diy solar

diy solar

Faster way to test single cells?

Shale MacGregor

Solar Enthusiast
Joined
Sep 5, 2021
Messages
641
I need to capacity test so I can match ~ 400 cells, is there a faster way to do this than limiting it to .2C ? I tried testing at .5C thinking I would adjust "expected" capacity and it would enable me to match cells, unfortunately this caused voltage sag below the 2.5V mark waaay too soon. These cells are rated at 2C and are 60AH.

I use a Siglent Load Tester that can do up to 200 Watts, and I have a shunt that works great using a load with an inverter for real world testing, but that only works for 4s.

I was considering hooking up a large enough array to max out my 1000 watt inverter at .2C ( 7p4S), and when one cell hits the minimum I could stop, give all of the batteries the minimum score (divided by 7) and then by testing voltages break them into more groups, or else just test each one from that point forward.

Is this a good idea, or does someone have a better one that would be advisable. Id rather not buy a bunch of shunts / load testers just to save time but at this rate it will take me a few months.


Thanks!
 
Looking for feedback on the method of pulling load on 4s or 16s pack then continuing one cell at a time and adding results.
 
Discharging 16s pack to weakest cell limit and continuing invidually from there is valid idea. 7p4s pack would be more difficult as you can’t be 100% sure of the current dividing equally between cells/strings.
 
Last edited:
Discharging 16s pack to weakest cell limit and continuing invidually from there is valid idea. 7p4s pack would be more difficult as you can’t be 100% sure of the current dividing equally between cells/strings.
Thanks! I've actually come up with a way to quickly add/remove batteries to testing rigs so my only real hurdle is settling on a methodology that is legitimate before burning through a few weeks of work.

1632565501828.png

I have an overkill solar 16s BMS, I will top balance each group of 16s at 3.65V using 120A charger, and then discharge at 11A (.2C) which should take around 5 hours, using a 48V Inverter and a ~550 watt load (could i increase this C rating to .5C if I am more worried about cell matching than what the actual rated capacity is? IE, if I discharge at .5C and these cells which could be 55AH give a result of 40AH, I am more worried about which cells fall into the grouping of 38AH, 40AH 42AH, etc).

Otherwise at 5 hours per day, I might get 2 of these tested each day which would be a few weeks just for this step, the comes individually discharging the cells to get the total.
 
Shale what did you end up doing to test your 400 cells? I would have hoped that it might be possible to use e.g. internal resistance of a cell at a known soc to determine expected capacity. With 400 cells you would have been the perfect person to find out... Also, I guess that once you used your .2c discharge rate on some cells to measure capacity you could discover their behaviour at .5c and rest the rest at .5 to save quite some time...
 
Shale what did you end up doing to test your 400 cells? I would have hoped that it might be possible to use e.g. internal resistance of a cell at a known soc to determine expected capacity. With 400 cells you would have been the perfect person to find out... Also, I guess that once you used your .2c discharge rate on some cells to measure capacity you could discover their behaviour at .5c and rest the rest at .5 to save quite some time...
Still testing :)

You are in luck, I did undertake that experiment. The IR at a known SOC gives me a guideline that is correct, some of the time, but nothing so much as to make it easy to match capacities with any accuracy or reliability.

I have tested at .1C .2C and now ~ .5C and have had no significant differences. I am testing them all at .5C now.

I ordered some more load testers, but that is also because I have another batch of other batteries to play with.
 
Thanks a lot for the information, that will save me quite a bit of time!! With so many cells to test and more coming in, I am not surprised you are closely looking at test speed. I was looking at this, might also be useful given a good base cell:

 
Thanks a lot for the information, that will save me quite a bit of time!! With so many cells to test and more coming in, I am not surprised you are closely looking at test speed. I was looking at this, might also be useful given a good base cell:

thank you for posting this research, very interesting
 
S/NCapacityIRstarting
00159,8280.543.375
00255,7020.693.344
00357,9780.553.570
00460,2200.523.570
00560,0030.553.570
00660,4950.483.570
00758,9530.593.570
00858,6270.493.580
00960,8030.583.613
01060,7030.563.588
01159,8200.543.344
01257,6100.613.343
01357,8850.583.341
01459,8520.543.342
01559,3610.613.344
01656,6520.623.342
01758,4860.653.342
01859,7190.583.342
01958,7190.603.342
02057,9440.593.342
02158,4360.563.342
02260,6860.593.342
02360,3860.553.342
02459,6110.563.342
02552,5420.723.343
02653,6090.703.342
02757,7270.673.343
02859,6520.593.343
02958,0350.653.343
03060,3610.583.343
03158,9690.523.341
03256,7180.783.340
03360,0280.683.340
03456,9270.583.389
03554,2260.603.350
03654,4340.523.343
03758,7690.613.343
03858,4020.623.339
03958,2520.623.337
04060,5440.573.339
04156,2930.703.340
04259,7700.673.340
04361,3110.563.350
04457,2770.603.360
04559,8360.603.360
04657,0110.683.337
04752,7420.733.343
04854,9430.713.425
04959,1860.533.398
05059,4190.533.376
05155,9430.623.354
05260,3620.603.354
05360,2030.553.435
05454,5010.693.351
05560,2780.553.351
05659,2110.533.353
05759,7120.553.335
05859,8950.623.333
05955,1850.673.345
06058,6880.593.345
06161,3030.563.440
06260,9370.593.453
06355,6350.783.337
06458,8270.613.338
06559,5910.523.341
06656,8600.613.340
06755,6910.653.338
06856,6510.693.341
06957,4850.653.339
07056,3930.603.343
07158,2270.643.335
07260,8780.583.386
07355,8180.583.334
07457,0770.593.337
07558,3020.653.337

Feel free to poke through what I have so far and see if I missed anything in terms of correlation.
 
thank you for posting the data! very appreciated!

good going getting through so many cells
 
Back
Top