Why difficult? It's just like a HV spark generator but in reverse. As long as it's got enough turns in total and roughly the right turns ratio.

Say 1000v : 12v is 1000t : 12t
but with expected low efficiency you need a higher output turns number like
1000t : 30t
then 1000t is a pain to wind so I would start with something like 300t : 10t or so and see what happens.

After that it is a matter of experimenting, and fortunately with low output turns like 10 it's pretty trivial to add or subtract a few turns on the output.

For experimenting I'd just hook it up and put say a 1k resistor on the output winding, then get the input current sparking and look at the output waveform on the 'scope. That should give an idea of how much output voltage and current should be available and turns can be tweaked accordingly.

 

Why difficult? It's just like a HV spark generator but in reverse. As long as it's got enough turns in total and roughly the right turns ratio.

Say 1000v : 12v is 1000t : 12t
but with expected low efficiency you need a higher output turns number like
1000t : 30t
then 1000t is a pain to wind so I would start with something like 300t : 10t or so and see what happens.

After that it is a matter of experimenting, and fortunately with low output turns like 10 it's pretty trivial to add or subtract a few turns on the output.

For experimenting I'd just hook it up and put say a 1k resistor on the output winding, then get the input current sparking and look at the output waveform on the 'scope. That should give an idea of how much output voltage and current should be available and turns can be tweaked accordingly.

I didn't say it would be difficult to wind one. The OP had asked for a part number, so he could buy one.

 

My apologies Ron! On re-reading your post I realised I made a mistake and thought you said "it would be difficult to WIND a pulse transformer". Sorry about that.

 

Found a ccfl inverter 12v / 1000v for operating a uv light at:

http://www.digikey.com/product-detail/en/BXA-12529/289-1025-ND/268545

Any thoughts on whether this transformer, used in reverse, might reduce 1000v to 12v?

 

The complete PCB cannot be used in reverse, and there's no way to know the properties of that pulse transformer were you to remove it. I'd keep looking.

You might learn a lot from using some transformers you can find easily, such as an automotive ignition coil or a TV flyback transformer. You can probably find either of those for next to free, and they'd tell you a lot about what you might want to buy. Any transformer might "work" but most power or audio transformers don't have the 100:1 turns ratio you need.

 

Okay, if I wind my own transformer to reduce 1kV pulses down to about 12 volt (+/-) pulses, I have a few questions, keeping in mind that I am dealing with very low current pulses and conversion efficiency is important if this is going to work.

What core material is suggested, iron or ferrite? Torriod or standard shape? What guage for the magnet wire on the primary, the secondary?

 

The transformer (removed) from that inverter is probably not a bad place to start.

If you want to experiment with windings I would suggest buying a few different sizes of potcores or ferrite E core pairs from a hobby supplier as they pick fairly typical magnetics and then just playing with the number of turns. When it comes to pulse transformer winding I'm a fiddler, as doing the calcs may get you close but won't get it all the way you still need to fiddle with turns and material grades generally if you want a really good result. Plus it's fun.

 

What core material is suggested, iron or ferrite? Torriod or standard shape? What guage for the magnet wire on the primary, the secondary?

I suggest iron because it has the highest magnetic permeability, and I believe a standard shape will be easier to find and wind. I have some iron toroids that are very cool, but they're not cheap and I could only find one supplier.

Choosing wire size will be a challenge. I'd try to get a core large enough that you can use the same wire on both sides, because wire is cheaper if you buy more at a time. But with your high ratio, this just may not be possible. Ideally you'll use the largest gauge that still allows you to pack all the turns into the space you have. This will reduce losses due to DC resistance of the wire. But don't go thicker than 18 gauge or so - it starts to get hard to wind by hand.

I developed an Excel spreadsheet for estimating coil size (on my toroids) and wire length for different gauges and number of layers. I can share it with you if you're interested.

Maybe someone else has an idea on this, but I have in the back of my mind that having the same amount of copper on both sides is optimal. If that's true, you'd choose the smaller gauge to be about 1/10th the diameter of the larger gauge. That would put 100 more turns through a given cross section of space.

 

Just keep in mind that permeability of core material is frequency dependent. Your pulse will be short, with a fast risetime, so you may want to get a core material that has permeability that holds up at ≈50MHz.

 

Just a tiny bit of information,
When I needed a pulse transformer for a High Pressure Sodium ignitor, I used a tiny audio transformer from Radio Shack. Something like 1K to 8 ohms.

Point is, you can run a transformer backwards and amazingly cheap stuff that never thought about high voltage might work.

 

Just a tiny bit of information,
When I needed a pulse transformer for a High Pressure Sodium ignitor, I used a tiny audio transformer from Radio Shack. Something like 1K to 8 ohms.

Point is, you can run a transformer backwards and amazingly cheap stuff that never thought about high voltage might work.

Just keep in mind that the turns ratio is the square root of the impedance ratio. In your example, the turns ratio is 11.18:1.

 

Bottom line: It worked.
In fact, I can't remember what year I did that (somewhere in the 1990's) and it's still working.

 

cold cathode inverters are cheap, can be found at some PC shops.
Simple two transistor oscillator, small transformer, and some kV output.

Even cheaper:
http://www.ebay.com/itm/New-Rotatable-USB-Ionic-Ionizer-Fresh-Air-Purifier-PC-Laptop-/270955324774

Has a similar circuit inside, small transformer + voltage multiplier.
You would have to solder out the transformer somehow, and work it in reverse.

 

For under US$3.00, I ordered it, not much financial loss if it doesn't help. Will see if the transformer in it is of use.
http://www.ebay.com/itm/New-Rotatabl...-/270955324774

 

Here a photo I have taken a while ago of this purifier circuit.

If you unsolder the transformer be careful the pins may break off from the transformer. If you have not done this before for instance from CFL PCBs, at first add solder generously and heat with a 50W soldering station, carefully, so at best it literally falls out.

I don't have a guarantee if this may work for your purpose but worth trying isn't it?