post Spark-gap Tesla Coil

Filed under Electronics, Videos by Iain (6:35 pm, November 18th, 2009)

This was a couple of years ago. I never got round to posting it here at the time and the video’s a bit rubbish. It’s much prettier in real life! But seeing as I’m going through the backlog, why not. For some reason the camera would only capture half the streamers – it would be fine half a second, then go streamer-blind for half a second. Likewise it would only show the lighting of the flourescent tube at 2:01 intermittedly, although it was lit constantly. Very wierd. Quantum effects?

Hmm, that topload’s none too impressive either.

Read on for the tech stuff.

OBLIGATORY WARNING: Don’t actually build this. The dangers are real if you get things wrong or aren’t confident in your abilities.

I don’t want to appear authoritative on HV only to have you blow yourself up after taking my advice, so I haven’t given a design recipe or schematic; instead I’ll just describe my choices and results.

The Tesla primary is powered by a bike ignition coil being oscillated by a 2SK899 MOSFET at 12V, switched by a 555.

To protect the LV side I have a .22uF 440V cap across the LV side of the bike coil, in parallel with a disposable camera neon indicator (to monitor back EMF) and a 5W 4K7.

The HV from the bike coil is half-wave rectified by a diode string (15 x 1N4001 and 4 x 3108) and charges a single DIY 2L saltwater capacitor that I estimate to be around 1nF. The cap is charged to about 2kV and discharged via a spark gap to the Tesla coil primary. The capacitor is made from a lemonade bottle. A 6mm threaded aluminium rod forms the centre electrode and passes through the cap for connection at the top. The bottle is 75% filled with saturated saline solution, with a 1″ layer of clean engine oil on top for corona suppression. The plastic of the bottom is the dielectric. The other electrode is a piece of wire soldered to aluminium cooking foil covering the outside of the bottle up to the level of the saline. The cooking foil is covered in insulation tape.

The Tesla coil itself consists of 8.5 turns of 1mm copper, with 2mm spacing on a 64mm PP former. The secondary is hand wound – it’s a little over 800 turns with no overlaps and took a little over an hour to wind. I wouldn’t want to go any finer than 0.2mm wire for hand winding though. I shot the secondary with a few coats of clear lacquer over a week or so to keep it nice and tight.

The whole thing runs cool; that heatsink on the MOSFET was complete overkill. If not the bottle capacitor (I have doubts about the long-term suitability of the plastic dielectric), the weak link is probably the poor ignition coil, salvaged from an old Triumph and surely being overworked here. It’s possibly oil-cooled though and hasn’t complained so far. I have a nice flyback transformer from a valve TV to try if that dies.

The output voltage from the Tesla secondary is over 15 kilovolts, which is enough voltage to produce 2″ streamers to air (the breakdown voltage of air is around 3kV/mm).

Yes, it nips a little bit when it bites you! Note that the current involved is very low so it’s “reasonably” safe. With only 10W input (which it doesn’t have), even 100% efficiency (which it doesn’t have) only gives a current of 6.6 x 10^-4 amps at 15kV (0.66 mA); so I felt OK playing with the final HV from the secondary. That was just as well as it much preferred arcing to fingers rather than the handy grounded screwdriver it was supposed to arc to!

I was a lot more wary of the HV on the secondary of the bike ignition coil however, in the Tesla’s primary circuit. I had no doubt that could kill me.

I was extremely careful at all stages of this, but even so I managed to get a nasty DC jolt at one point from the fully charged capacitor, which gave me a sore arm for a couple of days. As all the HV sites say, playing with this stuff WILL KILL YOU if you don’t know what you’re doing.

Besides pretty streamers to air, there are other fun things to do. Placing a flourescent tube anywhere near a Tesla coil will cause it to light up, and a regular filament bulb will turn into a plasma globe when brought close to the coil.

A clear-envelope bulb gives the best effect:

teslacoil

Power input: 500mA @ 12VDC (6W)
Primary: 8.5 turns 1mm copper, 2mm spacing on 64mm PP former
Secondary: ~850 turns 0.2mm enamelled copper on 40mm PVC former
Topload: ~60mm alu foil-covered Xmas tree ball
Streamer length: ~1.5-2″ to air
Arc length: ~3″ to grounded rod
Spark Gap: Static, 4mm
Tank cap: ~1nF saltwater cap (2L lemonade bottle)
BPS: ~ 120?

Leave a comment

ruld rurd
© 2001-2009 Iain Craig RSS Feeds Entries | Comments