How to keep a power up spike out of a battery charge controller?

[Deleted]

I'm having a problem with a circuit that I did not design. The unit uses
Lithium Ion batteries (laptop battery pack) when AC is not available. The
problem is, when the unit is turned on with the batteries, there is a large
initial spike (approximately 8 amps) which the charge controller sees as a
short circuit and immediately shuts down the battery. My first thought was
to use a capacitor to send the spike to ground. Then I realized that it was
the capacitors already in the circuit that are charging up to make it look
like a short. A resistor was added to attempt to reduce the impact of the
spike but that failed.

I then thought of putting a coil in series with the battery. The engineer
shot the idea full of holes and mumbled something about back emf when the
circuit was shut down and also complained that the spike could jump over the
coil. Not having an engineering degree, I shut up but continued to think
about this problem that should be solved electronically instead of settling
for cracking open the $100 battery pack and taking battery power directly
from the batteries to avoid the spike getting into the charging circuit -
which works fine but is kind of hokey and voids the warranty of the
batteries.

In the circuit below I show my thoughts (albeit in a poorly done schematic).
I figured I could use a coil with a ferrite core and a diode to ground on
each side. I figured the diodes would satisfy his emf concerns and the
choice of core would keep the spike from going across the coil. Note there
is an on/off switch (not shown) between the battery and the rest of the
drawing.

Is this an idiot idea? Should I use an iron or any other type of core for
the coil? Maybe only one diode? Get a job not even remotely connected to
electronics? Any better recommendations?



___________nnnnnnnn___________
| | |
+ _ _
BATT ^ ^ Out to rest of circuit
- | |
| | |
______________________________

 

[neon]

HERE IS THE IDEA When you turn on a switching regulator like The inrush current can be realy big as much as 10x of what is required to run normaly. What i don't understand is how can you get a spike across a battery that is realy hard to do. if there is a spke is a downward type because of drain on the battery. like your car when you start it deeps into 8v. I think eventualy you will find that the batteries are if not bad at least weak and replacing them will cure this spike. either way there is nothing passive that can solve your dilema. I sugest you do not try adding anything to cure this problem except battery replacement

 

[Deleted]

Why not just use an emitter follower, driven from a capacitor which
charges relatively slowly? Possible failure if the destination dislikes
intermediate voltages but if you keep the time constant to something
sensible (20ms or so?) then it would rise slowly enough to avoid the spike
but quickly enough to keep the destination circuit happy.

Main snag is the .6v drop that you'd be stuck with thanks to the BE
junction.

You *could* get around the above with a second transistor but that's
starting to head in the 'sledgehammer->nut direction.

Dave Johnson.

Thanks for your response. I didn't want to go nuts with this design change.
Adding transistors is like redesigning the charge controller and I can't do
that.

Interestingly, my supply guys accidently gave me a brand new battery with a
larger capacity and it had exactly the same modification to it that we have
been using - that is, separate the load from the controller and tap power
directly from the batteries. Unfortunately, said battery is physically too
large for this particular project. I guess the manufacturer realized they
had a problem with their controller but chose not to upgrade this model of
battery for some reason. I don't speak Chinese, so I can't call them with
any questions...

Bart