Li-ion cells cycle ageing

eMark said:

It seemed to take longer bulk charging my 10S5P 30Q battery (with 2.5A charger) from 4.00V to a 4.10V than say from 3.40V to 3.50V?

Click to expand...

That's because the charge current at 3.5V is maximum, e'g 2.5A in your case.
Above 4V the charge current is just a fraction of 2.5A as this is close to finished.
If one had forced 2.5A charge current into the battery during the whole charge cycle one would live in danger for fire, not to talk of short cycle life.

@ eMark
You need to compare either rest voltage with rest voltage, or load voltage with load voltage to get an apples to apples comparison.
Also during charge the battery will be under "load", therefore the voltage will be higher as it would be at rest.

When looking at red (1A) curve from following graph, you can see that between 4.00V and 4.04V there are 250mAh available, while between around 3.61 and 3.65V (same 0.04V) there are just 125mAh or half as much available. The flatter the curve, the more energy can be stored per given voltage difference.

john61ct said:

So perhaps not just better for longevity, but fewer balancing issues, to just set the 30Q stop-charge point lower for normal usage charging. . .Maybe even 4.00?

Click to expand...

For my eriding enjoyment no problem charging to 4.00V and discharging to 3.4V three times a week ... if it means increased cycle life longevity versus charging to 4.1V. Hopefully, a 4.0V charging goal would achieve 1000 useful cycles). So i could still get two recreational outings and running errands at 8-10mph with round trips averaging only 3-5 miles (occasional pedal assist for exercise) before reaching 3.4V bottom termination. Repeating this as often as three times a week.

docware said:

No, no, there is a LOT capacity between 4.0 and 4.1 V. I mean resting voltage. ... So optimal charging voltage to exploit capacity may be about 4.12 - 4.15 V (termination point).

Click to expand...

My eriding applicaion is NOT "raw performance" rather cycle life "longevity" so being the PRIMARY objective purchasing the 30Qs; however for my application the 30Q is probably overkill. That Said My Reasoning Was: (1) The price was right during IMR's Black Friday extended sale on 30Q cells this past November; (2) Suspected 30Q had a low DCIR; (3) My QC technical background interest in further testing the newer 141 cells to see if 30Q self-discharge tendency is still an inherent problem.

As you're undoubtedly aware the 30Q 15A rating is conservative. Perhaps, the reason Samsung didn't rate it closer to or at 20A is because they were aware of its self-discharge tendency, but because of it's low DCIR they decided the trade-off was worth it for those that use the 30Q year-round re-charging every day so any resting discharge isn't a big deal (e.g. daily charge/discharge cycling twice a day for ebiking commuting). So will be putting my Vruzend 10S5P 30Q battery with provision for splitting it into two 5S5P packs for balance charging and bench discharging over the next couple winter months. Currently both packs are at 50% 3.75+ resting volts (min-max 3mv) as of last night. Rested them for 3 hours after discharging from 3.85V before recording their resting voltages with my BG-8S. Comparing it to my multi-function Capacity Controller is like comparing a Swiss timepiece to a Timex watch. Well, i might be exaggerating a tad, but that ISDT BattGo BG-8S is sweet (as in beloved) and more accuarte then that Capacity Controller.

madin88 said:

@ eMark
You need to compare either rest voltage with rest voltage, or load voltage with load voltage to get an apples to apples comparison.
Also during charge the battery will be under "load", therefore the voltage will be higher as it would be at rest.

Click to expand...

The two 5S5P split packs are currently in an established resting mode (having rested after discharge from 3.85V to 50% capacity for three hours before beginning what i'll call the official resting/testing period. Allowed three hours to make sure all bounce back voltage had ceased. Will track any self-discharge with each 5S5P pack and record any apparent self-discharge every 24 hrs. and report back Monday.

I think 1000 cycles is not realistic?

Unless your high Ah capacity is enough to bring the average discharge C-rate down much lower than with most ebikes.

