I have a 10 kV 23 mA ignition transformer.

Apparently the way to make a jacob's ladder is to just connect the electrodes to the secondary, plug the primary into the wall, and call it good.

But of course, as a transformer this thing has no current limiting, and I have no way to calculate the effective resistance of the arc. (I'm guessing a theoretically perfect transformer has no current limiting, but a real life one probably might limit current in some manner I am not familiar with, magnetic saturation, etc, but I'm guessing also this is not the "correct" way to limit current when designing a circuit)

What's the best way to limit current here? I can use a power strip that will trip at 1800 W, but this thing will draw 230 W at the max of the spec.

Do I just fuse it at, say, 0.5 A on the primary side and use a slo blow to avoid blowing the fuse on the inrush?

Or do I use a single resistor on one half of the secondary side to limit the current?

Or use a variac on the primary side along with a killawatt meter to measure the current draw as I turn up the voltage?

What's the correct approach here?

Yes, I have worked with HV before and I understand this thing will kill me easily, I have built plenty of woodburners with scrap MOT, etc and I know how to handle HV.

  • \$\begingroup\$ Some people call me crazy, but I used a neon sign transformer, a wire coat hanger cut in two, and some insulated stand-offs mounted to a piece of wood. Connected the wires that went to the neon to the coat hanger pieces, plugged it in and held my breath. It took less than two minutes to get the gap right and the thing arcing nicely. I AM NOT a trained professional and am just sharing my experience. \$\endgroup\$ Commented Jul 18, 2018 at 20:43
  • \$\begingroup\$ Any fluorescent ballast has enough inductance to define current limit. Since arc has negative resistance it drops in voltage as current rises and the “- ESR” depends on current while the ballast is high “+” reactive X(f) that limits the current and thus the holding current rises with gap as it rises and extinguishes at an electrode gap limited by the current threshold, Ih. Similar to an SCR. \$\endgroup\$ Commented Jul 18, 2018 at 21:59
  • \$\begingroup\$ The more current, the higher the ladder goes. Some use a variation. Wear sunglasses 🕶 \$\endgroup\$ Commented Jul 18, 2018 at 22:05
  • \$\begingroup\$ Just as a note: 10 kV accelerated electrons striking metal target implies "soft x-rays." Pretty soft. And yes, there is air in the way to hinder the final velocities. But x-rays still occur. See any of several papers by J. R. Dwyer from about 2003 to 2008. (Dental is somewhat 'hard' and is 70 kV, typically.) \$\endgroup\$
    – jonk
    Commented Jul 18, 2018 at 22:52
  • \$\begingroup\$ Not to worry too much. Current-limiting is good practice, but older transformers are designed to run for hours igniting oil in the furnace across a spark gap. See pupman.com/listarchives/1997/may/msg00486.html \$\endgroup\$ Commented Jul 19, 2018 at 1:19

1 Answer 1


What has worked well for me with such transformers is 100 watt light bulbs in series with the primary voltage. Add more bulbs in parallel for more current.

The reason is that incandescent light bulbs are very non-linear about current consumption. As the current rises the filament in the bulb get much hotter and its resistance goes way up, limiting the maximum current flow.

If the transformer shorted the lights would come ON full brightness. As it is they will flicker with changing loads on the high-voltage side. Do NOT try resistors as they will simply get hot as a linear device. Do NOT try them on the high-voltage side as the voltage will just arc across the resistor and burn it up.

I know farmers who do this for their electrified fences to protect the transformer, and to limit the current when cattle or horses touch the wire.

  • 1
    \$\begingroup\$ Awesome! This seems very reasonable. I'll use a variac also, and a killawatt meter to ensure I'm drawing the correct amount of power. \$\endgroup\$
    – cat pants
    Commented Jul 19, 2018 at 19:48
  • \$\begingroup\$ Sounds like a good idea. Please be careful with HV. Normally I would never mention this but if you accepted or liked this answer could you please check the box's at the upper left of my answer. \$\endgroup\$
    – user105652
    Commented Jul 19, 2018 at 21:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.