Help. 8s 105ah cells yielded 173.16ah?!

[nebulight]

I setup my test setup with the following:

8 EVE 105AH cells
EPever ip2000 24v inverter wired to the ExtIO to disconnect for low voltage (switch only draws 15ma)

DSSR20 for solar (10 amp power supply acting as solar for now since it's inside)

500amp shunt 50mv shunt (it's what I have on hand as I plan to replace with a 150amp shunt for final install).

I fully charged all the cells to 3.6v, then paralleled them and again charged to 3.6v and let sit for two days in parallel. I then arranged in 8s for a 24v pack. I ran my home office off the inverter drawing anywhere from 5-9amps throughout the day (lights, computer monitors, laptop, phone charger, small fan, ect). I left it run for several days. It finally died and the inverter shut off, the counter was at 173.16ah. This doesn't seem right at all. Here are the screen shots from the test:

I don't think it's an issue with the shunt calibration because when I charge the battery with no other load, it's the SBMS is reporting the correct current (close enough):

I'm in the process of recharging the cells 10ah at a time (10 amps for one hour when I'm at a place where I can have the power supply fans running) so I can get a good idea of capacity going in since I don't have a PV shunt.

Since the discharge was at such a lower rate, i figured I might squeak out a time bit more capacity, but this doesn't seem right. Any ideas? Did I win the lottery with battery capacity?

[Dacian Todea]

The 173Ah is sure not correct.

It seems that you set the current shunt to correct 0.1000mOhm (guessing that was set correctly from the beginning).

I also guess you reset the energy counter when you started the discharge test.

Based on photo with the 10A charge current it seems current shunt was set correctly and was measuring correctly at that point in time.  Was the SBMS0 ever powered OFF while all this images where taken ?

Since all look OK based on photos and description I can only guess that you have a bad connection on the shunt sense wires ADC1n/ADC1p and at some point ADC1n got disconnected maybe for a few minutes but if you checked the graphs you should have seen that spike to 900A (tho your test lasted for a few days so you may not have noticed).

Since there are maybe around 70Ah extra that means ADC1n should have been disconnected for about 5 minutes (could be all at one or a few times adding up to around 5 minutes).

If the ADC1n is disconnected (either on shunt side or on the SBMS0 connector) the battery shunt measurement will show a 900A load (max what this shunt can measure) and in a bit less than 5 minutes that will add up to around 70Ah

Maybe when you lifted the SBMS0 to look at it (possibly a few times during the test) the wire did not made contact and the SBMS0 incorrectly seen 900A adding to the energy counter error.

[Dacian Todea]

One thing I notice and forgot to mention is your 4100 error code that means the Under voltage Lock flag is set and that should never happen in normal operation as that means one for the cells drop below 2.5V (unless you changed settings in advanced parameter settings).  I can not see in any of the photos cell's below 2.5V as all are always above 2.8V so not sure how that happened or if you changed things in advanced parameter settings.

[nebulight]

Thanks for the pointers Dacian. It's possible that may have happened as I was picking it up several times a day to check the status. I'm doing some work around the house today so I'm letting the charger fully charge it, then I'll do a higher current discharge test with a space heater or something to draw it down quick and I won't touch anything.

As for the under voltage flag, not sure how that happened as the only setting change I made from your default was to set the discharge voltage to 2.8 from 2.9. Maybe I changed the wrong setting. I'll check a little later once I'm done with all this testing.

I do have another question.  If the BMS is shutoff, is the counter information and SOC stored in memory? This is going to be in a portable box and was thinking about discharging to about 50% then turning it off when not in use. Is that a bad idea? 

[nebulight]

and btw, I did take a photo of the screen, it was just under voltage, not under voltage lock out:

On Saturday, January 23, 2021 at 12:27:57 PM UTC-8 electr...@gmail.com wrote:

[Dacian Todea]

Sorry I interpreted the errors incorrectly it is just normal Under voltage not Under voltage lock so normal operation (the battery was just fully discharged) the 2.8V is the default if I remember correctly.

With LiFePO4 you do not need to discharge the battery for storage anything in the 80 to 99% SOC will be just fine. Other type of cells are much more affected by being fully charged but not LiFePO4

You need to secure the wires either to front panel or in your case since is a test setup and SBMS is not installed to a front panel then secure the sense wires to the SBMS0 body so that wires do not put strain on the connectors.

[nebulight]

Second test was good. Thanks for the pointer, maybe I nudged it when picking it up several times.

[Dacian Todea]

Thanks for the update. That looks in line with what is expected from those type of cells.