Another thread, „Samsung 40T and other cells like it?“ https://endless-sphere.com/forums/viewtopic.php?f=14&t=102682, reminded me that I would like to have here information relevant to cycle aging of the Li-ion cells. Therefore I copy here my post from above mentioned thread to have all relevant info in one thread :

The rate at which a battery or cell can be charged or discharged is limited by the rate at which the active chemicals in the cells can be transformed.
Forcing high currents through the battery results in incomplete transformation of the active chemicals reducing the battery’s effective charge capacity and it also causes unwanted, irreversible chemical reactions to occur because the chemical transformations cannot keep up with the current demands.
The unwanted chemical transformations consume some of the active chemicals causing the battery to lose capacity and thus age prematurely.

From the above we can expect that with each charge/discharge cycle the accumulated irreversible capacity loss will increase. Although this may be imperceptible, ultimately the capacity reduction will result in the cell being unable to store the energy required by the specification. In other words it reaches the end of its useful life and since the capacity loss is brought on by high current operation, we can expect that he cycle life of the cell will be shorter, the higher the current it carries.



The capacity reduction at high discharge rates occurs because the transformation of the active chemicals cannot keep pace with the current drawn. The result is incomplete chemical reactions and an associated reduction in capacity. This may be accompanied by changes in the morphology of the electrode crystals such as cracking or crystal growth which adversely affect the internal impedance of the cell.

Ageing Accelerators
The previous paragraph indicated some of the basic ageing factors inherent in the battery chemistry. From this we can see that certain external environmental and usage factors, such as those following, can be considered as ageing accelerators :

High and very low temperatures
High energy throughput rate (Charge and discharge rates)
Mechanical stress or vibration which can give rise to open or short circuits or seal failures.

Operating Environment
In addition to the ageing accelerators noted above there are some less obvious environmental factors which can influence battery life.
A temperature gradient across the battery can increase the rate of battery ageing. From Arrhenius we know that, with a 10°C difference in temperature across the battery, some cells will age at twice the rate of others giving rise to unbalanced stresses on the cells resulting in premature failure. High pressure or cyclic pressure changes can cause mechanical failures of the cells.
High humidity can give rise to corrosion causing increased contact resistance at the battery terminals.


Source : THE ELECTROPAEDIA https://www.mpoweruk.com/

https://www.mpoweruk.com/Software_Configurable_Battery.htm
https://www.mpoweruk.com/failure_modes.htm
https://www.mpoweruk.com/life.htm
https://www.mpoweruk.com/lithium_failures.htm
https://www.mpoweruk.com/reliability.htm
https://www.mpoweruk.com/chemistries.htm

Actual predictions for a given use case are indeed going to be inaccurate.

But a set of standardized tests revealing their **relative** performance

is very valuable.

Certainly better than relying only on vague anecdotal reports over the years.

Uf.. too much informations. So only one question where to get samples of "141" line of Samsung 30Q in EU for testing?

I made no personal attack, which I agree would indeed be very inappropriate.

This is what I was responding to.

eMark said:

Similarly, extrapolating the performance of the whole population of batteries from a sample of one [cell] can also give unreliable results and erroneous conclusions . For reliable predictions, testing for each stress factor should be performed on several samples
…
docware's bench tests only represent (IMO) at most 50% of the total story of a DIY battery's cycle life longevity before EOL. There are innumerable real-life variables (read entire article) that determine/decides the EOL of an 18650 DIY battery.
…
Whether, docware's or flippy's or Pajda's bench tests how reliable predictor are they of a DIY battery's cycle life longevity before EOL???

Click to expand...

I'm usually not even aware of who the author is of the post I am responding to. Don't take everything so personally, this is a public forum. . .

Absolutely.
And I would add that really proper testing should be like automotive is doing. For example complete battery has to be mounted on the bike frame. This assembly is then mounted on the vibration testing table, which is in the enviromental test chamber. Actual testing then covers electric testing as well as mechanical vibration and temperature cycles, all together at once. Electric parameters testing itself shell simulate real conditions, not only continuous discharging at constant current.

Congratulation on beginning of realizing of the Li-ion cells aging complexity.

By the way, for example DCIR of Samsung 29E7 at 25 °C is about 30 – 31 miliohm, at 2 °C is about 65 miliohm.

eMark said:

To get a true representation of battery ageing or wearout, the tests must be run with complete batteries ... just as they would be in the planned application.

