After I put the Headway pack in my Vectrix, I discovered (2) of the cell junctions were getting voltage to the frame through their balance wires - I think the insulation got chafed/cut/pinched while I was sliding in the bottom half of the pack. I would have to dissassemble the pack to replace/re-route the wires (ah, the tradeoffs we make to shoehorn cells into a space

) so I just blew the fuse to those two cells, and now they don't have balance wires anymore.
I've been running this battery pack for a year and a half now, bulk charging daily and balancing it every 3 months. Of course, the cells on either side of the blown balance wire have been getting charged as a pair in series, instead of 2 individual cells, as the two single cell chargers loop around the blown balance wire. If the two different cells have different rates of self-discharge, it is possible for one cell to have a higher voltage than the other. In an extreme imbalance, one cell could be charged to a voltage that could be a fire risk.
Getting to these cells seemed impossible, so I have ignored the issue, and bulk-charged to a lower voltage. Recently I decided I have been running the pack long enough that I needed to become very concerned about those risks, and I stopped riding/charging the bike until I could address it.
I realized that if I took the body off and did some careful measuring, I could drill holes in the frame/battery box at the point in the pack with a bad balance wire, and insert a temporary wire for balancing. I did not relish the idea of drilling holes, and taking off the bodywork was an onerous task - one basically needs to remove the whole back end of the bodywork to remove the parts on the side. I waited until my next vacation, and budgeted several days to get through everything.
Armed with my Vectric Service manual, I decided to see if I could remove the bolts holding the parts on the side, and pull the parts out far enough to access the frame, without removing all the bodywork. That not only worked, but I was reminded that while the bottom of the battery box is solid metal, the upper part is a framework skinned in plastic. When I went to locate the junction between cells 13 and 14 with a bad balance wire, I found it right where the plastic skin meets the solid metal box. I was able to easily jam a solid-copper wire into the junction as a temporary balance wire.

- 13-14.jpg (33.43 KiB) Viewed 389 times
Then I went looking for the junction of cells 20-21, right at the back of the pack. I pulled back the corner of the plastic cover, and voila, I was right where I needed to be - so I attached the yellow alligator clip lead you see in the picture as a temporary balance wire.

- 20-21.jpg (65.56 KiB) Viewed 389 times
When I built this pack, I chose to sort the cells by resting voltage, instead of IR. I charged all the cells up at the same time so they were at the same voltage, let them sit a few days, then I sorted them in order of resting voltage. I figured that the cells with the highest self-discharge (represented by lowest voltage) would be the ones more likely to fail, so I put them at the top of the pack (towards cell 45) where I would have the easiest access to them. The cells with the best resting voltage went to the bottom of the pack (towards cell 1). As a side benefit, that means each supercell should have a similar rate of discharge to the supercell next to it.
I did a balance charge of the entire pack before I started, I wanted to see how high the voltage got on the stronger cell of the pair. I think sorting cells by self-discharge rate helped a lot, because when I went tested cells 13 and 14, they were within 0.012v of each other - already balanced, IMO. With cells 20-21, I found a much larger delta, but it took only about an hour of 2a charge (on a 60ah cell) to bring the weaker cell into balance. Considering that the pack has paralled (5) cells of mid-grade LiFe getting hammered daily for a year and a half, I think that is pretty good. When I was done, I just pushed the plastic back into place, with a layer of duct tape to seal it. The bodywork slid back into place easily, and now that I know where the cell junctions are, I won't need to remove it for my next balance charge.
Instead of days of work, the solution took mere hours. The blown balance wires have been a source of angst for me since I built the pack, so I am hugely relieved to have found a simple way to resolve the issue. I am also pleased to see how well my sort-by-resting voltage approach worked over a long period of time.
-JD