diy solar

diy solar

The downside to commodity cells...

snoobler

Solar Honey Badger
Joined
Jul 10, 2020
Messages
8,390
Location
HBR, AZ
...is you never know what you're going to get.

Ordered from Amy and received about 7 weeks later.

In January I tested these 9 Eve 280Ah cells. I ordered a spare in case there was an issue with one of them, so I didn't have to wait for a replacement.

I left them fully charged and they sat that way for almost exactly 5 months. Yes. I know. Better to store them at 40-70%, but that wasn't the goal.

No one ever runs the self-discharge test in the data sheet. Neither did I; however, I conducted an alternate test that was a little less painful and likely has comparable value. Instead, I charged 'em up, let them sit 5 months and then measured the amount of mAh needed to recharge to 3.65V @ 20A with 2A tail current.

Here is the table of test results:

SNDateDCH1SoCCHG1DCH2CHG2DateVoltsCHG3SoC
02YCB66710000J89S00009711/11/20217929.10%2712742726/18/2021NA27899.90%
02YCB66710000J89600002571/11/202110939.20%2772802786/18/20213.4356399.80%
02YCB66710000J8CH00005541/11/202110238.17%2672632686/18/20213.3867699.75%
02YCB66710000J8CJ00021191/11/20217627.97%2722692726/18/20213.3955499.80%
02YCB66710000J89R00024151/13/202110839.56%2742722746/18/20213.4532099.88%
02YCB66110000H89E00035221/13/202110739.53%2712732716/18/20213.4161799.77%
02YCB66710000J8B700020241/13/202110840.43%2682722686/18/2021NA41599.85%
02YCB66710000J8A400020481/13/202110839.73%2712722716/18/2021NA30599.89%
02YCB66710000J8CJ00027381/13/202110138.12%2642652646/18/20213.30307098.84%

All cells were tested individually.
Discharges were conducted at 20A to 2.5V.
Charges were conducted at 20A to 3.65 and a tail current of 2A.
Discharges followed charges after a 20 minute wait.

I didn't record them, but all 9 voltages were the typical 3.29X with IR in a .15 to .19 mΩ range.

DCH1 is the as-received capacity via discharge to 2.5V @ 20A.
CHG1 is is the post-DCH1 charge to full.
DCH2 is the capacity of the cell.
CHG2 is the post DCH2 charge to full (the amount input prior to the 5 month sit).
CHG3 is the mAh input to achieve full charge.

Items of note:
Cells were shipped at 28-40% SoC.
Cell voltages ranged from 3.30 to 3.45, yet they were all above 98.8% SoC (everybody's favorite chart says 3.30V is 70%).
All failed to meet datasheet capacity.
All met self-discharge datasheet capacity, but one was a clear outlier losing over 1% capacity after 5 months.
8 of them are reasonably well matched in capacity and self-discharge (268-274Ah)
The ninth had both the lowest capacity and the highest self-discharge.

Musings:
Discounting the outlier, a sit for this period of time and the resulting drift in SoC would likely have caused at least a slight headache in a 8S battery. Even the .15% difference in those 8 cells would likely cause an annoying runner at peak charge.

Had cell #9 been used in the pack, it would have been a shit show. The runner would be violent, and extremely frustrating to restore the pack. It would have needed 72 hours of passive balancing during charge to burn off the excess from the other cells assuming it can do so on all 7 at the same time. This 72 hours would likely have been spread over 2 weeks of typical operation.

Conclusions (speculations):
"Grade A" commodity cells are not grade A - failure to meet datasheet capacity.
Voltage and IR matching are insufficient to catch all problems.
Cell selection process at the "distributors" while filling orders is not robust.
Top balances have a shelf life.
Don't top balance until you're ready to put the battery in service shortly thereafter.
If they sit too long, be prepared to hit each cell with 10A to 3.65V individually. In most cases, this will likely only take an hour or two.
If you use the battery intermittently without regular cycling, don't expect the top balance to hold.
Don't order an oddball number of cells.
The outcome is an acceptable trade-off for the lower cost of commodity cells, i.e., I would do it again.
 
Last edited:
Thahis. I’m getting ready to order from her and am waiting a quote. Mine will likely show up at the end of September if I choose to purchase.
 
Yeah, it seems the quality of these gray market cells has dropped off a bit over the last year. Thanks for the detailed testing. I am looking at building a 30kwhr bank, and it seems conservative voltages, a higher rate balancer, and a few spares are called for when using these lower grade units.
 
Back
Top