Need help with Ignitor Circuit

[RayRay1132]

I've recently gotten hold of some old high pressure sodium fixtures and they have bad ignitors, (yes I know LEDs exist this is just for fun...). Anyways, I was wondering how I could make an ignitor. I'm assuming its just a standard voltage multiplier. If so, what would be the easiest way to build one?? (I just want to build one, not buy one.)
Thanks in advance -Ray

 

[crutschow]

They probably use some kind of ballast (which sodium bulbs need), such as the old fluorescent lights had, that generate a high voltage for starting.

 

[RayRay1132]

They probably use some kind of ballast (which sodium bulbs need), such as the old fluorescent lights had, that generate a high voltage for starting.

It is a simple choke with a center tap.

 

[KeepItSimpleStupid]

if they are anything like xenon, it's an Arc lamp. The electrodes need to face the right direction, but there is no asymety. A 1000 W arc lamp operates at about 22V 40A and uses a 40 kV pulse to ignite.

It's best to keep the ignitor close to the lamp. You have this large wire gauge transformer in series with the power supply.
it has to generate lots of voltage and pass a lot of current on the secondary, There was about 80V DC prior to ignition.
You can power control, but you want voltage and current limits. We actually controlled intensity.

We had manual ignition, so I was allowed to push the start button for 10s, if it didn't start, it needed a cooling period/

The start function used an Spark gap oscillator and therefor you got like a 600V spike on the primary.
Remember, this transformer is in series with the lamp.

I think there was a spark gap at the input too. Then the lamp itself had a gap. the high voltage had to jump to gaps to start. Then you ignite a plasma and go into current control or power mode.

A neon transformer is lossy, so it has a very high voltage secondary and when the secondary shorts, it essentially current regulates.