Samsung INR18650-25R cycle test.

[cwah]

Could you do test with:
20A discharge to 3v
1A charge to 4.05v

Should be truer to what happens in real life!

 

[electricbike]

Uppdated 500 cycles, will extend to 600 or when 80% is reached.

Could you do test with:
20A discharge to 3v
1A charge to 4.05v

Should be truer to what happens in real life!

If you mean a long term cycletest? don't think so.
The cell will be to hot for that, will not burn my apartment down.

DrkAngel: Yes i will do a singeltest of some new cells to get startvalue.

 

[DrkAngel]

500 cycles completed 1C C\D

Graphed vs 20A discharge rate

Capacity deterioration plotted ... estimated based on 95% >> 90% deterioration.
8C discharge plot seems to level noticeably after 85% point ...
If 1C plot reacts similarly, 1C discharge should reach 80% capacity mark near 1000 cycles.

 

[electricbike]

Differencetest uppdated.

 

[DrkAngel]

#1 cycle test would likely not suffer any variance in accuracy if expanded to 40cycles, or more.
All previous similar tests showed minimal deterioration variance after the 1st few hundred cycles.
Should speed up long term test nicely.

 

[zener]

How about a cycle test with constant capacity rather than voltage?
So instead of charging always from 4.15V to 3.1V.
Charge from near empty to 75% (1875mAh) of the initial capacity 2500mAh .
Than always cycle 1875mAh till 500cycles.

 

[liveforphysics]

As the 25r was designed around being a cordless power tool battery in which cycle life is of minimal importance in the cell design (as opposed to EV cells which sacrifice C-rate and some capacity in exchange for high cycle life), I'm actually amazed they cycle so well.

 

[DrkAngel]

... or 22.7% loss vs rated:

"11.2% loss over 500 cycles, or 22.7% loss vs rated:" "22.7% loss" is not accurate.
A full range cycle should demonstrate a true 11.2% actual "loss", vs rated ... with additional 11.5% "reserved" ... might be more accurate.

 

[Wheazel]

Surprised over the durability of thee cells. The number of cycles tested will probably never be reached in my applications.
Reassuring to know they hold up so well. I instantly wonder what other cells with more cycle life as a design criteria would show under similar "shallow cycle" test conditions.

 

[madin88]

Surprised over the durability of thee cells. The number of cycles tested will probably never be reached in my applications.
Reassuring to know they hold up so well. I instantly wonder what other cells with more cycle life as a design criteria would show under similar "shallow cycle" test conditions.

don't forget that this was only 1C charge and 1C discharge, so far byond the specs and normally not how the 25R will be used in real life applications..
i really would like to see same test with a samsung 35E or panasonic 3500mAh GA for comparison.

 

[DrkAngel]

1C discharge cycles will be supplemented with 2C then likely 4C.
8C is graphed to <70%.
All will be graphed together for direct comparison.

Alternate graph will also compare charge rate ... 2.50A 1C vs 4.00A 1.6C (maximum rated)

 

[cwah]

Doesn'the make any sense to test it at 1C discharge if no one use it at 1C.

 

[DrkAngel]

Interesting is that:
8C discharges deteriorates cell to 80% capacity at ~125 cycles.
1C discharges deteriorates cell to 80% capacity at ~1000 cycles.

So 8C damages cell at a 800% rate
1C = 1/1 100% cycle life
8C = 1/8 12.5% cycle life

Am curious to see 2C and 4C cycle tests:

1C = 1/1 100% cycle life 1000 cycles
2C = 1/2 50% cycle life 500 cycles ?
4C = 1/4 25% cycle life 250 cycles ?
8C = 1/8 12.5% cycle life 125 cycles

 

[zener]

Interesting conclusion. Didn't expected that the rate of discharge have so much impact on cycle life.

I am curious to see cycle test for a 80% capacity cycle from the beginning till 500 cycle.
So every cycle 2000mAh (@ 1C/1D ) charge and discharge from the beginning rather than fixed voltage charge discharge.

 

[madin88]

@zener Interesting thought. If charging to 4.2 instead of 4.15 cycle life would suffer a bit, but would be marginally the same loss as was graphed with the previous 1C test which yielded 2200mAh at the start.

