diy solar

diy solar

Weak Cell?

DiploStrat

New Member
Joined
Sep 8, 2020
Messages
12
Built three 280Ah batteries with EVE cells and Overkill Solar BMS. to mount in my overland camper. Charging is from solar, alternator, and Magnum inverter/charger.

Two batteries have been textbook perfect in about two months of use. This includes induction cooktop, a ton of fans, and overnight air conditioner runs. Cells remain balanced within 0.00x volts, except at the very top.

The third battery has one cell, number four on the BMS readout, that is almost always low. Here the delta between cells is typically about 0.05 volts. Battery voltage tracks the other two, but the BMS generally rates it at about 2-3% lower SOC.

Under load it appears to drop a bit slower. One night, when it was turned off, it appeared to drop its voltage much faster than the others.

All 14 (was advised to order extras) cells were originally top balanced in parallel.

I have double checked the BMS leads several times, and cleaned and reconnected the BMS leads.

Soooo, I propose the following:

-- Remove the offending batteries and top balance along with the two spare cells.

-- Reassemble with the orignal cells, changing the order.

If all is well, have a Merlot. If not,

-- Remove again and remove offending cell. Reassemble the battery with one of the spares.

-- Top Balance again and reinstall.

Questions:

-- Any comments?

-- With Overkill Solar BMS, which is cell 4? Closest to + or closest to - ?

Many thanks!
 
The first cell is the cell the minus wire connects to and the 4th cell is the cell the + connects to.

You seem to have a valid plan .... but I would include carefully cleaning and inspecting the bus bars and apply a little noalox or similar when I re-assembled.
Make sure they are clean and perfectly flat with no burrs. I would also clean and burnish the terminal pads on the cells.
De-greasing with some alcohol or acetone would also be a good thing to do right before applying the noalox.
 
-- Remove the offending batteries and top balance along with the two spare cells.

-- Reassemble with the orignal cells, changing the order.
I started with 13 cells so was in the same position you are. I've done a mixture of changing the order and swapping spare cells in and out.
I eventually figured out which 8 were the strongest and assembled that into a 2P4S battery. I took the next strongest 4 and built a 4S battery.

Both your suggestions are good but with 2 spare cells and one suspect cell, i think i would opt for substitution. I would also mix up the cells so the new one did not fall on one of the ends. Cell 1 (negative end) is usually where the weird chit happens. Others have seen things on cell 4 too - the last one i recall was from sitting for months with a BMS being powered (possibly from cell 4).

With that many cells and batteries, you should really consider using letters for your cell labels!
 
The first cell is the cell the minus wire connects to and the 4th cell is the cell the + connects to.

You seem to have a valid plan .... but I would include carefully cleaning and inspecting the bus bars and apply a little noalox or similar when I re-assembled.
Make sure they are clean and perfectly flat with no burrs. I would also clean and burnish the terminal pads on the cells.
De-greasing with some alcohol or acetone would also be a good thing to do right before applying the noalox.

Did. Will redo.
 
With that many cells and batteries, you should really consider using letters for your cell labels!

Pedantic question. Letters vs. number? The BMS uses numbers. Might not batteries on letters and cells on the same numbers as the BMS work as well?

Not trying to argue, but to understand.
 
Letters vs. number?
The battery cell positions are numbers. The negative end starts with cell 1.
If i number my cells and move them around, my notes might look like this.

Cell 1 in cell 3 has 3.4V. Moved cell 3 to cell 5 and settles 3.5V

If i label my cells as letters:
Cell A in cell 4 has 3.4V. Moved cell C to cell 5 and settles 3.5V.

Its REALLY easy to write undecipherable notes with numbers representing both cell label and cell position.
 
The battery cell positions are numbers. The negative end starts with cell 1.
If i number my cells and move them around, my notes might look like this.

Cell 1 in cell 3 has 3.4V. Moved cell 3 to cell 5 and settles 3.5V

If i label my cells as letters:
Cell A in cell 4 has 3.4V. Moved cell C to cell 5 and settles 3.5V.

Its REALLY easy to write undecipherable notes with numbers representing both cell label and cell position.
Got it - you want a fixed identification of each physical cell, vs. it's functional position.

OK, so the batteries are , "George", "Sally", "Algenon", etc. ;)
 
Some images. Batteries 1 and 2 look like this:

Screenshot 2021-08-12 at 13.54.29.jpg



Battery 2 looks like this:

Screenshot 2021-08-12 at 13.53.34.jpg

The delta can be over 0.2v. I would guess that that would mean that the low cell is pulling the others down, no?
 
UPDATE: Went out and measured the voltages with a multi-meter. Each was spot on 3.33v. The BMS gave cell 4 as 3.239v.

So I removed the balance leads and tested each wire. Resistance was 0.0. Testing required running a paper clip into the plastic connector so that I could get access to that end of the wire. Then, for good measure, I gently re-crimped every ring fitting at the battery end and re-tested the resistance.

That done, I reassembled the leads to the battery and the BMS. After 24 hours, that is a day of solar charge and a night, the cells balance perfectly.

So I am prepared to say that there was some resistance in the BMS leads. Should have used DeOxit on the paper clip.

Also a mild shout out to Amy Wan, from whom I bought the batteries. She sells decent stuff!
:)
 
Back
Top