Battery Pack Testing

[retepsnikrep]

I have been playing about recently with an idea to develop a semi automated pack testing system. I have a spare battery pack with some weak cells/sticks but I don't know which yet.

I have a few things assembled on my bench and am developing some Qbasic software to control it.

Basically it comprises an old PC power supply with a Velleman K8000 PC interface kit and a 300w dummy load.

I have partly dissasembled my pack and removed one side of the cell connectors to reveal all the cell sticks. I can now connect fairly heavy current cables to each stick in turn in the pack and let my tester do it's business.

Basically it looks at the stick voltage and then charges it up at say 1C to max V (Any thoughts on what Max V should be for one stick of six cells?)

It then allows a rest (how long?) then discharges stick at 2C (is that enough?) through a dummy load until min V (Any idea on minimum V?) then another rest, and charge again, cycle will repeat as many times as you program. Voltage, current and temp are monitored at all stages (10 second intervals) and recorded to a text file for later comparison in excel or similar.

Temp is Monitored using the oem attached sensing strip, this increases in resistance as cell temp increases. I have to calibrate my software to match this.

So quite a bit of work to do but once done I could lend this device to members as it only needs a pc/laptop with a parallel port to run the software and control it. You could test your own failed/salvage packs and make some good ones up! We could build up some useful data on cells.

Peter

[Mike Dabrowski 2000]

Peter
I have been giving the same thing a lot of thought.
I built a pogo pin based clip on back probe that clips on the BCM and brings out all of the battery pack taps, and it has a .1" pitch header for a breadboard. I built it back in 2002, but since my pack is still working good, I have not used it.

I believe the low current you are proposing could uncover a bad subpack cell, but the car will present a much higher load, and could show the problem subpack more easily.
I have the drawings of the clip on probe if you want to make one, and then you could diagnose the pack in real time in the car with minimal disassembly.
I used Labview to do some test runs using relays to switch in the subpacks to the acquisition system.
As far as the end of charge determination, the cell temp and the fast rise then fast dip that a fully charged NIMH has at end of charge, are both dependant on the charge rate, and may be difficult to pick up with a 1C charge.
If you would like the drawings or any other help, send a PM