Sanyo 18650GA low voltage cutoff

wannesd

100 W
Joined
May 23, 2012
Messages
243
Location
Belgium
Guys,

I know the recommended LVC here is 3,5V, so batt will bump up to 3,65V resting voltage.
Could it be that this just doesn't hold through for this particular cell?

I did a discharge from full to 3,5V on a brand new genuine 18650GA, only got 2292mAh out of it, instead of the rated 3450mAh.

I found some tests of this cell, with discharge curves. Even at 10A discharge, the cliff only appears at 3V.

What LVC would you guys recommend for this cell, based on this graph?

ncr18650gavsncr18650bnzury.png
 
wannesd said:
Guys,



I did a discharge from full to 3,5V on a brand new genuine 18650GA, only got 2292mAh out of it, instead of the rated 3450mAh.



ncr18650gavsncr18650bnzury.png

2 thoughts: 1) What current level did you test these at? 2) Where did you buy them from?
 
The cutoff voltage should heavily depend on your typical discharge current. If you want to maximize cycle life, the best thing to do though is not to focus on the bottom cutoff voltage, but instead focus on how full of a charge you are giving the cells. For example 4.10v charge voltage has been proven to yield about 2x the cycle life, while providing 90% cell capacity. 4.0v charge voltage is about 4x cycle life but still about 80% cell capacity. Even more important than that is how long the cells spend above 4v. If you are regularly storing the pack at voltages above 4v especially ~4.20v there will be much more cell degradation. In some long life military battery applications Li-ions are charged to 3.92v in order to get a 10-20 year service life.

If you do a 3.5v @ 10A cutoff you have almost no capacity. But 3.5v at 0.2C and the cell is almost depleted. The above poster is right. The industry standard for testing and rating cell capacity is 4.20v CC/CV with usually around 50mA cutoff current then discharged at 0.2C @ a temperature of 25C. But generally the cycle life of most cells in this environment will be in the 300-500 cycle range to 70-80% original capacity output.
 
Yeah,

I charged at 1,4A to 4,2V, let them rest for 30 minutes, and discharged with 1A.
These cells will go in a 20s10p config, max batt amps for my controller is 65 amps.
That gives me a max draw of 6,5 amps per cell, but that will probably be sporadic.

I'd be willing to sacrifice the top ,1-,2 volts for better cycle life, if it only costs me 10% of capacity.

So 3 volts LVC it is then?

edit: I bought them from here: http://www.nkon.nl/rechargeable/18650-size/sanyo-18650ga.html
 
18650 cells are not rc lipo. LVC will be different.
 
Merlin said:
this is a trustfull seller for cells..... but not really cheap.
is this your first battery build?

First with 18650, yes. Previously I've always used rc lipo.
 
Alright, new data:

Programmed in a new cycle:
charge to 4.1V (cutoff at 0,1 amps) @ 1.4A
Rest 30 minutes
Discharge to 3.1V @ 1A
rest 40 minutes
charge back up with same settings as first charge.


Results for first 4 cells:
Cells IR Capacity
1 34 mOhm 2926 mAh
2 34 mOhm 2899 mAh
3 32 mOhm 2841 mAh
4 35 mOhm 2887 mAh

the 3.1V under load went back up to 3.21V resting.

Still a long way from the 3450mAh that they are rated for...
 
joeyan310 said:
yes I am getting the same results discharging to about 3.1v

Ince this is just a test setup to weed out any bad cells, it does not really matter, just looking for consistency.
In my bike though, I'd like to get the best blend of lifespan and range...
maybe 4.1 to 3V?
 
If you doesn't know about the "dampfakkus" website...

Here you can check or compare results to your cells

https://www.google.de/url?sa=t&source=web&rct=j&url=http://www.dampfakkus.de/akkutest.php%3Fid%3D609&ved=0ahUKEwjEkOCm94nNAhWkdpoKHVTQCrMQFggbMAA&usg=AFQjCNGjwvR58idtlYMmNWqsQ0NopcpaUQ
 
wannesd said:
Alright, new data:

Programmed in a new cycle:
charge to 4.1V (cutoff at 0,1 amps) @ 1.4A
Rest 30 minutes
Discharge to 3.1V @ 1A
rest 40 minutes
charge back up with same settings as first charge.


