> my DC voltage is still a little high, I can afford to include this
If you bleed the first cap, voltage drop is slight. The other caps will bleed to the first cap's bleeder fine through the 300r resistors.
> I'm unsure about the resistance value or rating.
First: Define The Situation.
How much current is the supply delivering to Useful Load? You have the meter and resistors to get an exact value. I'll estimate "about 10mA".
What is a comparable problem?
You get your house-water from a pump, the rocking-bird type that outputs a gush-gush-gush. To get steady water in the kitchen faucet, you pump to a tank on the roof, to store water and smooth-out the gushes. This is really all your AC-DC-filter does.
Let us say the pump moves 10 gallons per hour (keeping the number "10" from the electronic question).
But it is getting cold here in Maine. The tank will freeze. I need to empty it.
Of course I could run the kitchen faucet until the tank drains. In fact your tubes will drain-off quite a bit of charge before they cool too much to suck.
But I might forget to open the faucet. Or worse: the faucet might jam shut (I got rocks in my well) and refuse to drain (just as you can't count on your tubes to do a good bleed-down).
Of course I could only open the leak when I want to drain the tank. That would normally be wise. But again, I may forget, or the valve may stick. A constant leak is sure to work every time.
So I put in a constant "leak", to be *sure* the tank will drain "quickly" after the pump is turned off.
How big a leak?
100 gallons an hour is MUCH more than the pump is supposed to give. I'd need a bigger pump just to cover the "leak" and still have water in my kitchen normally. It would drain the tank fast, but does it need to be that fast?
OTOH, 0.01 gallon/hour leak is insignificant extra load on the pump, but is liable to drain slow. Depending on tank size, I might have to shut-off the pump a month before a freeze (or before working inside the tank).
As a general thing, when you have a Major Load and a Minor Load, pick the minor load about 1/10th of the major load. Smaller won't save much, larger adds cost. This rough-guide may be wrong in specific cases, but is always a good first-guess to see where it gets you.
So we put a 1 gallon/hour leak in the 10 G/H pump and tank.
So you consider a 1mA leak in your 10mA supply.
This 1mA comes from a 300V supply.
300V times 1mA is 0.3 Watts. Because the bleeder must NOT ever fail, we more-than-double the estimated power to buy a resistor.
We are in the area of 300K 1 Watt.
Another point. Resistors have voltage ratings. Conceptually, high voltage will scoot around the outside of the resistor body. Certainly 30,000V will go-around a 1/4W resistor. Actually 1/4W parts are often rated 250V, 1/2W parts may be rated 350V. So a 1W part is probably safe for a 300V supply, but the wise specifier will look-up the data-sheet. Two 1/2W 150K resistors series is also safe, but if either resistor fails (for any reason), the bleeder don't bleed, the fail-safe now fails un-safe.
We should also compute the bleed-down time. In the water-tank, if it was a 100 Gallon tank at 1 gallon per hour leak, it takes 4 days to go down.
One real difference. The watter tank WILL drain to the last drop in linear time. The capacitor and resistor will drain along an exponental, and "never" hit dead zero. As you have a LOT of capacitance there, I fear any reasonable bleeder will not bleed down to a "safe" voltage before you can open the box and stick your hand inside. False "security" may be worse than respectful fear.
The bleed-down rate can be computed, but I have to go scare some trick/treaters off my driveway.
