Balancing question

H2Guy · Jul 4, 2021

I’ve built 2 separate 7s10p packs with 18650’s for my golf scooter project (see Golf E-Caddy thread) and although I selected cells and balanced each P with a consistent group capacity, when I use them they both exhibit the same performance. That is, groups 1&2 deplete faster then the other 5. Both packs it’s the same … group 1&2 drop at least 0.3V below the other groups at 50% SOC and I end up having to top balance the packs after charging. The cells were sourced from the same place and the BMSs are different but nothing fancy.

Anyone have any thoughts or should I even be concerned?

fatty · Jul 4, 2021

Could be the BMS. Bypass it and test.

john61ct · Jul 5, 2021

check wiring connections

Sure you don't have a load pulling from just part of the pack?

stan.distortion · Jul 5, 2021

Do you have a pic showing how they're connected? Somewhere there's a really good thread on battery building, well worth a read as there are lots of potential pitfalls that aren't really obvious until explained and many can be easily avoided.

At a guess, you're pulling power from one end of a parallel connection, the end with the low voltage? Just the resistance along a parallel connection can cause a load imbalance, even with damn big strips, more power is pulled out of the cells nearest the lead. I'll dig that thread out and edit this post as I need to re-read it anyway, iirc pulling power off the parallel connection with multiple equal length cables spaced along it's length and joining them together to a single output point is one way around that issue.

edit
https://endless-sphere.com/forums/viewtopic.php?t=84412

H2Guy · Jul 5, 2021

I don’t have any external connections to any part of the pack other than the main leads so no obvious explanation for why bank 1&2 draws down faster on both packs. Since each pack is using a different brand of BMS it is still strange why they behave the same ad unlikely to be the BMS … probably rooted in my build design.

I should have taken pictures of the build before I shrink wrapped it but in my cheesy hand drawn version below each parallel bank is connected with 8mmx0.15mm nickel strips spot welded and the series connections are 9 nickel strips (same size) spot welded to the parallel strips between the parallel groups. The main connections are 8awg stranded copper soldered to the entire length of the -ve and +ve rails (obviously on the opposite side of the pack than the series connections). Hope the diagram helps …

fatty · Jul 5, 2021

stan.distortion said:
Do you have a pic showing how they're connected? Somewhere there's a really good thread on battery building, well worth a read as there are lots of potential pitfalls that aren't really obvious until explained and many can be easily avoided.

At a guess, you're pulling power from one end of a parallel connection, the end with the low voltage? Just the resistance along a parallel connection can cause a load imbalance, even with damn big strips, more power is pulled out of the cells nearest the lead. I'll dig that thread out and edit this post as I need to re-read it anyway, iirc pulling power off the parallel connection with multiple equal length cables spaced along it's length and joining them together to a single output point is one way around that issue.

edit
https://endless-sphere.com/forums/viewtopic.php?t=84412

This affects current draw within the parallel group, and would not result in imbalance between series strings

fatty · Jul 5, 2021

H2Guy said:
Since each pack is using a different brand of BMS it is still strange why they behave the same ad unlikely to be the BMS

Many (if not most) BMSs are known to draw their own operating power from the first series string or two. This makes sense, since it's cheaper and easier than including a DC-DC converter across the whole pack.

Run a search through this subforum for examples of packs with dead string #1 and BMS.

H2Guy · Jul 6, 2021

The BMS drawing from banks 1&2 makes perfect sense now that you’ve pointed it out. I’ll do a little research to confirm. Thx!

Balancing question

H2Guy

100 mW

fatty

10 kW

john61ct

1 TW

stan.distortion

1 kW

H2Guy

100 mW

fatty

10 kW

fatty

10 kW

H2Guy

100 mW

Similar threads