It would be nice to have a chart with total used / drained Ah after given cylces.
Yes, 4,2V will be ahead the first 100 - 200 cycles or so, but after some point it starts to get worse compared to the cell that was charged to 4,15V. IMO this depends all on the cylces the user wants to have. For instance if someone only needs 200 cylces in 10 years than charging to 4,2V makes sense, but if you going to use 500 or 1000 cylces, it definitely will be better to charge to only 4,15V or even lower in view of total useable Ah.
Fact is if we charge to 4,2V the cell will wear more as if you charge to 4,15V, so even after lets say 50 cylces the cell charged to the lower voltage will be in better shape as the other.

 

[Teh Stork]

As the 25r was designed around being a cordless power tool battery in which cycle life is of minimal importance in the cell design (as opposed to EV cells which sacrifice C-rate and some capacity in exchange for high cycle life), I'm actually amazed they cycle so well.

What common cells would you rate as "EV cells"?

Common Samsungs: 22P, 22F, 26F, 29E, 30B, 32E
Common Panasonics: PF 2,9Ah, BF 3,35 Ah
Common LG: E1, HE2, HE4

or others off the top of your head?

Curious

 

[Offroader]

Surprised over the durability of thee cells. The number of cycles tested will probably never be reached in my applications.
Reassuring to know they hold up so well. I instantly wonder what other cells with more cycle life as a design criteria would show under similar "shallow cycle" test conditions.

don't forget that this was only 1C charge and 1C discharge, so far byond the specs and normally not how the 25R will be used in real life applications..
i really would like to see same test with a samsung 35E or panasonic 3500mAh GA for comparison.

I will have a bunch of extra SANYO 3500GA cells that I can run tests on soon.

I can run tests on these with my two hobby chargers that will allow me to run them continuously back to back tests when interfaced through the computer.

So I can set both of my chargers to run a specified cycle and leave it run continuously for weeks if needed. Discharge up to 8A 50W max on one of my chargers and 10A on the other. I can go higher 20A and 40A if I hook my charger to a battery to discharge to which is easy.

I just have to decide on the values to run them at.

 

[Wheazel]

Surprised over the durability of thee cells. The number of cycles tested will probably never be reached in my applications.
Reassuring to know they hold up so well. I instantly wonder what other cells with more cycle life as a design criteria would show under similar "shallow cycle" test conditions.

don't forget that this was only 1C charge and 1C discharge, so far byond the specs and normally not how the 25R will be used in real life applications..
i really would like to see same test with a samsung 35E or panasonic 3500mAh GA for comparison.

100% aware of this, and those are higher rates than where I use my current ebike batteries. (I am not using any 25r for the bigger builds tho)
My cargobike peaks at ~0,5C and sees 0,1C - 0,3C most of the time riding. I didnt have a clue of that the currents had this big of an impact on the cells before reading this thread.
I thought that as long as you were within the cells discharge specs, there would be minor differances in aging. I find it very positive that there is more to it.
Mapping out the "aging expectancy chart" for different conditions gives a good base for decisionmaking in regard to projects/builds.
Maybe we will see that rough conclusions can be drawn to other cells aswell based on the results to come.
I am also very curious how the graphs go for the cells you mentioned!

 

[Hillhater]

As the 25r was designed around being a cordless power tool battery in which cycle life is of minimal importance in the cell design (as opposed to EV cells which sacrifice C-rate and some capacity in exchange for high cycle life), I'm actually amazed they cycle so well.

What common cells would you rate as "EV cells"?

Common Samsungs: 22P, 22F, 26F, 29E, 30B, 32E
Common Panasonics: PF 2,9Ah, BF 3,35 Ah
Common LG: E1, HE2, HE4

or others off the top of your head?

Curious

I guess we are referring to the Tesla EV packs, which in majority are Panasonic, but the problem is they are unique spec for Tesla, so their actual details are not public.....other than those test results that have been reported in other threads.

 

[electricbike]

Uppdated difference test.

 

[DrkAngel]

Starting a chart including the new test cycles.

Oops! corrected % labels.

 

[madin88]

shouldn't the curve with 4A charge not start from the same point as 2,5A charge? or is the CV cut off current also higher?

 

[cwah]

Starting a chart including the new test cycles.

any way to do something like 5A discharge constant with few 20A peak of 10s every 2 minutes? it would be closest to real life. It would be closest to real life as nobody would discharge at peak power until empty either

 

[DrkAngel]

It would be closest to real life as nobody would discharge at peak power until empty either

I fear you underestimate several members ...

Take into account that an 8C discharge (60min / 8C ) is a maximum of only 7 minutes 30 seconds till dead empty.

 

[cwah]

yeah but I suppose most members wouldn't use their bike only for 10mn I suppose?