VagueDirector
Battery Addict
- Joined
- Jul 20, 2020
- Messages
- 10
I recently noticed the cell voltages of my 2.3kWh LiFePO4 system were unexpectedly deviating. After running my computer setup with it (~25A@12V) for a few hours, one of the cells was much lower than the others. This is odd as I know they have closely matched capacity. So I opened the case and went probing around with the multimeter. (The cells are arranged as 2P4S for reference.) That's when I noticed that the first parallel pair was at completely different voltages/SOC, one cell was at ~3.15V while the other was at ~3.25V. When I probed a cell terminal bolt and its attached busbar saw a 70mV difference! These bits are directly bolted together, there should be no voltage across them at all when the system is at zero load.
What happened is when I built the system ~6 months ago I didn't sand the aluminium terminals to remove the oxide layer. I was fixated on issues such as cell balancing, cell compression and terminal bolt over torque that I completely overlooked this in the excitement of building a new system. I assumed the cell voltage discrepancy was normal and down to manufacturing tolerances.
To fix it I completely redid the connections by sanding the terminal posts and busbars with fine grit sandpaper, and then immediately applying dielectric grease to prevent oxidation. After re-assembly the cell voltages as read on my well calibrated BMS are now always within 5mV of each-other. There is also <0.5mV between the terminal bolts and the busbars. I've also noticed the batteries now run a little cooler under moderate load, as they are no longer sinking heat from poor connections.
Some images of the system for context:
What happened is when I built the system ~6 months ago I didn't sand the aluminium terminals to remove the oxide layer. I was fixated on issues such as cell balancing, cell compression and terminal bolt over torque that I completely overlooked this in the excitement of building a new system. I assumed the cell voltage discrepancy was normal and down to manufacturing tolerances.
To fix it I completely redid the connections by sanding the terminal posts and busbars with fine grit sandpaper, and then immediately applying dielectric grease to prevent oxidation. After re-assembly the cell voltages as read on my well calibrated BMS are now always within 5mV of each-other. There is also <0.5mV between the terminal bolts and the busbars. I've also noticed the batteries now run a little cooler under moderate load, as they are no longer sinking heat from poor connections.
Some images of the system for context:
Last edited: