```> Original Poster: "Ted Rosenberg" <TRosen1-at-Tandy-dot-com>
>
> Adam: I would agree that the MMC be enclosed but perhaps slightly vented to
> encourage circulation if warm.
> The question: How to determine the correct value of a bleeder R. Base it on
> the total voltage of the MMC? And if so, to achieve what current?

Well, in keeping with common sense, I would design it so that the time
constant is short enough such that by the time you walk from your variac
over to your coil the cap has only a non-lethal amount of energy stored in
it.  Preperably a lot less than a lethal amount. :-)

Of course, you need to make sure that the resulting current through the
resistors does not exceed their power handling capability.  Assume worst
case scenerio where you pull the plug with the cap full charged, and your
other connects to the cap have just mysteriously failed at that same
instant.

Real world example:

Your Tranny Vt =           15,000 volts RMS
Your bleeder resistor =    R ohms (we'll find a good value)

So, say you pull the plug with 2 * 1.414 * Vt volts on the cap.  You can
start with Vc = 1.414 x Transformer RMS output voltage (Vt), but if you've
got a cap-xfrmr resonance match it could be *a lot* higher than this.  So it
would not be unreasonable to multiply the voltage by 2 or more to be safest.
So I'm using 2 in this example.

So, Vc = 2 * 15,000 * 1.414 = 42,420 volts

Now figure that this Vc is appearing across your R ohms resistor chain.
Your time constant is RC, so after RC seconds, there will be 0.367 * Vc
volts on the cap.  This means that after 1 time constant, there is still
15,570 volts on the cap, and 2.42 joules (E=0.5*CV^2).  This could still
give a nasty shock.  Since the capacitance is small, a good figure to look
at is I^2 integrated over t for the time constant of the pulse, RC seconds.
It is desired that this be at least 0.01 or lower.  Again, lets assume worst
case scenerio, of a human body resistance of 500 ohms.  If you grabbed the
cap at this point, it would discharge from 15,570 volts to 5700 volts in 10
micro seconds, giving you about 0.004 A^2-sec.  This would give you a chance
at life, but it is still more than you would want to experience, so figure
that 2 * RC should be less than the time it takes to walk over to your cap.

If it takes you only 1 seconds to get to the cap, you will need a resistor
of 0.5 seconds/ 0.00000002 farads = 25 megaohms or less.  If it takes you 2
seconds, you could use up to 50 megaohms. 4 second walk... 100 meg.

Now you have to choose between walking slow, or buying power resistors.  At
25 megaohms, you would need a 70 watt resistor to safely handle the initial
1.7mA current through 25M.  Since duration is short though (this is not RMS
continuous power), you can get away with a lot less.

OK, now here's where I say that I've been out of school for 4 years, and
even when I was an EE student, power was not my thing (audio DSP was).  So
I've probably made some mistakes in the above calcs, which will be
pedantically combed over in future replies by other readers, so be sure to
follow this thread ;-) I disclaim all knowledge of ever having written this.

Visit here for some electrical safety:

http://plop.phys.cwru.edu/repairfaq/sam/cord/c4/c4_mod01/mod04-01.htm