Similarly, extrapolating the performance of the whole population of batteries from a sample of one [cell] can also give unreliable results and erroneous conclusions . For reliable predictions, testing for each stress factor should be performed on several samples. ... https://www.mpoweruk.com/reliability.htm

Click to expand...

Some may assume that the individual cells being tested are giving us a real-life DCIR application that's representative of a DIY battery of new brand name cells (e.g. 10S5P, 14S8P, etc.) after say 150 cycles, 250 cycles, etc. Capacity inpedance is crucial (DCIR) to cycle life longevity and yet docware's bench tests only represent (IMO) at most 50% of the total story of a DIY battery's cycle life longevity before EOL. There are innumerable real-life variables (read entire article) that determine/decides the EOL of an 18650 DIY battery. Possibly the most important real-life variable (IMO) that is near impossible to track is abuse and then defining abuse.

Click to expand...

Thats right, but imagine how many different tests you would have to perform already with one given cell in order to find out "the other 50% of the story".
You would have to do individial cylce tests at different ambient temperatures, different charge and discharge current, different cut off voltages during charge and discharge etc etc..
Then if you go with the type of cell which suit the most for your purpose and building a DIY pack with dozends or hundred cells, other things will come into play which will effect lifetime of this pack as a whole, like equal curren flow through all cells (symmetric load), equal temperatures between the first to the last cell etc etc.

I totally agree that one cannot conclude lifetime or cycle life from one single cell to a (DIY)pack, and most annoying thing i find is that when you really did everything properly when assembling your own pack, there still can be those unknown things like the issue with self discharge because of production problems or whatever it has been why 30Q sucks so much.
The best cycle life on paper does help you quasi nothing if your 1000$ pack starts to drift apart after just 50 cycles, whether it is because of issues ex factory, or those "self made problems" like uneven current flow, soldering for too long time on some cells, leaky cans from spot welding which get overlooked etc

I am really thankful that docware is doing those tests and share his findings here, so big thanks to you :thumb:

eMark,
would you be so kind and restrict size of your post and your agressivity toward other members ?
Take it easy, forget please words like battery cycle life, just sit on your ebike and enjoy the ride.
Thanks for not stirring up hate and bitternes here any more.

madin88 said:

I am really thankful that docware is doing those tests and share his findings here, so big thanks to you :thumb:

Click to expand...

Amen *100 to both those, attaboy and thanks!

Not much to add, but interesting to read :wink:
So thanks from me too.

New batch of the pictures : M36 700 cycles, 29E7 633 cycles, 30Q 336 cycles, LG MJ1 300 cycles, SONY VTC6 64 cycles, Samsung 50E 56 cycles. As there is a lot of the curves in the capacity decay chart, I added new one, zoomed.










LG M36 achieved the goal 700 cycles showing excellent results both in capacity and DCIR having 93,9 % of initial capacity in the cycling interval 4,1 – 3,4 V /2,5 A. Measuring capacity in standard 4,2 – 2,5 V range at 1 A shows 96 % of initial capacity and 94,8 % of initial energy. We can hardly speak about any internal resistance rise of this cell after 700 cycles. Absolutely excellent !!

This is awesome work. Thank you Docware, Flippy , and Pajda for your reliable and valuable contributions . You've consistently done the electric powered community a huge service. Less educated newbies ( such as myself) - to this complicated and sometimes overwhelming niche of present and future tech- are extremely fortunate to have people willing to give their time and money to provide this level of support. Can't say thank you enough. There are also several others that provide tons of technical advice every day that keep me from wasting lots of money and burning my house down. In particular , Amberwolf and John61ct , have been extremely helpful over a long period of time. Thank you.

One of those is not like the rest 8-D

Seriously, I am not worthy being mentioned on that list

I mean I appreciate the thought, but let's just call it an "A for effort"

john61ct said:

One of those is not like the rest 8-D

Seriously, I am not worthy being mentioned on that list

I mean I appreciate the thought, but let's just call it an "A for effort"

Click to expand...

