Pack slowly discharging

Alec_J

Alec_J

Member
Joined
Mar 16, 2017
Messages
89
I balanced my 5 packs of 20p cells to 4.02volts by paralleling them together. All of my cells are new from new replacement laptop batteries and hover board packs. A couple days later one of the packs was at 3.96volts. I thought it was odd at the time but didn't think much more. Yesterday I paralleled the 5 packs again for 14 hours. They all balanced at 4.01volts. This morning I checked and the offending pack was at 3.98 and the other 4 packs were still at 4.01.

So my questions are:

1. is this possibly just voltage sag on this one pack? I am doubtful but thought it worth asking.
2. How do I figure out which cell/cells in the pack are causing the problem?

All ideas welcome

Thanks,
Alec
 
The only way to check for leaky voltages in a parallel pack is disconnect the cells. This one of the main reasons I'm not going to solder/weld my packs.

What I would do, is re-balance the pack. Then , disconnect rows of cells (so, if it's a 10 x 6, have 10 rows of 6, or 6 rows of 10). Then let it sit for a cpl days checking it periodically. This will help narrow down the offending area. Then repeat by cutting the row in half that is now lower. Keep doing that till you find the problem child.

Another 'possible' way to find the cell, is to do a high draw on that pack alone and watch it with a thermal camera/probe. In theory, the problem cell will start to get hotter than the other cells. Lay the pack flat, and probably put in an enclosure to help keep heat dissipation balanced across the pack (if the bad cell is on the outside, it may loose the heat so fast to air you may not catch it). This, of course, is just theory.
 
Best way is to totally disassemble all the packs and charge them all up and let them rest. I only keep.cells that stay within 0.05 of charge after 2 weeks
 
must be at least one cell sucking the power out of the others and bringing the total voltage down. I also had cells that don't hold charge, that's why I measure the voltage of all cells twice. First after they were charged full, then again after at least 3 weeks. If they don't loose more than 0.1 volts, they go into my packs.

In order to find the bad guy, just balance your packs again, then disassemble the offending pack. Measure the voltageof each cell after a while and I'm sure you'll find it...
 
Pick pack apart and leave for another 2-4 weeks and you have your answer on what cell/cells that are the issue.
 
I think I may have found the bad cell upon further inspection.

image_sbnhwy.jpg

Looks like when I soldered it it burned through the side a bit. What do you think?

@Korishan how do you plan to connect your cells? You may have addressed this elsewhere. Just curious.
 
Yep. You made a leak in the negative electrode. I had one Sanyo battery that had that kind of defect out of the pack, and it was discharging quite quickly, and started bubbling. I would get it out of the pack as quickly as possible.
When charging, since it is a small leak compared to mine(about half the size), the solvent inside will release very slowly, discharging the cell, and dragging down the pack with it, since when parallel, cells balance charge themselves, and you constantly lose voltage because of this if a cell is faulty.
 
I had this happen in one of my 120p packs. It slowly bled down to 3.99v after sitting for a few months. Doesn't sound bad, but all other 13 packs were still at 4.15v. I ended up removing the positive busbar and found a 0v cell that was slowly burning off some current. The cell wasn't even warm or anything.
 
A thermal camera might pick up the heat in the fuse wire, but just barely.

All of the other cells are giving off a little current, That one cell is drawing 19 times more current than the others are giving off.
 
Disconnect that one cell and see what happens on another charge up. If it holds, you found your culprit! Nice find! :) A lot faster than just tearing the whole pack down first.
 
BlueSwordM said:
Yep. You made a leak in the negative electrode. I had one Sanyo battery that had that kind of defect out of the pack, and it was discharging quite quickly, and started bubbling. I would get it out of the pack as quickly as possible.
When charging, since it is a small leak compared to mine(about half the size), the solvent inside will release very slowly, discharging the cell, and dragging down the pack with it, since when parallel, cells balance charge themselves, and you constantly lose voltage because of this if a cell is faulty.
Click to expand...


Yep I cut its fuse wires to isolate it from the rest. I will retest the pack voltage in a day or so. My guess is it will be fine now. I will pull it and replace it. I wish I had realized the error earlier. Oh well I will know now moving forward to be conscious of that sort of fault. I love learning this stuff even if it means I have to tear the pack apart... [emoji19]

egam said:
A thermal camera might pick up the heat in the fuse wire, but just barely.

All of the other cells are giving off a little current, That one cell is drawing 19 times more current than the others are giving off.
Click to expand...


Great points. I wish I had a thermal camera. Anyone know of an inexpensive model?

Korishan said:
Disconnect that one cell and see what happens on another charge up. If it holds, you found your culprit! Nice find! :) A lot faster than just tearing the whole pack down first.
Click to expand...


I clipped it's fuse wires and will monitor it over the next week or so.

Thank you @Korishan for all of your help and posts! I truly appreciate your freely given knowledge. This is truly the best form I've ever been part of.
 
Learning from each other mistakes makes all of us safer, smarter, and more intelligent. We all move forward as an enlightened people :)
We all now know (at least those who actually read this post) that to watch out for burn marks when soldering cells and how to narrow down a leaky situation.

Glad you found it before you started ripping it apart. That would of sucked. Especially if that would of been the last cell you took out.
 
You must log in or register to reply here.
Back
Top