A123 cells charging super fast/charge on LiPo setting?

[drutledge]

I bought 15 A123 cells via the Midwest group buy from Victpower and have assembled them into a 15S1P pack which can be seen here:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=38761&start=30

When put on a 600W charger at 54V (11A, about 1/2C rate) three of the cells really take off from the group. Like a +150mv gap within a few minutes. They will slowly come back down to the rest after much time but I have charged from 3.45V or so and watched these cells quickly hit +3.65V before I pulled the plug. If I charge at an even slower rate on my Hyperion 1420i (250W max output), they exceed the group but not nearly as much. (Picture below FWIW)

I'm thinking this is either the cells or something with the pack design, perhaps residue left over from the stickers on the tabs though I think I cleaned them fairly well. My plan is to test each cell individually and compare their charge curves.

Q1: I can't use my bulk charger to charge individual cells, the Hyperion only allows for a 3.3V charge on it's LiFePo4 setting. (I don't know why) I was going to use the LiPo setting and set it to stop charging at 3.6V not 3.7V; since they are both CC/CV. Any problems with this?

Q2: Assuming the cells still exhibit this behavior, is it normal? Can they be expected to even out when 'broken in' by cycling a few times? I only have experience with Lipo (and SLA) and while some cells are more "energetic", less resistance (?), I've never seem anything like this. Maybe a 20-25mv difference while charging a 12S2P pack.

I posed this question in the middle on my build post but it really seems better on it's own. Sorry for the double duty.

 

[DAND214]

Use the LiFePo setting. it should peak at about 3.65. 3.3 in nominal setting I think.

Dan

 

[heathyoung]

If they are at a higher SOC than the others they will charge faster than them (ie almost full already).

Easiest way to equalize the SOC on these large cells is to measure their terminal voltage - if they are very close, put them all in parallel for a week or so (I did this with 44 thundersky 40AH cells - too lazy to wait to charge individually.

 

[drutledge]

Use the LiFePo setting. it should peak at about 3.65. 3.3 in nominal setting I think.

Dan

Ok I must have been doing something wrong, I got a cell to go to 3.6V on LiFe setting, thanks.

If they are at a higher SOC than the others they will charge faster than them (ie almost full already).

Just so I'm sure/clear what you're saying, LiFePo cells have a very flat curve and that even though they all show within a 5mv of each other, some might be at a much higher SOC (more charged) and that LifePo voltage really takes off towards the end of charging (something I saw last night charging a single cell). That would seem to be a good explanation. It is very easy to break apart and reassemble my pack so I think I'll try to parallel them for a few days and see if there is a noticeable difference in their charging behavior.

I would have hoped that these Victpower A123s would have come at the same SOC; makes me wonder about their quality and where they've been...