Author Options:

How to make a lightweight 10kv 500 watt power supply? Answered

I need to convert up to 500 watts of 12 volt to at least 10 kv. It's important that it has both +10kv and -10kv. Right now I'm thinking flyback converter driven by a 555 to 1000 volts then voltage multiplier to 16kv. It's important that it's light so I don't want a regular flyback transformer. I'm wondering how duty cycle effects a transformer. Does it effect it like it would with an inductor-based boost converter? If so, then according to this calculator http://www.ladyada.net/library/diyboostcalc.html I would need an 88% (88% high?) duty cycle to get 12 volts up to 100 volts with a boost converter. Does this mean with a 1-10 transformer I could get 1000 volts from 12 volts with an 88% duty cycle?

If so, then I need a light weight 1-10 transformer. I'm thinking winding a toroidal transformer will be best? So I could get a ferrite toroid (I'm thinking material 61, size 114) and wind a 1-10 transformer, but I need to make sure it doesn't saturate. Do I find out the saturation just like I would an inductor in a boost converter? If so, then according to the same calculator, the higher the output current, the less henries I need. So that means if I don't draw enough power it will saturate? I'm probably going to use 25 khz, so according the the calculator, if it's 100 watts then I need a little over a 25 henry primary. But if I use 500 watts, then I only need a little over 5 henries? This doesn't seem right. Soooo according to this toroid winding calculator http://www.66pacific.com/calculators/toroid_calc.aspx with a material 61, size 114 toroid, I would need 2 primary turns if it was drawing a minimum of 100 watts (the lower inductance/higher current can't be right can it?) and then I would need 200 turns.

Sooo my question are (haha) is this a good idea for a light weight high power hv power supply, if so are the similarities I'm assuming between boost and flyback converter correct, and if so, how many/how do I figure out how many windings I would need on each side of a toroidal transformer/ how do I figure out what inductance I need for the primary. I have pretty good experience making boost converters but I've never made anything hv. Very long question thank you this is for an ion thruster.   


LOL, all this negative nancy make me laugh :)
you can use a old automotive ignition, they convert 12V to 10KV to ignite the spark plugs, they are alot of you tube videos about that.

Hey my wife's name is Nancy.....

Now, a spark coil is at most only a couple of watts

Leaving 498 watts to come out of the zero point aether... :)



6 years ago

Power is proportional to weight
weight is inversely proportional to frequency
Saturation is inversely proportional to frequency
( lower frequency saturates sooner )
12V to 10,000V is a turns ratio of 1:833 twice for +- 10kv
500watts at 10kv is 50ma reflected back to 12v is 42 Amps
at a 100% efficiency .  A 12V 50 Amp battery is heavy !
To support the energy transfer your core material BH curve
must be known, how many watt seconds it can support
I don't think you can get away with 2 primary turns more like 30
and you will not be able to fit the secondary winding even on a
split core toroid .


Okay so it is probably better to use a microcontroller, maybe 20mhz. How does saturation relate to current?

A's running through the outlines of the power transformer design you need to follow to make a 10kV 500W, high frequency transformer. It's a difficult thing to do even when you're a professional designers, like A. and me. In fact, unless our palms were crossed with enormous amounts of money AND there was an overwhelming need to make a special supply, we'd probably buy a suitable design in from a specialist.

Switchers at this kind of power/voltage are HIGHLY specialist designs.

Okay but couldn't I use a flyback transformer to generate 1kv then voltage multiplier to get it up to 16kv? And I'm not sure about this but can't I get 1000 volts from a 1-10 transformer with the correct duty cycle?

You are under a mistaken concept,
  • Duty cycle control can only reduce a transformer's output voltage at the secondary.
  • A flyback is never going to deliver 500W at 10KV.
  • A Multiplier also cannot deliver 500 Watts.
  • A 1:10 transformer can only multiply the primary voltage by ten at the secondary.  It also divides the available current by ten at the secondary.

+10 (kV ?)

This isn't a project for a beginner.

Not really. Like I said, you've chosen a very hard problem.

One way to look at it is,

The 12V supply must be capable of 550 Watts
When running depending on efficiency of design, the transformer
will take 15 Watts  85% eff.

If the core Saturates fore some reason ( power transistor shorted )
the windings will begin to absorb all of the 550 Watts and get very hot ! ! !


Saturation of a magnetic core occurs when the amps × winding_turns
generates more flux then the core can contain forcing additional flux to
permeate the surrounding space which does not require energy to align magnetic core domains ( inductive reluctance ).. 
Now as long as there is voltage the current will raise up to the limit of
the DC copper wire winding resistance turning the coils into a toaster !


You wouldn't lie to us would you; it's not for a suicide-machine?


500 watts at 12 volts means the power supply will be supplying 41 amps!

I think this may be your first problem.

10Kv isn't something you should be playing with unless your fully experienced - Your talking about the sort of voltages that you get on the electricity transmission pylons - Would you mess with them?

Sorry to be negative but
a) I don't think your going to get this to work.
b) it is going to be costly
c) It would be VERY dangerous.