diy solar

diy solar

Pack assembled but cell voltages all over the place

hedgehog911

New Member
Joined
Oct 5, 2021
Messages
3
My first time building a pack, but after reading here for weeks. I've got a really odd problem....

So this is a 16 cell 48V unit. I top balanced all the cells by connecting all in parallel and charging at 3.65V until current dropped to zero. This looks weeks and weeks! Then I assembled into the pack as shown (sorry for the messy balance leads)…

20220407_174450.jpg

I connected the Overkill 100A 16S 48V BMS, and then completed the -ve connection to a Victron Multiplus-II GX 48V 5000VA. I got a spark on connecting the -ve, suggesting the capacitors in the Victron were drawing current. However the Victron didn’t appear to switch on.

I then logged into the BMS using the Android software, and found that the 5 cells nearest the -ve end of the pack are showing higher voltages than I’d expect...

Screenshot_20220407-174227_xiaoxiang.jpg


Somewhat alarmed, I’ve disconnected the pack by taking out the link between cells 8 and 9 (top of my picture above) which I assume must stop all current flowing.

I’ve checked the voltages with a Fluke multimeter, and this confirms that same as shown on the BMS above.... that the 5 cells nearest the -ve connection of the pack appear to be much higher voltage than I’d expect from a resting battery.

I don’t really want to reconnect anything until I understand what’s going on. Has anyone experienced anything like this?

Dave

PS - I can still connect to the BMS at the moment, even though the battery has that link taken out and so is reading 0V. How can this be?
 
Why do you think those 5 cells are that high?

only thing i come up with:

-Those 5 just got off top balance and had not had time to reach 3.34. The 5 that just got off are at 3.6. Can’t remember how long my cells needed to be off to reach resting voltage, but somewhere shorter than overnight.

Looking in the Overkill manual, this is how wire sensors are supposed to be wired, which you may have correct, I am having trouble checking BC11-16:



9591F055-B00D-482C-A471-E19DF5C55856.png
During assembly, i got some funny readings when my 8S overkill was wired incorrectly. When i fixed the wiring, it was fine. For me, multimeter readings said it was OK, but BMS said was bad. For you both show the same.
 
When you top balanced, all in parallel did you connect the charge wires to one end of the pack?
Did you label your cells and take voltage readings at a couple intervals after your top balance?

We’re the cells all the same voltage before you built the 16S pack?
 
For the top balance, I connected all the cells in parallel and in a line. Imagine cell 1 being leftmost and cell 16 being rightmost.

I then connected to the PSU the +ve from the leftmost cell (cell 1) and the -ve from the rightmost cell (cell 16). I'd heard that this was better than connecting to say both the +ve and -ve of cell 1?

However I didn't take voltage readings once the PSU had ramped down to practically zero current. Wish I had now of course.

On the plus side the Victron is working now, it was a simple loose connection.
 
However I didn't take voltage readings once the PSU had ramped down to practically zero current. Wish I had now of course.
It’s never too late to label your cells and take notes. I like using letters (A, B, …) because the cell positions are already numbers. And, the negative end of a battery is cell 1. It’s odd but getting your notes right at the start will be helpful next time you need to sleuth your cells.
 
For the current we are talking about hooking up a small 10 or 20A supply, where you connect up the charge leads is pretty much meaningless. Once you get up to the 3.65V all the voltages will be identical since all the currents will be identical...ZERO.

I put my charge leads on one side...charged up to 3.65V...took the cells off and measured them with a capacity measurement device...put them back in parallel and top balanced again. All of my 10 280AH Eve cells measured 280AH except for one that measured 279AH.
 
To the OP...did you confirm the voltage measurement with your own good meter across each cell?
 
Yes, I confirmed the BMS cell readings by using a good quality multimeter and they were accurate.

However, as soon as I got the Victron hooked up and working, then within about 2 minutes of applying a decent load (around 1.5kw) then the cells all equalised. Since then they've all been almost exactly the same voltage.

I'm guessing that any balancing issue could rear its head again once I come to charge the pack, so will run it down for a bit and then experiment with charging and see if any of the cells cut off due to overvoltage.

On thing... is 3.65V * 16 (58.4V) an appropriate voltage to set for the Victron to bulk charge? Since the BMS is set I think to disconnect if any individual cell exceeds 3.65V, then I'm wondering if 58.4V would be too high, as surely one cell is likely to exceed 3.65V when charging?
 
Yes, I confirmed the BMS cell readings by using a good quality multimeter and they were accurate.

However, as soon as I got the Victron hooked up and working, then within about 2 minutes of applying a decent load (around 1.5kw) then the cells all equalised. Since then they've all been almost exactly the same voltage.

I'm guessing that any balancing issue could rear its head again once I come to charge the pack, so will run it down for a bit and then experiment with charging and see if any of the cells cut off due to overvoltage.

On thing... is 3.65V * 16 (58.4V) an appropriate voltage to set for the Victron to bulk charge? Since the BMS is set I think to disconnect if any individual cell exceeds 3.65V, then I'm wondering if 58.4V would be too high, as surely one cell is likely to exceed 3.65V when charging?
I think 3.65 per cell charge is too high and could get things like you mention. I charge at 3.475 * number of cells and have my cell over voltage is 3.54.
 
Take your balance lead off of cell 1 and reassemble the cables/wires, making sure the balance lead is the topmost ring terminal on the stack. From your picture, it looks like the balance lead is between two of the big cable lugs.

Reorient all your balance leads so that the insulated part of the ring terminal is hanging off the side of the bus bar, not on top of the bus bar.

On thing... is 3.65V * 16 (58.4V) an appropriate voltage to set for the Victron to bulk charge? Since the BMS is set I think to disconnect if any individual cell exceeds 3.65V, then I'm wondering if 58.4V would be too high, as surely one cell is likely to exceed 3.65V when charging?

As @chrisski said, 3.65 volts per cell is too high on an ongoing basis. The Victron default for LiFePO4 is 3.55v as I recall.
 
It’s never too late to label your cells and take notes. I like using letters (A, B, …) because the cell positions are already numbers. And, the negative end of a battery is cell 1. It’s odd but getting your notes right at the start will be helpful next time you need to sleuth your cells.
Thankyou that's very useful advice!
 
Yes, I confirmed the BMS cell readings by using a good quality multimeter and they were accurate.

However, as soon as I got the Victron hooked up and working, then within about 2 minutes of applying a decent load (around 1.5kw) then the cells all equalised. Since then they've all been almost exactly the same voltage.

I'm guessing that any balancing issue could rear its head again once I come to charge the pack, so will run it down for a bit and then experiment with charging and see if any of the cells cut off due to overvoltage.

On thing... is 3.65V * 16 (58.4V) an appropriate voltage to set for the Victron to bulk charge? Since the BMS is set I think to disconnect if any individual cell exceeds 3.65V, then I'm wondering if 58.4V would be too high, as surely one cell is likely to exceed 3.65V when charging?
This is what has been happening to me - under load all cells equalise fairly quickly with the voltages immediately reading somewhat differently to what I see when charging/balancing. However as soon as I charge I get exactly what you suspect would happen above. Like you I also thought that 3.65 V was the maximum charge for a 3.2V LiFePo4? as all chargers sold for my setup 8S 24V are always 29.2V. Currently working through all the awesome advice provided in this thread and will post those outcome when I'm done - thanks everybody!
 
Back
Top