Li-ion cells cycle ageing

[Pajda]

I think I remember reading a while back that 35E is very sensitive to charge current , so they may be very decent when charged at low rates around 500-700ma . Just a hunch.

That is true, they does not like 0.5C charging rate, the same as GA which is slightly better. On the other side M36 and MJ1 are completely fine at 0.5C. But still 0.3C for 35E (aka 1A per cell) is ok. This value of charging current is recomended in datasheet and I got good results at 0.3C as well.

 

[Pajda]

However, if you want to be able to see subtle DCIR changes, temperature and soaking time during measurement should be always the same as much as possible.

You are absolutelly right, my excuse was that I am aiming on the announced 20% capacity wear, which is usually tied with DCIR rise of about 100% or more. And so ca 5% error in DCIR measurement caused by the RT fluctuation is not significant.

By the way what chamber temperature would you suggest? Please round the suggested value to two decimal places.

 

[john61ct]

25/77 is very common

 

[docware]

New data : SONY VTC6 600 cycles, Samsung 35E 500 cycles, Samsung 50E 500 cycles, LG HG2 474 cycles.

 

[Dak77]

HG2 though. Wow. I'm super stunned. That thing is a beast.

 

[docware]

New data : SONY VTC6 664 cycles, Samsung 35E 557 cycles, Samsung 50E 550 cycles, LG HG2 532 cycles.

 

[Centurio]

Thank you for performing this tests!! There are interesting statistics about these cells. The Sony VTC6 will have reached the 700 cycles soon. The other cells soon after. What cells are you going to test after?

 

[Pajda]

Thank you for performing this tests!! There are interesting statistics about these cells. The Sony VTC6 will have reached the 700 cycles soon. The other cells soon after. What cells are you going to test after?

That is actually a very good question

 

[Upperfoot]

Thank you for performing this tests!! There are interesting statistics about these cells. The Sony VTC6 will have reached the 700 cycles soon. The other cells soon after. What cells are you going to test after?

That is actually a very good question

I've always been intrigued about RC LiPO cycle life and decay, and haven't found much testing done that's publically available, would be an interesting topic of testing and would help validate/invalidate common assumptions.

 

[docware]

Thank you for performing this tests!! There are interesting statistics about these cells. The Sony VTC6 will have reached the 700 cycles soon. The other cells soon after. What cells are you going to test after?

I am about to quit the testing as indicated few weeks ago.

 

[Pajda]

That is actually a very good answer

 

[docware]

Maybe it´s time to ask Pajda for his comment on the results.
„I will add general/well known comments to his results. „
https://endless-sphere.com/forums/viewtopic.php?f=14&t=102387&p=1507776#p1507776

I believe that his remarks are highly wanted not only for me. 8)

 

[Centurio]

Haha, nice hint doc.

To attach your linked comment of Pajda.

It would be kind of interesting how the same cell behaves under different conditions. Let's say one will be charged always to 4.2 Vol while an other has a higher discharge rate. The third is discharged with your standard discharge current but to 2.5V. Another could maybe have all the standard contiditons but "suffers" fast charging. So each of the of the cell has one not standard stress.

Hope you can understand me...

 

[Tommm]

Can you test 40t and 30t samsungs? Would be nice to know if they made some improvements.

 

[docware]

I´ll finish the testing within next 2 – 3 weeks. If I start again in September, 40T can be one in the test.

 

[Stealth_Chopper]

Thank You for documenting your work in these informative charts and graphs here for us to review.
After 32 years analyzing graphical turbofan engine test data, I find myself happily reviewing each of your updates because perhaps I'm not seeing a 'downside'. No failures, and seemingly superior performance by all the batteries compared to common expectations. ( I think?

 

[docware]

Uf, fortunately I din´t know that my graphs are regularly checked by professional.

You are right, cells are performing better than in the real life. We should realize that this is only comparative test done at specific conditions, in the real life are conditions different. For example mildly increased temperature 25 – 33 °C of the cells during whole test is very friendly both for charging and discharging.




