Charging groups in pack with MC3000 smoking

[Obsequious]

Hey there everyone, longtime lurker, first time poster. I'm in the process of rebalancing a rather classic, badly designed Geoby e-bike battery (Samsung cells have GK2T printed on them, so likely 2016). One of the packs was broken off (nickel strip torn) during cycling (this pack belongs to my 80 old neighbor, you know how these seniors are ), and for some reason the pack now has serious imbalance. So much that the charger BMS can't handle the balancing, and started smoking when I tried (passive balancer, way too much current through the resistors I guess). The two broken off groups are at 3.9V, all the others around 2.9V. So I thought: let's rebalance it 'manually', group per group.
I thought I had done my homework, and ordered a SkyRC MC3000 for this, then created a few leads with nickel strips glued onto fake batteries to charge each of the 10 paralleled groups in this 10S3P pack, like this:



This worked well for one cell group, nothing wrong, voltage steadily went up. I kept the amps low and my eyes on the cell temperature with IR camera. So I thought: let's add another group with another fake battery + leads. Very soon after I started charging on slot #4 on the MC3000, the dreaded "Warning MCP3424- 1 Err" came up on the charger's display, and smoke came out of it. I immediately shut everything off, and to my surprise, all slots seem to be working well.

After doing some more homework, it seems that what I'm doing is okay, in theory. I think it's the MC3000 that freaks out in this setup. It looks like it's shorting somehow. I've put another Li-Ion cell charger next to it now, and they work perfectly in tandem.

So, was I wrong to believe that the MC3000 had 4 fully independent charging slots? Can anyone guess what actually happened?

 

[fechter]

My guess is the charger doesn't like having the cells connected to each other. Possibly the negative charger outputs are all connected, so this creates a short when you attach series connected cells. You can use an ohmmeter to check if the slot negatives (or positives) are connected together.

 

[Obsequious]

You can use an ohmmeter to check if the slot negatives (or positives) are connected together.

Yes, it seems the positive sides of the MC3000 are indeed connected, at least when I do a continuity check. I wonder why.

 

[fechter]

They have a common source. It would be much more expensive to have all the outputs isolated. You can do multiple cells as long as they aren’t connected to each other.

 

[Obsequious]

You can do multiple cells as long as they aren’t connected to each other.

Too bad, I had hoped I could charge four packs at a time with the MC3000. But it's not a big deal, it just took more time to balance the pack. Which was finally ready today, and is now fully functional again! Thanks for the info