James Lerch wrote:
On 28 Aug 2004 18:27:56 -0400, (DoN. Nichols)
wrote:


In article ,


James Lerch wrote:


I'm currently on the hunt for a nice water cooled 2mf cap, but not
having much luck. Most of the ones I'm finding or for high voltage,
not high ripple current.



Am I reading this correctly? You're looking for two
*milli*-Farad capacitors? (0.002 Farad or 2000 uF (micro-Farad)?
(Granted, back in the old days, 'm' was used as an abbreviation for
"micro", but that got dropped about the time that they started calling


Yes, no, maybe... ahhh dang it

(after some research, I say) YES.... 2000uf or 0.002 Farad.

This can't possibly be right! 2000 uF as a parallel resonator (tank)
capacitor? For 25 - 100 KHz? It sounds way too high! Or, is this
in the DC power supply? If you are burning up the caps on the DC
supply side of the power transistors, it is because you have "shoot
through" currents. That is where both the power transistors in each
half bridge are turned on at the same time. The currents could be a
pulse of 10 - 50 nS width and thousands of amps!

You need a MUCH better circuit to control "dead time" between the
two transistors. I use a pair of IR2113 FET gate drivers, and use
a resistor-capacitor circuit at the input of the driver to delay
the turn-on signal. I put a doide across the resistor to make the
turn-off transition faster than the turn-on. (In my particular designs
for servo amplifiers, I use 3.3 K Ohms and 100 pF, and get about 100 nS
delay between the turn off of one transistor and the turn on of the
other. It works really well.)

Jon