# Finding the optimal voltage for a sealed transformer

I stumbled across a sealed transformer (step-up) which i believe came out of a neon power supply. Its completely covered in epoxy like goop and the internals cannot be seen. However,on the tin it indicates a output voltage of 10kv but i have huge doubts if it takes 230V in the primary.

I have taken some readings of both the primary and the secondary using a LCR but not really sure if i can get a good voltage ratio.

I'm just willing to know what's the best way to drive this.. (P.S no feedback coil either) The right hand side was the input for 230vac. Left side is the output. And the soldered points is the primary winding. I have NO idea what happens in the right side all is assumed. I also do not have a lot knowledge in transformers. (learnt physics up until 12th grade) Thanks,

• Neon transformers are designed to have lots and lots of leakage inductance. It's kind of "what they do." So I'm not sure what to make of your readings, right now. What are you intending? (And yes, I'd probably try out the primary quickly and see what happens after adding a $\ge 20\:\text{W}$ $4.7\:\text{M}\Omega$ load on the secondary. The worst is that I lose the transformer. Oh, well.)
– jonk
Commented Feb 10, 2020 at 6:39
• I just need a voltage source of 1000V which i can control. This is the only transformer i got. I just tried attaching a 5khz 12v square wave to the primary and got a output of 452V output. Still do not understand how can you power this with 230Vac at 60hz. Commented Feb 10, 2020 at 6:42
• Just set up a ~1000:1 voltage divider with some high voltage rated resistors so you can measure the output voltage and apply 120VAC or less to the primary (use any sort of step down transformer if you only have 240 available). That way you can measure the ratio and figure out what primary voltage would give you 10kv on the output. That gives you the rated input voltage Commented Feb 10, 2020 at 6:42
• Does the frequency matter in this case? Its confusing me to point where i had to pull out a LCR... Commented Feb 10, 2020 at 6:44
• is it iron, or ferrite? Is it intended for 60Hz, or 5kHz? Why are you driving it with 5kHz? What frequency is your LCR bridge testing the inductance at? Post a picture of it. For only 1000V, you may be better off with something else, like a Cockroft Walton. If it's designed for it, you power it with 230VAC by connecting it to 230VAC. A safe (safer, safety is relative with 230V and 10kV), then you connect an incandescent filament mains bulb in series, if you're foolhardy enough to connect it to mains, for the first time. That's current limiting and indicating! Commented Feb 10, 2020 at 6:51

It's probably a switching supply built to drive a neon tube (they require a high output impedance since the tube itself has a negative resistance characteristic). The device would rectify the 230VAC input to create about 300VDC and then chop that to get 10kV using a high-frequency ferrite-core transformer. I would expect if it worked you might get about 1/20 the output with 12V in, which is what you are seeing.

If you search for "neon power supply 10kV" (without the quotes) you'll find examples of this kind of product.

So you should be able to feed it about 30-35VDC and get 1kV out, but it will be deliberately poorly regulated for load changes (more than a handful of mA drawn will significantly drop the output voltage).

That's just an educated guess.. it's hard to tell what's inside there.

• If so then is it possible use a AC bridge rectifier and hook it up to a mosfet (driven from a function generator)? Commented Feb 10, 2020 at 7:34
• It should run from DC, and it has it's own oscillator inside if I am correct. Just give it DC (either polarity) from a bench supply. And be careful, tens of mA even at 1kV is not to be trifled with. Commented Feb 10, 2020 at 7:35
• It turns out that the wires are cut off. Completely stuck in the epoxy. Is there an external driver which could do this? (As in what is it called) Commented Feb 10, 2020 at 7:44
• It's potted so I have no idea what might be connected to whatever you can reach. The transformer if completely disconnected from the circuit would use a custom power switching circuit operating at tens of kHz or more. Commented Feb 10, 2020 at 7:46
• Interesting, Thanks for that. Commented Feb 10, 2020 at 7:50