I have now covered the windings and lam edges with epoxy to hopefully solve the short issue. In the process of removing the nut to take motor covers off, I rubbed off some of the insulation on the sensor harness, which I believe shorted one hall. So I have one replacement hall on the way, cause I can't find my spares... If you use magnet wire for hall sensor wires, make sure there isn't a huge jut in the splice between the 12 gauge deans (in my case) and the cheaper wire from axle to motor. got to be able to clear the axle and washer over the joint.
Also just received 35 more A123 cells for 20s5p or 21s5p. I would like to hear your advice in regards to whether or not I should cycle the new cells twenty times before supplementing my current pack which has minor use. I suppose cycling them is the best option, though only ~.22% of their expected 10-12% life diminishment over 1000 cycles has been achieved. (11-15% decrease over 1000 cycles multiplied by 20/1000). I don't know the realistic cell-gap that will develop, and I rather not have to balance them all the time cause I only have 14S with my hyperion..
Adding more cells will require moving the controller out of the enclosure. I haven't decided whether it will be mounted on the downtube or rear rack, and have not figured out how to weather proof the controller.