No I agree. Just because you aren't testing doesn't mean you aren't bringing a lot to the table. The information the testers are providing is critical and worth it's weight in gold , but the one's that answer tons of questions on a daily basis are just as responsible for the overall benefit this forum brings. Nobody has to spend their time explaining this stuff , but you guys do it and it's all equally helpful . It's difficult to remember every poster that has been instrumental in helping , there's a bunch , but I know I also appreciate the hell out of it.

Well, here we have New Kid in Town on the testing, or maybe rather the old acquaintance : Samsung 35E. Parameters are standard, 4,1 – 3,4 V 2,5 / 1 A 5 minutes rest.




https://www.youtube.com/watch?v=1mJbk2CuQ0I

Always can´t believe to my ears how well they sound live.
I just sitt down, listen and relax …..

docware said:

Frustrated and tired by never ending unfruitful debate about cycle life of various Li-ion cells, I finally decided to start some long term cycling to see the real results. Parameters were chosen in effort to imitate real utilization, but on the other hand to get some results in acceptable time.

Click to expand...

Really do appreciate your willingness to provide comparative 18650 results using brand name Grade A cells. They're performing better than the manufacturer's datasheet when it comes to the number of charge/discharge cycles even at 80% capacity with EOL nowhere in site (except PF & GA).

The Samsung 30Q datasheet rates it at 60% capacity (1800mAh) over 250 cycles. The Samsung 30Q being tested has over 300 cycles and is nowhere near 80% capacity, but is dropping with the possibility that it ends up as the worst cycle life performer among the cells being tested. On the other hand at the rate M36 is going we could see over 1500 cycles before it even reaches 80% of rated capacity. Sooo, is your and Pajda's plan to cycle LG M36, no matter how long it takes, before it retains 80% of rated capacity? M36 may outlast you or your equipment at 60% capacity :wink: ...

docware said:

... but on the other hand to get some results in acceptable time.

Click to expand...

One thing is for certain -- your test parameters point out the distinct advantage of charging to only 4.1 volts and discharging to only 3.4 volts. "Parameters were chosen in effort to imitate real utilization" ... WELL DONE! :thumb:

YOUR very WORTHWHILE thread is one of the most valuable, if not the most valuable, ES thread in that it makes very apparent that cycle life longevity is not pie-in-the-sky, but real utilization. Docware's revealing graphs (visual worth 1000 words) once and for all points out the benefit of charging to only 4.1 volts and discharging to only 3.4 volts.

Am deeply appreciative of docware's expert contribution. Docware's common sense practically of "real utilization" for extending a battery's cycle life beyond that of a manufacturer's datasheet is not only possible, but preferable :thumb:

Worth noting, long term beyond the current graphs , M36 may end up being the undisputed king , but up to 650 cycles , if availability is there and price is less with 29E , it looks like a better option to me. In the 4.1>3.4v range , it has more usable capacity and reportedly can handle a bit of abuse. At least in the USofdumbsuehappytards , LG cells are becoming harder to come by. This test actually has me excited to go with 29E for my next build.

So are LiPo completely off topic here?

If not, would like to include that option in this here REQ

What low cost 5-8000mAh / 6S choice would y'all recommend, for a **low power** use case?

Maybe 2C peak, usually under 0.2C sustained

QA / consistency of production and longevity being the top two priorities

weight / density being pretty unimportant, so I think in fact heavier is actually better for longevity, cost per year / cycle lifetime

So for example CNHL G+PLUS series is inferior to Black? even with both at the same $50 price?

https://bit.ly/313tZfE

Yes, Lipo are off topic in this thread.

@docware

You have started testing 21700 Samsung 50E
Is it Samsung 50E or Samsung 50E2

docware said:

Frustrated and tired by never ending unfruitful debate about cycle life of various Li-ion cells, I finaly decided to start some long term cycling to see the real results. Parameters was chosen in effort to imitate real utilization

Click to expand...

Thank you docware for your excellent graphs and your parameter choices for testing these lithium-ion 18650 cells. :thumb:

From the test by Mooch :

The two 50E2’s I tested delivered 5051mAh and 5062mAh at 1A (0.2C) down to 2.5V. The DC internal resistance measured 20.7mOhms and 20.8mOhms. This is astounding consistency and fantastic to see.

Click to expand...