mAh.
it would give a tesla 10kWh extra storage but it would absolutely murder the lifespan of the battery.

 

[docware]

Does it mean flippy that you don´t trust results from Pajda ´s cycle testing ? Why ?

 

[flippy]

Does it mean flippy that you don´t trust results from Pajda ´s cycle testing ? Why ?

i dont say i dont trust them, he does great work. the GA lifespan conclusion just dont seem to line up with real world results.
once you have the GA or other high capacity cells in the outdoors instead of perfect 20c conditions their performance falls apart.
this is partially the reason why tesla cells seem to preform so well, they are used in a thermally controlled box with the most advanced (and most expensive) bms on the planet today. then you can get away with something like that.
other problem is that they are too expensive per Wh. the PF and 29E is simply too cheap for the GA to compete.

this is why my data differs from his, i do not put my lifecycle tester in a constantly temperature controlled room, the room i have them in swings from 10~30C depending on the outside and if i am present, much more realistic and its visible in the data.
i do not find "perfect" lab conditions perticulary useful in lifecycle testing as it gives a wrong impression and can result in making the wrong desisions. its also a hard conversation when a battery that costs over 10k dies 20% faster then calculated/expected and trying to build a business on repeat customers.

 

[docware]

I understand that you have lot of experience from real life of building and using of the battery packs, which I appreciate a lot.

From Pajda´s comments here I understand that he highlight mainly LG cells MJ1 and M36, not GA. From 21700 only Samsung 50E.

Yes laboratory and real life condition are different, that is good observation.

Frankly, I have a little bit problem to accept your conclusion as you have too much random temperature condition. I am not sure if you compare apples to apples.

Have you tested any other cells than 29E or PF ?

I know one case of the ebike rider who was able to kill battery with 29E cells in two years, because the battery capacity was too small to the motor.

 

[flippy]

so far i have detailed data about the 29E(6 and 7), PF, GA, BD, 35E, 25R, MH1 and probably a couple i forgot. and ofcouse a crapload of chinese junk.

and my tests are realistic or done on cells that have been in use for years of daily cycles so that is actual usable data and it does give a better idea on what to expect in terms of realistic lifespan. no pink glasses data.

next week i am planning on taking a couple new GA's, PF and 29E's and stuff them in a test mount and dunk them in a fridge set to slighty above freezing and run them hard and see how they cope.

 

[cwah]

I just ordered MH1 cells from Nkon. Was it good?

 

[Jan-Erik-86]

I was able to change my order to M36 cells instead. Ships tomorrow, and can't wait to test them!

 

[flippy]

I just ordered MH1 cells from Nkon. Was it good?

depends on your use case but they preform well. but again like most other cells: they cost too much.

 

[docware]

While i want the battery to last some years, i also want to have fun, and it's not fun if i have 30% capacity left on a 3 year old battery, hit the throttle and reach LVC. For me that's the same as having 30% less capacity left, even if i could theoretically pedal home in lowest power setting. Do i risk this being the case with a a PF/29E?

Your choice LG M36 seems to be for such case better option than Panasonic PF.