Questions on Voltage Mulitplier and Capacitors

I have almost finished my electrostatics and am now in the need of a slightly variable voltage multiplier. There are a few different solutions to my problem (that I could come up with) and I would like some feedback.
The voltage that my ESLs arc at is somehow dependant on the humidity (no suprise there). What I need is a knob that will slightly change the voltage output from ~1400V to ~2000V.

Option 1) Use a voltage multiplier on the input to the step up transformer. Then, I could use a potentiometer to control the input voltage. The only problem is that I'm not sure about using a voltage divider on wall voltage.

Option 2) Possibly use a voltage divider on the output of the high voltage. As I'm typing this I realise that this is a horrible idea. But, I'll throw it in just to make sure.

Option 3) Use a Rotary Switch to vary the output of the voltage multiplier. For example, on a voltage multiplier with 220 Volt steps, I can use a 4 step rotary switch to vary it from 1320 to 1980 Volts in steps of 220. Of course, this will require many steps (nine, actually) on the voltage mulitplier. Which brings us to another question. How many steps are possible? Surely, one cannot voltage multiply to infinity. What is the limiting factor? The capacitors? The transformers?

Also, are electrolytic capacitors viable for a voltage multiplier? I have some spare ones that I've pulled out of an old regulated power supply. But, when they are used the output of the multiplier is much less than expected by an order of magnitude. Some smaller orange caps that I have, however, seem to be up to the job. Can anyone give a reason?

Thanks for your input,
-Dan
 
Hello Dan,
Modern microwave ovens contain a step-up transformer and high voltage rectifier and 1.1 uF capacitor that gives the magnetron a 2000V+ supply that can deliver 800W.
Supply the primary with a variac and you would have an instant variable HV supply.
These supplies however are LETHAL, so some form of limiting would be required - perhaps a suitable capacitor in series with the primary.

Another approach might be to use a TV line output transformer (LOPT), or PC monitor LOPT.
These have integral rectifier, and when the primary is excited by 120V give a rectified/tripled output of 25KV or so.
Varying the primary voltage would give output variation, or use the focus or screen voltage pots to give variable output.
A standard TV drive circuit would be suitable, and typically run at 15,625 Hz.
Computer monitor LOPT run at higher scan speeds than this, and typically have a lower step-up ratio.

Your local recycling center would be a cheap source of suitable parts.

Regards, Eric.
 
Why not just use a variac

For voltage multipliers, the capacity required is a function of the load current. The ARRL handbook has a nice chart on this, and you can prove it yourself with a modeling program.

If you need to alter the secondary voltage, why not just employ a 5 amp Variac which will run you anywhere from $20 to $45 on EBay. Use the variac to control the voltage on the primary of the transformer.