I added 2 new v1 slaves to my existing v1 setup 6 months ago.
Installed as per guides on here, ensuring exisitng stack was 70% soc, got all green lights etc, now have 11.2kwh capacity.
However since day 1 ive experienced sudden unexplained drops in the soc as the soc drops below 40%
I understand the existing batteries were 2 year old and so may not have 100% health, and that it can take weeks for the system to balance.
However 6 months later my system is still dropping between 12 and 20% in one sudden dip as the soc drops below 30%
It's only really become noticable again in recent weeks as less sunlight means more battery use.
For the last week Ive been following guidance from here re cycling the batteries from 10 to 100%, slow charging at 5amps, and letting it sit at 10% for a few hours to bottom balance before charging. But it doesn't seem to be improving.
Ive used a tool found on here to check cell voltages remotely but am not sure if its telling me something useful... cell 18 on all 4 battery units consistently is 0.02v lower than the other 17 cells in each unit... is this normal?
Can I just double check you actually meant 70% soc when you added the new batteries? - as the recommendation is 50% +/- 3% a 20% difference will take time - admittedly 6 months should be plenty but only if you have done a number of regular grid charge cycles to 100% and back down to 10%.
The graph you have posted looks ok, cell 18 usually does that - i’m guessing that is an internal design thing with lead lengths or sensor pickups so that all looks fine, but the best time to run the graph is when the battery stock does (or has done) it’s soc drop as that will show which of the batteries (or cells) is falling off - could you repeat the graph when you have the rapid fall off.
The graph you have posted looks ok, cell 18 usually does that - i’m guessing that is an internal design thing with lead lengths or sensor pickups so that all looks fine, but the best time to run the graph is when the battery stock does (or has done) it’s soc drop as that will show which of the batteries (or cells) is falling off - could you repeat the graph when you have the rapid fall off.
Thank you Dave,
Glad to know the cell 18 voltage is not out of the ordinary.
I found an installation guide on these forums that said to have the battery at 70% and followed that guide to the letter, unfortunately it looks like it got me off to a bad start.
They've had a dozen or so full cycles since being installed, but not very often as the battery has been topped up with solar during the day so hasn't discharged down to 10% much at all over summer.
they're getting fully depleted by 9pm daily from here on, and charged over night.
Ive used a home assistant automation script I think you wrote to reduce charge current once it gets to 93% so will run that for a week or so and see how it goes.
Thank you
Glad to know the cell 18 voltage is not out of the ordinary.
I found an installation guide on these forums that said to have the battery at 70% and followed that guide to the letter, unfortunately it looks like it got me off to a bad start.
They've had a dozen or so full cycles since being installed, but not very often as the battery has been topped up with solar during the day so hasn't discharged down to 10% much at all over summer.
they're getting fully depleted by 9pm daily from here on, and charged over night.
Ive used a home assistant automation script I think you wrote to reduce charge current once it gets to 93% so will run that for a week or so and see how it goes.
Thank you
ok keep an eye on it, the batteries are designed to maintain small differences but a 20% swing will take it much longer and the slow charge at the top will really help with that.
Sorry I should have said, also make sure your battery firmware is the same on all of your batteries
Sorry I should have said, also make sure your battery firmware is the same on all of your batteries
Hi please Can you help me understand the normal ranges for cell voltages?
I've got one battery with quite a large range within the battery unit which becomes more pronounced the lower the level of charge, and theres also a difference between the newer and older batteries (which I expected given the starting point was 20% off) .
From what I can deduce, 60v (18 cells at 3.33v) indicates the battery is fully charged, and 56.7v ( 18 x 3.15v) indicates the battery is around 10%.
Is this roughly correct?
I've got one battery with quite a large range within the battery unit which becomes more pronounced the lower the level of charge, and theres also a difference between the newer and older batteries (which I expected given the starting point was 20% off) .
From what I can deduce, 60v (18 cells at 3.33v) indicates the battery is fully charged, and 56.7v ( 18 x 3.15v) indicates the battery is around 10%.
