Active balancer.... anyone want to speculate how this work?

[paulca]

[mikerj]

Looks like a buck/boost balancer.  AFAIK these simple non-coupled inductor designs can only transfer charge between adjacent cells, so on anything with more than two cells you can't transfer charge between the end two cells.

[SparkyFX]

Underside of the PCB would be interesting, should there be anything.

I suspect it just makes use of an inverting boost converter, which could enable charging the lower cell and might be able to switch to non-inverting for reverse operation.

[paulca]

Looks like a buck/boost balancer.  AFAIK these simple non-coupled inductor designs can only transfer charge between adjacent cells, so on anything with more than two cells you can't transfer charge between the end two cells.

That's the thing.  It "seems" to be working on a 3 series and... bidirectionally too.

[mikerj]

Looks like a buck/boost balancer.  AFAIK these simple non-coupled inductor designs can only transfer charge between adjacent cells, so on anything with more than two cells you can't transfer charge between the end two cells.

That's the thing.  It "seems" to be working on a 3 series and... bidirectionally too.

Yes it still works on a 3S, but the limitation on transferring energy between adjacent cells means it may have to transfer the energy multiple times across cells which drops efficiency but still a lot better than the simple resistor load style balancers.

Lots of info on this and other topologies if you google "buck boost cell balancer"