David Brandt wrote:
> 
> Boy, am I steamed.  I put all that effort into getting good pack isolation,

David, yes, it's possible but unlikely with new batteries.

Disconnect the pack from the controller.

Connect the resistor between the most positive end and the  frame.
Resistance is pack voltage/5mA.

Measure the current through the resistor. It will be be 5 mA in worst
case
(if negative end happened to be shorted to the frame). You likely get 
less value (than 5 mA)

If you get no current, connect to the negative end and repeat.

If no current, connect controller and repeat.

Don't measure it while charging.

If still no current, your charge's GPF itself is faulty because
>5mA must leak to the frame somewhere to trip it.

If you get current, move resistor connection point toward middle
of the pack until the current is minimal. This battery is at fault.

Now, connect 100k resistor parallel to your meter to lower its
resistance.

Now if you measure voltage with it between positive terminal of
suspected battery and the frame and get *steady* +4.2V (not disappearing
in a few seconds to 0V), and between negative of the same battery
to the frame -8.4V, you have the leakage path from internal cell #2
through the microscopic crack in the battery case (which wicked
conductive 
electrolyte) to the frame. It is very hard to wipe clean and insulate 
bottom and sides of the battery. If it has side crack/pinhole with 
connection to internal cell through electrolyte, it may be enough
to trip GFI no matter how clean the top surface is.

To verify, remove this battery, put it near by onto insulating 
surface, (better outside the car, in 3-4 polyethylene shopping bags),
connect it back electrically (long thin hook up wire is OK for test)
and try to charge. It should not trip (if GFI itself is OK).

If this is the case, find the leak. Connect one lead of voltmeter to a
square piece of thick foil (2x2") and other end - to neg. or pos post.

Pressing the foil against battery body sliding over its sides and bottom
will locate the leak point - the voltage will indicate that. For 
internal cell leakage it will not be zero whether neg. or pos post
used for measurement.

Victor

Reply via email to