Further, the cells are loaded mainly by cycle aging, calendar aging is not significant here.

Battery lifetime may affect lot of other factors like vibration, thermal gradients, inhomogeneous current distribution, DOD, loads, temperatures, BMS failure, …..
Also from statistical point of view is 1 cell under test absolutely not sufficient, battery contains much more cells.

 

[Dak77]

If there's a website or forum with this type of in- depth cycle testing by equally competent observers, I haven't seen it yet. There's Mooch and Thunderheart with their load/capacity/temperature/resistance measurements, but it seems you are of the very few (possibly the only ones in the world) doing this level of testing for public viewing. I'm sure lots of professionals quietly watch the progress of these tests. Vendors in particular.

 

[Stealth_Chopper]

Thanks Docware,
Temperature effects, one of my favorites, especially at 30K feet , Mn 0.8, -20F
Another new graph with fresh data and yummy.
Just great!

Doc,and all:
We know that Energy is conserved:
however on jet engines we use some fan compressed air to cool the cases,
( @ cruise / altitude ) for increase turbine performance.

Is your data telling us;
A warm battery performs better during discharge: and a vehicle system should dedicate some energy toward warming?
Mike

 

[john61ct]

way OT start a new thread

 

[docware]

New data : SONY VTC6 700 cycles, Samsung 35E 620 cycles, Samsung 50E 600 cycles, LG HG2 600 cycles.








SONY VTC6 finished testing. At 4,2 – 2,5 V / 1 A discharge has VTC6 after 700 cycles 93,1 % of initial capacity and 93,4 % of initial energy, in the cycling interval 4,1 – 3,3 V / 2,5 A then 86,6 % of initial capacity.



SONY VTC6 No1 DCIR after 700 cycles :



Updated SOC comparative chart after 700 cycles, SONY VTC6 added.

 

[john61ct]

Incredible stuff!

SONY VTC6 finished testing. At 4,2 – 2,5 V / 1 A discharge has VTC6 after 700 cycles 93,1 % of initial capacity and 93,4 % of initial energy, in the cycling interval 4,1 – 3,3 V / 2,5 A then 86,6 % of initial capacity.

But I'm very puzzled, how **greater** deterioration over lifespan is coming from lower capacity utilization, avoiding the voltage shoulders at both ends does not pay off in longer lifespan?

Just speculating, if the tests were to be continued until an EoL spec of say 75% SoH, do you think that relationship would continue?

Or am I misunderstanding your summary?

 

[Honk]

These tests are so vital and important to sort out the lesser good models in the djungle of batteries.
And one important thing can be determined by the results, that is 29E keeping its golden crown among contenders in form of the best bang for the buck.
I'm glad I bought 29E a while ago by chance and got lucky in the process.

 

[john61ct]

29E7 specifically, apparently very different from previous versions?

 

[Centurio]

@john61ct

If you dont mind loosing litte capacity, by not using those shoulders, you still have over 85% initial capacity afer 700 cycles. If in seldom cases you use this shoulders, it might not hurt the battery that much.

Just do some math and assume a battery pack according your needs. Calculate 20% extra capacity. Then calculate how many km you drive with one cycle and finally multiply this with this 700 cycles. I am sure the pack is gonne take you very very far, and you might want to change it before end of life anyway.

Look at the chart. The first 4 cells (5 with HG2 arriving) have still over 90% of their capacity. This is really great!

But we have to consider, accoring the first post, the discharge current is 2.5A only. Yeah, its good over the datasheets discharge rate and closer to reality. But I think many e bike riders here load their batteries with higher constant current per cell. Or do you want to commute 30mph (45kmh) with a BBS02 750W assuming 25A consumption and carry a 14s10p (10p!!!) pack with you?

But still I am very happy about this testing. It gives us all a good impression about cycle life. I definitly will consider these results when building my next pack.