Spark and 0.3A of currrent flow when wiring battery circuit

6 views
Skip to first unread message

Mike Brown

unread,
Nov 6, 2011, 10:54:51 PM11/6/11
to 914ev
As I mentioned in the previous message, today I was going to make the
motor turn for the first time. I had all of the battery bus bars and
cables attached, except the final connection from Battery 14's Neg
terminal to the most-ground post. All of the partial-pack voltage
readings matched the manual's checklist. When I went to put the final
lug on the mount-ground post I got a big spark, which I assume is a
short somewhere.

I started trying to narrow down the problem by isolating the battery
boxes, removing cables from battery terminals, the controller, the
motor, the contactor. I couldn't get rid of the spark.

In frustration, I put my meter between the battery and the most-ground
post and found I had a current draw of 0.3A. Not really much, but
enough to make a spark and make me think twice about proceeding. This
is with the aux battery disconnected, the DC-DC disconnected, the
motor disconnected, key off and breaker off. None of the car's 12V
features are on.

The only thing out of the ordinary is a PakTrakr system, with a
voltage sensor wire on each of the 20 batteries. I disconnected each
PakTrakr unit's power supply wire and the LEDs all went dark. Even
when powered up, the PakTrakr should pull no more than 32 mA, and I'm
seeing about than 10x that.

Any suggestions?

Mike Brown
No, not that one, another one.

Mike S

unread,
Nov 7, 2011, 2:05:14 PM11/7/11
to 914ev

Not sure what motor/controller you are using. I too was surprised by a
big spark when I connected made the final connection of the battery
pack to the Azure Dynamics DMOC445 controller in Pumpkin (evalbum.com/
2466). The next time I made that connection and before bolting the
heavy wire in place, I used a 25-ohm resistor to get both sides of the
connection to the same voltage first, and had a digital oscilloscope
looking at the voltage across this resistor. What I saw was an
exponential decay from the full pack voltage to zero with a time
constant of 3 milliseconds. So it looks like the DMOC has a large
capacitor sitting across its dc input terminals and that capacitance
is 3 x 10^-3 seconds divided by 25 ohms or 120 micro farads! To charge
that capacitor to 185 volts requires an energy of over 2 joules. Max
current with the 25-ohm resistor is 185 volts divided by 25 = 7.4
amps. With no resistor to pre-charge the input capacitor, current is
going to be well up in the range where welding is done!

What is hard to understand is the lack of any advanced warning about
what happens when one makes this last connection. The big capacitor
does a fine job of protecting the controller from voltage transients
(like hooking up the battery pack) but the hapless soul who makes the
last connection for the first time is totally unprepared for what
might happen to his cardio-vascular system, the wrench in his hand,
the words coming out of his mouth, etc.

Note that pre-charging the controller input capacitor *safely* takes
takes careful planning. At that last connection there is lethal
voltage with tons of available current in close proximity to ones
pinkies. Think about it and be careful!

Mike Sasnett

Mike Brown

unread,
Nov 13, 2011, 3:14:21 AM11/13/11
to 914ev
Thanks for the background. I tried the spark test again this morning
- it sparked but after I left the wire on for a minute didn't spark a
second time. I quickly bolted on the lug and everything looks good.

Mike Brown
No not that one, another one.
Reply all
Reply to author
Forward
0 new messages