Samsung INR18650-25R cycle test.

electricbike

100 W
Joined
Oct 31, 2015
Messages
100
Location
Lund/Sweden
Hello. I have started a cycle test of Samsung INR18650-25R green cell.
Discharge 2,5A cutoff 3,1V, wait 5 min charge 2,5A to 4,15V, wait 5 min.
Maybe some of you find it interesting.

1250.jpg


New cycletest of different charge current.
Cell 1 @ 2,5A
Cell 2 @ 4,0A

Difference-test-960-1320pg.jpg
 
Nice...keep us posted as you go.

:D
 
This is my test equipment. Nothing special at all but it working.
Turnigy Reaktor is nice to have when you need to cycle cells.

reaktor.jpg
 
A 1C (2.5A) discharge rate isn't much of a load for real life testing. Crank it up to 4C (10A, still just half of rating) and it will be more of a real life test. For grins I'd like to see what it would do at max rating of 8C (20A).
 
1C 2,5A is what my cell will se in my batterypack for the moment 8P14S 20A controller.
I will keep the temperature down so i get long life of my pack.
 
Thanks for taking time doing that, i know what job it is! :lol:

Guys Please pay attention when using RC device for test.. these ar enot very precise and often will read lower value than the actual Ah value of the cells! the reason? The voltage loss in teh test lead.. these often have high resistance and contact also have high resistance making often 0.2 or 0.3V lower reading on the actual voltage of the cell, and stoping the discharge sooner because LVC kick in too soon... ( if charger read 3.0V, the cell might still be 3.3 or higher volt...

That's also why testing at high rate ( 2, 3 or more C ), make it worst.. The best testing device to avoid that is a device that use not 2 but 4 wires for measuring the cell state: two that carry the discahrge current and two that measur ethe cell voltage ... independently.. .. that is called Kelvin connections. I posted that already on E-S.

Some high reputation Battery testing lab use 4 wire testing too.. as well as me in my lab. :wink: result are more constant and acucrate this way.. but i agree that that's not everybody that want to buy such equipement...

The best in that case is to use the balance lead of the charger to monitor voltage and not the discharge lead.. so this way the charger knwo exactly the true voltage of the cell.

Doc
 
Overclocker said:
1c charging isn't what people usually do. So your results won't be representative of real-world
I will charge at 0,5C in my real pack, only 1C just now to speed up my test. What C rate do people usualy charge at?
 
Hillhater said:
Doctorbass said:
........The best in that case is to use the balance lead of the charger to monitor voltage and not the discharge lead.. so this way the charger knwo exactly the true voltage of the cell.

Doc
Isn't that what most RC chargers do when you have the balance leads attached ?

yes, most of them.. but many forget to connect the balance lead, or select in the menu to take account of these.. or even some charger dont have that option to use the balance lead :roll:

The test that electricbike did was NOT USING the balance lead witch make LVC to trigger sooner and give wrong Ah measurement.

Many person on E-S are generous to test things and that's good! :wink:
.... But problem is that this is not everybody that have the proper equipement to do that and give results that make sense and know how to use their equipement.... :?

Doc
 
Hillhater said:
Doctorbass said:
........The best in that case is to use the balance lead of the charger to monitor voltage and not the discharge lead.. so this way the charger knwo exactly the true voltage of the cell.

Doc
Isn't that what most RC chargers do when you have the balance leads attached ?
Icharger 3010B does measure on balance leads if you have them connected and LVC is based on balance led measurements, not on main leads. My cheap HK eco8 do not do that and I wonder if Turnigys do, BTW balance leads must be at least attached to the battery.
 
There is even bigger problem, when IR of the battery rises, discharge will stop on higher voltage every time. In any way, he is doing a very good job, documenting and sharing.
electricbike said:
Overclocker said:
1c charging isn't what people usually do. So your results won't be representative of real-world
I will charge at 0,5C in my real pack, only 1C just now to speed up my test. What C rate do people usually charge at?
I use about .5-.25C in real life. You could do some IR tests every hundred or few hundred cycles and one very slow discharge cycle to have real capacity.
Thanks for sharing!
 
Ok, perhaps some changes to the methodology can be done, but....!

@electricbike - don't change your testing procedure in mid-test. Finish what you are doing and once your test is done, like after 500 simulated cycles, then reconfigure your testing if you feel you have the need to do so. Then retest.

Thanks for putting in the time and posting your results!

:D
 
electricbike said:
Overclocker said:
1c charging isn't what people usually do. So your results won't be representative of real-world
I will charge at 0,5C in my real pack, only 1C just now to speed up my test. What C rate do people usualy charge at?

i believe 1c charging leads to rapid cell deterioration. so in the end your results may not be very useful. max 0.5c should be ok

to speed it up just use higher discharge current. most people using the 25R discharge it much higher than 1C. it would be a waste to use 25R on 1c applications :D
 
Overclocker said:
1c charging isn't what people usually do. So your results won't be representative of real-world
5 minute wait time is also not what people do.
So in the end these test show nothing useful, just to compare the cells one to another. But results in real world are very much different.

Overclocker said:
i believe 1c charging leads to rapid cell deterioration. so in the end your results may not be very useful. max 0.5c should be ok
Depends on cell in question. In my tests, panasonic B type died after 30 cycles, while LG D1 did 500 cycles to 80% capacity on same conditions (1C charge).
 
circuit said:
Overclocker said:
1c charging isn't what people usually do. So your results won't be representative of real-world
5 minute wait time is also not what people do.
So in the end these test show nothing useful, just to compare the cells one to another. But results in real world are very much different.

Overclocker said:
i believe 1c charging leads to rapid cell deterioration. so in the end your results may not be very useful. max 0.5c should be ok
Depends on cell in question. In my tests, panasonic B type died after 30 cycles, while LG D1 did 500 cycles to 80% capacity on same conditions (1C charge).
Yeah well I am not too sure about all this, I charge my 18650 pack often over a 12 hour period over night. And its currently 7P configuration and I am rebuilding it to 8P on a 500w S12S/Bafang setup so its not going to draw a lot from each single cell either. I only use a PAS setup these days and I often go riding on pretty tame speeds as well with assist level only being 2-3 most of the time.
 
This is great! Keep it up.

1c is well within the rating of the cell, it's a perfectly valid test. I charge my 20s10p pack of LG HE4 cells at just under 1c every day.
 
Interesting to compare to the Samsung pdf data sheet at 140cycles, which indicated a 20% capacity loss ( but 20 amp discharge and 2.5v lvc )..
..you are showing only 5% capacity loss with the lighter discharge and higher lvc !
infact you have more capacity at 140cls (2088mahr), even though you are not going below 3.1v,.. than Samsung indicates (2000mahr) using 2.5v ! :D
But how much of that gain is from a lower discharge rate , and how much is from your higher lvc is still unknown ?
 
Used a free graphing program to transfer info for comparison purposes.

Added the graph from the 4A charge 20A discharge
Updated:
Cycle Life vs Capacity.jpg

Free Graph program - use to open attachment after saving file and changing extension from .doc to .grf.

Charge voltage is a minimal differential. (4.15V vs 4.20V)
I would have to attribute capacity loss from manufacturers test as attributable to high charge\discharge rates.
Lower charge voltage combined with a not so deep discharge likely contribute partially but not substantially to OP cell's much more stable capacity.
 
Optimal discharge depth looks to be 3.30V at the 1C discharge rate.
This likely relates to a ~3.40V static voltage (After discharge removed)

file.php
 

Attachments

  • INR18650-25R.jpg
    INR18650-25R.jpg
    115.3 KB · Views: 2,456
Back
Top