diy solar

diy solar

36 EVE 280 cells capacity tested

cinergi

1.21 Jigawatts
Joined
Aug 9, 2020
Messages
1,419
I just completed testing 36 EVE 280AH cells purchased in the Group Buy thread.
Tester: https://www.aliexpress.com/item/32821857011.html
Wire gauge increased to 10. Connected to cell via lugs (in other words - a high quality connection, not the alligator clip method).
Amperage: 15
Temperature: 60-70F
Starting voltage: 3.65
Ending voltage: 2.2 at the tester, ~2.5 at the cell

The IR on all cells measured at 0.20 at 15 amps at the start of the test.

I am confident in the tester as the results line up with multiple shunt readings while the cells were in a battery, including Victron BMV-712.

I wanted to post this front and center so it's easy for everyone to get at this data.

268
269
270
270
270
270
271
272
273
273
273
274
275
276
276
277
277
277
277
277
277
278
278
278
278
278
278
278
279
279
279
280
280
280
280
281
 
So about 2.2% capacity gain from matching capacity in a 2p setup. 268+269 vs 268+281 in the worst vs best case. 537/ 549 = 97.8.

So for a power wall unless a cell is a real dud (which you can localize in a full pack test) or unless you have hardware to automate everything it's probably not worth the time/effort.

I'm happy planning as if they are all 268ah cells.
 
The testing wasn't for maximizing capacity; I wanted to document EVE performance for everyone and also have baseline for all my cells for warranty purposes as well as general history. I also would rather parallel matched-capacity cells so I don't have one cell trying prop up another cell while simultaneously providing power ... or get into a top-balance problem that I'd have every single time I ran one of the cells below the knees because the two cells would unbalance each other (and I wouldn't know this, because the BMS can't see this). The only way to sort of avoid that is to stop discharge using SoC cutoff instead of LVDC cutoff and programming the battery to a capacity of 268 * 2 ... and of course, I don't know what the low number should be until I've tested all my cells.
Time consuming, yes. Didn't take a lot of effort though.
 
Beginner question on math: ~19hr per cell, right?

280 / 15 = 18.67 so yup! One per day. Once I'd tested 4, I built a 12v pack with an old BMS and charged the group of 4 and then parallel-top-balanced them, ready for use.
Basically, every morning I'd check the tester and log the calculated Ah. I disconnected it, set it aside, connected a new cell and restarted the test. That took maybe 2 minutes per day.
Connecting up the 4s battery and initiating a charge took maybe 5 minutes every 4 days.
 
Why only 15A? Do you think the tester would have died at higher amps? I've got 16 cells coming and one of those testers on the way also. I'm debating whether I'll test each cell or just use the tester on any suspect cells. I'd like to test each cell, but I don't love the idea of going to sleep while the $30 gadget is running. I guess I'm not going to be watching it the whole time regardless of whether I'm sleeping or not, so maybe safer running it at low amps.
 
15 'cuz I didn't want to run at max rated. 14 hours vs 19 hours didn't really make a difference in my overall timeline.

I also had the same fear of leaving the tester unattended. After doing it enough times, I trusted it (both that it would stop the test at the right voltage and that it wouldn't start a fire). Same with the power supply.
 
280 / 15 = 18.67 so yup! One per day. Once I'd tested 4, I built a 12v pack with an old BMS and charged the group of 4 and then parallel-top-balanced them, ready for use.
Basically, every morning I'd check the tester and log the calculated Ah. I disconnected it, set it aside, connected a new cell and restarted the test. That took maybe 2 minutes per day.
Connecting up the 4s battery and initiating a charge took maybe 5 minutes every 4 days.
Thanks for this, I will test and post results of my 16 cells as well.
 
Back
Top