In May 2004, we received 95 of the 70 amp-hr batteries. We decide that our first priority is to check the consistency of the quality of the batteries.

This pre-use test cycle consists of charging the cell to full charge, then discharging the cell to 3.0 volts while recording amp/hrs, then recharging to full charge. Each cell will be charge and discharged at 1C (70 amps). Since the cells arrive less then half charged, each cell cycle takes about 3.5 hours and about 120 amp/hrs from our battery bank. We are time limited to testing 3 cells a day. With 95 cells, this is a major time commitment but could be crucial in our long term understanding of the batteries.

In testing, every cell delivers over 70 amp-hrs at 1 C. The maximum we see is 78 amp-hrs, the minimum is 71 amp-hrs. Watt-hrs are around 275-290.

Tracking amp-hrs recharging was a little funny. We seemed to be putting back less amp-hrs than we put in. Were we in violation of some law of physics so soon?

Sadly, no. When we compared watt-hours {(amps*volts)/time} in and out, they matched. Measuring amp-hours ignores that the battery voltage curves are different for discharge and charge.

We are charging at Kokam's maximum suggest rate-1C using their recommended method of charge: constant current (CC) until 4.2 volts is reached then constant voltage (CV) until current drops below .1 C. This charge method takes about an hour and fifteen minutes. In an EV application, it would be difficult to have a power source that would supply enough power to charge this fast. A standard 15 amp 110 volt outlet would be hard pressed to charge a 24 volt pack at 70 amps much less a 350 volt pack. Charging at a slower rate's only known problem would be that it would have taken us a lot longer to test all our cells.

We do notice some inconsistent behavior when charging at our high rate of charge. Most of the cells when charging at 70 amp constant current accept around 250 watt-hrs before their voltage reaches 4.2, then another 30 watt-hrs or so at constant voltage. This was not true for all the cells. About 7 cells reached 4.2 volts early. The most aberrant cell reached 4.2 volts after only 169 watt-hrs and then spent over an hour at constant voltage charging for the remaining 103 watt-hrs.

This means that without individual cell management under high constant current charge, this cell might go into over voltage well before the other cells.

With a battery management system (BMS), charging in series might be substantially slowed for a pack as the charger will be forced to lower it's constant current rate to accommodate the aberrant cell(s).

It is unclear at this point whether the cells that exhibit this behavior are defective or if this behavior is a characteristic of these cells. Testing time was limited by the team's goal of getting the car on the track but we did manage to do some cycle testing of our original test battery. Keeping in mind that this is an abused battery which has been twice discharged at twice the recommended rate, so any conclusions would be premature.

We did find that in constant cycles, the point where the battery hit 4.2 was pretty steady, but when it did vary it was by a large amount. From the expected 54 amp-hrs to a low of 34 amp-hrs. There also seems to be a very slight increase in amp-hr capacity as the batteries warm but that does not seem to effect the charge cycle.

We decide to run 88 cells. This gives us a nominal voltage of 325.6 volts and a full charge voltage of 369.6 volts.