DC high voltage multiplier

[jayc73]

Hi All,
Ive spent the last couple of days looking into Voltage multipliers and I've come up short.
I'm looking for a simple way to turn 5 to 12V dc into 20 to 25kv.
I've looked at Cockcroft-Walton Voltage Multipliers but cant seem to figure how to calculate the required capacitors ( KV and capacitance) and diodes to achieve the end result.

This is for a spark plug tester

I've gone down this route after buying an off the shelf " high voltage " spark plug tester that is utterly useless.
It ignites the spark plug for a few seconds then ( the best way to describe it) it fades , just like a flashlight running out of power.

It runs from a 12v dc power supply, to a small PCB, to a " high voltage generator" ( epoxy filled box of tricks) to the spark plug being tested.
Testing with my multimeter, everything is fine up to the high voltage generator

I've decided to go down the DIY route and try to replace the high voltage generator with a voltage multiplier so it works as it should.


Any help would be great.
Thanks,
Jay

 

[ArticCynda]

The traditional way to do this is with a high voltage transformer (the type used in CRT tv and pc monitors), because they can supply a larger continuous current than a long chain of voltage multipliers. A transformer can also be combined with voltage multipliers. If you start from a DC, a MOSFET power stage can be used to drive the primary side of the transformer.

If you want to multiply transformerless, then the amount of energy (and thus the output current) you can pump depends both on the capacitor sizes and the switching frequency. Increasing the caps will lead to higher charge currents and thus larger diodes, wires etc. A higher switching frequency might be an easier way to do it.

 

[jayc73]

Thanks Artic,
I did a little more investigating and found a couple of things.

There's a LM317T CC12CVW mosfet on the PCB that heats up very quickly. It does have a small heat sink but after 10 seconds use the heat sink becomes too hot to touch.
From the 12v input supply it looks like it feeds the mosfet ,on through to a resistor and diode , to a 25v 100uf cap, to another resister , then another 25v 100uf cap , on to another diode and finally to the high voltage generator.

By the time the 12v input reaches the last diode , it drops to 3.1v

It looks like the hotter the mosfet becomes , the more the voltage drops off .

 

[ArticCynda]

An LM317 is an adjustable linear voltage regulator, not a MOSFET. So if you control its reference pin the way you normally would with the gate of a MOSFET, it would explain why this component heats up so quickly. Keep in mind that the LM317 has a rather large dropout (voltage between input and output), so depending on the current you draw through it, a lot of heat is dissipated.

Voltage regulators have an internal thermal shutdown, i.e. they throttle down the output current if they get too hot to prevent their own catastrophic failure. If the load isn't adjusted accordingly, this indeed results in a drop of the output voltage.

 

[BradtheRad]

Here is a thread is about a similar topic:

https://www.edaboard.com/threads/275566/

It mentions a Cockcroft-Walton multiplier, and transformers.

One post suggests a transformer which has several secondaries, connected in series.

 

[D.A.(Tony)Stewart]

Any old TV/CRT will have the flyback + voltage doubler. 25kV. Monochrome used 15kV

Or you can modify a microwave oven transformer circuit with 1kW arcs.

What are your specs?

 

[chuckey]

If you drive your inverter from a 12V source that is pulsing on and off, your continuous spark will change into a series of pulsing sparks, which would be enough to measure the voltage and reduce the power in the circuit so it will run for longer. You could just dump the whole thing in a container of oil to make it easier for the heat to be dissipated.
If the transformer produces N V p-p, then the diodes and capacitors need a PIV of > 2 X N V.
Frank