Well, it turns out the resistor number 4 is just what I think is a 'saftey' resistor that drains the capacitors to ground when the unit is turned off. This resistor connects the charging circut input voltage (430DCV) and the ground.

I poked around with the voltmeter and found out that the power supply is AC and looks like it is split to about 400 v when it enters the series of blue "chokes." Well these are really probably diodes and they form a rectifier circut, because the output is about 430 DC volts. So, I had that wrong.

Anyway, after getting rectified to 430 DC volts the voltage is distributed evenly to all three exposure circuts (ie "10-2", "10-3" and "10-4" all have the same input voltage). Each circut then has its own set of resistors that then connect right to the respective high power capacitors.

Also, all three main capacitors are charged to the same voltage, irrespective of where the 3 position switch is set or if it is in neutral.

The 3 position switch, then, determines which capacitor is physically attached to the xenon tube. When the 3 position switch in in 'neutral' there is no voltage at the xenon tube. Pressing any of the 3 buttons applies 430 volts to the red wire of the tube.

My ignition capacitor charges up to about 117 volts and seems to fire OK. I did test the output of the ignition coli and it is well over 1000V. The xenon tube does show a faint blink, but there is not full discharge. This is very frustrating because there is certainly full voltage across the leads to the bulb socket and the capacitors are all charged appropriatly (I presume 430 volts is appropriate).

So I have narrowed it down to:

1) Not enough voltage on the ignition capacitor (may need more than 117V?)
2) Ignition coil weak and not 'sparky' enough (not high enough output voltage, need more than 1000V?)
3) Bad xenon bulb (though why would it sometimes seem to work fine)
4) poor bulb socket connection

At this point, after testing all the above, I could not get ANY full discharge from the xenon bulb. Just a faint flash from the ignition circut.

Still seems repairable, I just need to think about it a bit more...