At this point, I experimented with some Lithium Polymer batteries that I use for radio control helicopters. The packs were 5s 4000mah packs and I connected three in series to create a 15s 4000mah pack. The power was really incredible but at only 4000mah, I just did not have enough range. They are also very expensive and pretty delicate as well. I did not see these packs surviving every day use on the bike.
I finally started researching A123 cells and found quite a few people using them on bicycles. I had never really considered them because for radio control they are still too heavy for the models I fly. However, for bicycles they looked like a great cell.
I had a starting point from my experience with the SLA batteries. I knew that I used about 40Wh per mile and I needed a range of at least eight miles. So, I figured I needed a pack that would be rated at 320Wh * 1.25 (for a margin of safety) = 400Wh. A123’s are commonly quoted as 6.9Wh cells which meant I would need about 58 cells and A123 packs conveniently have ten cells so I bought six packs from Ebay for a total of sixty cells. Getting the cells out of the packs is not terribly difficult but it is time consuming.
I initially decided to create a 15s4p pack out of the cells. This is where I tried something a little different. I wanted to be able to get to each individual cell for charging or monitoring and I liked the idea of having somewhat modular packs to create a variety of configurations. I finally decided to create 4 cell flat packs with a six inch piece of 16 gauge Dean’s Ultra wire on each terminal with an APP connector at the end. These cells are not connected to each other electrically in any way, they are just connected mechanically to each other reinforced packing tape and heavy duty rubberized shrink wrap.
I was very concerned about how much voltage I would lose because of all the connectors and 16 gauge wire, so I did an analysis of the power loss due to the resistance of the wire and connectors. If I limit the current to 40A, each 15s pack will only see 10 amps. At this low current level the cells hold their voltage very well, actually above 3 volts per cell. I calculated a value of 3.061V per cell using 10A and an internal resistance of .0089 Ohms, which yields a pack voltage of 45.9V under the 10A load. When I figured the voltage drop from the wire and connectors, I calculated an additional drop of .73V for a pack voltage of 45.2V. This seemed perfectly acceptable to me, so I went on with the pack construction.
I made fifteen of these packs and then connected the first cell from all packs together in series, second cell from all packs together in series, third cell from all packs in series, and the last cell from all packs together in series. This gave me four 15s packs. I then made a couple of harnesses that could plug into these four packs using 16 gauge wire and plug into a large 50A APP connector. All of this fit very neatly into a Topeak MTX bag and weighs about 14 lbs.