Is this roughly correct?
Attachments:
As a rough rule an LFP cell is between 3100mV at 10% and approximately 3450mV at 100% - the upper voltage can be a bit higher if it has just charged and it sags after the charge has ended to ~ 3400mV which is still 100%.
Other complexities are the charge voltage has to be higher when the cells are colder and so you may see them as high as 3550mV in winter.
When approaching 100% the cell voltage rises rapidly and that indicates to the BMS it is getting close to maximum, likewise as it approaches 3100mV the cells voltage drops rapidly and so the BMS cuts power and starts a maintenance charge at that level.
The ‘problem’ with multi-cell batteries is that you have a number of different cells that should all perform the same in the lab but in a real environment some of them will be warmer than others and so you usually see the same cell fall off at the bottom and at the top of the charge cycle - but it’s how the pack works together that is important and they should be in reasonable harmony from 15% up to 97% outside of those limits you’ll start to see some separation.
All of that said if you have what appears to be some cells that are still 100mV above others at the bottom or the top that indicates they are not in perfect balance.
The individual battery software is responsible for the inter-cell balance, the BMS is responsible for maintaining the pack.
If you hard discharge your batteries regularly (i have a friend who does full dumps to the grid at nearly 6kW) - the pack and individual cells never have the chance to balance - their individual charge circuits can manage ~30 watts/hour and so if the batteries are out of balance and you never give them chance to recover they won’t, and hard discharging causes balance problems.
Make sure all your slaves are on the same version of firmware and also your BMS is running latest.
Where you think your cells (or packs) are out of balance - to aid with ‘bottom’ balance allow the batteries to discharge to minSoC (10%) and then leave them there for at least several hours, longer if possible - you will see the BMS do occasional low power charges and the individual cells that may be low will be prioritised for that charge helping them balance.
Likewise at the top (at least from 94% on) reduce the charge to 5A which allows them the time to balance for longer - it still won’t happen overnight as a 20% difference (580 watts) at 30 watts per hour balancing will take 20 hours of doing that before they will be in balance. At 5A the 94%-100% will take maybe 30 minutes so that 20 hours can take a lot of charge cycles to recover.
Other complexities are the charge voltage has to be higher when the cells are colder and so you may see them as high as 3550mV in winter.
When approaching 100% the cell voltage rises rapidly and that indicates to the BMS it is getting close to maximum, likewise as it approaches 3100mV the cells voltage drops rapidly and so the BMS cuts power and starts a maintenance charge at that level.
The ‘problem’ with multi-cell batteries is that you have a number of different cells that should all perform the same in the lab but in a real environment some of them will be warmer than others and so you usually see the same cell fall off at the bottom and at the top of the charge cycle - but it’s how the pack works together that is important and they should be in reasonable harmony from 15% up to 97% outside of those limits you’ll start to see some separation.
All of that said if you have what appears to be some cells that are still 100mV above others at the bottom or the top that indicates they are not in perfect balance.
The individual battery software is responsible for the inter-cell balance, the BMS is responsible for maintaining the pack.
If you hard discharge your batteries regularly (i have a friend who does full dumps to the grid at nearly 6kW) - the pack and individual cells never have the chance to balance - their individual charge circuits can manage ~30 watts/hour and so if the batteries are out of balance and you never give them chance to recover they won’t, and hard discharging causes balance problems.
Make sure all your slaves are on the same version of firmware and also your BMS is running latest.
Where you think your cells (or packs) are out of balance - to aid with ‘bottom’ balance allow the batteries to discharge to minSoC (10%) and then leave them there for at least several hours, longer if possible - you will see the BMS do occasional low power charges and the individual cells that may be low will be prioritised for that charge helping them balance.
Likewise at the top (at least from 94% on) reduce the charge to 5A which allows them the time to balance for longer - it still won’t happen overnight as a 20% difference (580 watts) at 30 watts per hour balancing will take 20 hours of doing that before they will be in balance. At 5A the 94%-100% will take maybe 30 minutes so that 20 hours can take a lot of charge cycles to recover.