Results for first 4 cells:
Cells IR Capacity
1 34 mOhm 2926 mAh
2 34 mOhm 2899 mAh
3 32 mOhm 2841 mAh
4 35 mOhm 2887 mAh

the 3.1V under load went back up to 3.21V resting.

Still a long way from the 3450mAh that they are rated for...
What did you expect ?
You are undercharging them , and not fully discharging them ! :roll:
 
Yes I don't understand why you continue to test the cells in conditions outside of the ways the cells are rated.

0.2C Discharge (0.7A) with a 4.20v charge voltage and 2.50v cutoff voltage @ 25C temperature.
 
Because I'm not testing to see if they will get the same Ah as rated, I'm testing to see what dV I should use. 4,1 - 3,1V seems to yield good enough results, maybe I'll even go down to 3V.
The reason I'm testing at 1A, is because my charger will only let me test at 1A max...

I guess I'm trying to find what the safe limits are, without hurting lifespan too much.
 
charge them to 4,15v and discharge to 3,2v maximum (resting voltage)

if you use this pack every day you get easy 5 Years to ride with it. in 5 Years you want the new cells with 25C and 5ah capacity ;)
 
I run my battery packs based on two things:

Capacity - Using the CA is the best thing ever since you can actually know how may AH you have discharged. From this measurement I try to discharge my pack to a 80% max, ie. if it is a 10AH pack I'll only discharge 8AH.

LVC @ Max Controller Rating - I do this so that the controller doesn't cut at a higher voltage. In my newer 25R packs drawing 10-20 amps from each cell for 5 secs is fine but it will sag. I usually just take my max controller current divide it by the number of cells I have in parallel and then find the current per cell. Then I go to a discharge graph and find the voltage at that discharge current and add maybe .1v or a bit more to be safe then multiply that by your total of cells in series and that's the LVC.

Both of these things you can control from the CA as well. I'm not sure there is a controller out there that'll let you cut throttle based on AH discharged.
 
Here are some minimal discharge rate graphs.
Note: there are 3 specific formulations that "map" 3 very different discharge cliff points.
3.6V and 3.4V and 3.2V
Discharge rate is so minimal that these voltages could be considered as the static discharged voltages, in practice DOD would be determined by personal discharge rate.

file.php


Based on your supplied graph I would recommend a 1A discharge to 3.0V with a resultant 3.20(?) static voltage as the result.
This might meter as only 3250mAh, but I would recommend as much better for battery.
Discharging deeper supplies minimal capacity with steep voltage sag.

If your discharger has programmable discharge V ...
discharge to 3.40V record mAh & recovered "static voltage" then continue
discharge to 3.35V record mAh & recovered "static voltage" then
discharge to 3.30V record mAh & recovered "static voltage" then
discharge to 3.25V record mAh & recovered "static voltage" then
discharge to 3.20V record mAh & recovered "static voltage" then
discharge to 3.15V record mAh & recovered "static voltage" then
discharge to 3.10V record mAh & recovered "static voltage" etc?

or
Meter discharge towards 3.00V, record mAh at every increment - .1V or .05V or .02V etc
Discontinue when mAh between increment reduces sharply.
Allow to settle-recover and record as static DOD.
Apply your "typical" load and record resultant voltage as "your typical" maximum DOD.

Cliff - DOD will be optimal at last mAh reading before drastic drop in capacity.

See also - Capacity Mapping
 
Found up a discharge graph that includes a .2A discharge rate

Sanyo NCR18650GA 3500mAh (Red)-Capacity.jpg

Based on low discharge rate data ... (unlike most Lithium formulations there is no abrupt cliff where voltage and capacity drop off. )
I would revise my recommended LVC near a static 3.40V
I added green line to indicate my "Safe" recommended discharged voltages at different discharge rates
I added yellow line to indicate my "use Caution" maximum discharged voltages at different discharge rates
I added red line to indicate my "Danger" maximum discharged voltages at different discharge rates

Note the 15.0A rebound to static voltage lines!
Maximum charted DODs look to rebound to a static 3.20V.

See full tests Sanyo NCR18650GA
 
Back
Top