I always suggest you add the Min/Max Cell voltage graph in VRM. You will see there that the max should have a spike around when the alarm was raised. The alarm is from the BMS. The balancers will sort it out which is why it is fine later.
If you don’t want to get the alarm you have to make sure that the cells are all reaching close to full at the same time. If one cell is much earlier than all the rest the battery might think it needs to add a lot more power, but that one cell cannot take anymore. If they all are close to full the same time the battery can see it is getting close and will lower the charge rate the last part so and this keeps the alarm from happening.
Top Balance = you balance the cells to reach the top(full) at the same time.
@ebendl High voltage alarms are normally triggered when any cell in the battery reaches 3.6V
I suggest your first step be to make sure the battery hasnt swelled. Make sure you check the battery from behind (the seems will open up slightly) and run a straight edge over the top and bottom to check for bulging.
Any swelling would be a warranty claim and the battery should be removed from the bank immediately.
Once you have determined no physical damage to the battery then proceed to balance as others have mentioned.
Don’t have this available - I’m guessing this is because I’m not running the older firmware on the batteries (US3000a or US3000b batteries from 2020, never touched the firmware).
I have seen some reports that the newer firmware may even reduce these alarms (thought is that they get smaller spikes of high voltage which quickly disappear and the older firmware just sent through the error where the newer firmware monitors it a bit further before raising the alarm).
So if I see 88% for this long, I take it it is seriously struggling to match cell voltages? Been like this (on “Keep charged”, with maximum charge current on 5A and maximum charge voltage on 52.1V) for 8 hours…
Pylontech batteries give an average of the SOC of the modules. I’m not sure if it is a pure mathematical average (which would suggest that one of them is WAY below 88%), or if it is some kind of weighted average.
In any case, I would think it is fine to bump it, and resume after load-shedding.
A nice way to see if the BMS balancer is working hard on a battery is the first temperature reading on MultiSIBControl.
The top reading is the BMS temperature and the two lower temperature readings are individual cells.
When the BMS is balancing due to the cells being at different voltages, the top temperature reading will be quite a few degrees higher than the cell voltages.
So I’m guessing Pack 2 Cell 12 is the culprit here. When I bumped the current it started climbing much faster than the rest - seriously thought it was going to get to 3.55v but then the battery decided that we hit 100% SOC and dropped the charging. I’m guessing now it should slowly drop over time.
Ugh, irritated with loadshedding. Would have liked to contine doing this nice and slow
@Thaelian yeah the pack temperature was 29 degrees while the cells are 25 or 26 degrees C.
Delta - Highest 3.520v - Lowest 3.475v = 0.045v which is ok, but can be better at charged. More like 0.005v Delta. That is where the balancers must do their work.
The peak current of 1.502a vs lowest of 0.622a - cool data!
Yeah, just haven’t found a way to log all of this. Fine if one happen to be there while it occurs.
Buddy of mine apparently wrote a python script to run on a Pi to do exactly this - will see if I could get it going over the weekend. Would also be good to see if it is actually getting better…
Looking at those cell voltages, I don’t see an imbalance. When practically all the cells are above 3.485V or very close to it, and one jumps out to over 3.5V (but still less than 3.55V)… that’s not an imbalance. That’s pretty good.
There must be another reason why the SOC is so badly out of sync.
@ebendl You should consider upgrading the firmware on the batteries, the 111% SOC is a red flag for me. Get some assistance if you are unsure exactly how to do it though as it can brick the BMS
Nope, but all good ideas. I will give some of it a go over the weekend, or consider the firmware upgrade.
Edit: everything very close to one another now, and hopefully enough time today to charge very slowly. I hit loadshedding last night, SOC dropped to 91% and then when power came back on charging resumed slowly. It’s been sitting there since.
I recall someone from Pylontech mentioning to me that they DO go over 100% in their internal measurements. Of course they never show that to a customer.
We also cap it to 100% in the can-bus-bms code, also because the protocol allows sending up to 65535%… and one day… someone did. So a customer will never see more than 100%, but internally, if a cell is charged slightly higher than design spec, it could be over 100%.
Just like dams. Dams are sometimes 110% full.
That is to say, the 111% isn’t too weird to me. It just tells me the cell is very high, which I already know.
Ok I shut down everything today and took out all three batteries.
I could see no split seams of any sort yet
I thought there was bulging on the middle battery (pack 2, the one with the cell that “rapidly” gained charge previously, but when I measured all three it looks like it’s the same on each battery.
I swopped them around (pack 2 became pack 3, pack 3 became 1 and 1 became 2)
Started them again. The new pack 1 (master battery) made a “clang”, had the “Alarm” light flash for a while but then returned to normal… not sure what to make of this???
Everything seems to be back to normal. Still charging at a slow 5A (DVCC limits).
OK so couple of days later, I am still charging fairly slow (10A) to lower voltages (52.7v, but I am increasing it with every 100% I reach by 0.1v). Loadshedding pulls the batteries to about 85% or so, and afterwards “Keep batteries charged” basically immediately starts recharging them at a lower current. Haven’t had a high cell voltage alarm in a couple of days now.
None of the 3 batteries seem to go completely out of wack in terms of cell inbalance - all three has basically the same lowest and highest cell voltages.
