Recharged the pack overnight, and we see that the pack not only had more capacity, but the sticks tracked each other much better on the discharge, until the same stick 3 dropped out. The reversal of the cells during the really deep discharge does not seem to do any damage (need more cell level analysis to be sure), and the sticks seem to be much better matched.
If you have a pack that has not responded well to the standard charger settings, it may be worth giving this a try?
Procedure:
1. enter password(call or e-mail me to get it)
2. change the MinDischV* to a voltage below 1V/cell.
3. set mode to 6
4. option bit 7 set to no precharge
5. with discharger connected, press start.
6 when the discharger turns on, press the test key
the message SlopeStopDisable now turned ON will be displayed, and the discharge will proceed to where it would have stopped, it will start beeping at you, but not stop until the pack is fully discharged to 10V, then it will recharge. The SlopeStopDisable will remain active, and since you cannot turn it off while recharging,you will not be able to turn it off until the next discharge cycle starts.
You could let it deep discharge for 3 cycles, turn the SlopeStopDisable off on the next discharge cycle, or change the number of cycles to 1, so you only get one cycle.
At this point, this procedure is experimental, and we have no idea if the reverse charging does any long term good or harm, but if your ready to give up on the reconditioning anyways, this may be a way t6o at least get all the cells to see a deep discharge.
I will follow this test with some more detailed cell level analysis very soon.