From Grahams Wiki

This is a stupid project that i originally started for a victoriana game. I've made two of them, but there both in different states of repair, cos i didnt bother finishing them when the game got cancelled. Its basically a climbing spark created by a one or two car ignition coils switched by an Oscilator.

In both cases, I used a battery for safety purposes, and both have different safety features which i'm going to replicate in both.

Unsurprisingly, there still both quite dangerous and need to be used with care.

These pictures are from the second one, which is the more powerful one with two coils.

Contents 1 Examples 2 How they work 2.1 Mains (NOT RECOMMENDED!) 2.2 Battery 3 The ones i have built 3.1 The original one 3.2 My second attempt 4 Running

[edit] Examples

http://www.youtube.com/v/2vWsdr8PweA

http://www.youtube.com/v/sXmynGvDaeM

[edit] How they work

Really simple, There are two ways to make these things. You need an AC current from somewere.

[edit] Mains (NOT RECOMMENDED!)

If you use the mains, you can just transform it. You need some form of current limiting, and there are other dangers, not least of which is that you will be wired to the mains, and you will have to custom wind a transformer, or use a flourescent light transformer. Also, if your mains is on an RCCB (And most are!) you will reliably trip it as some of the current will be lost.

[edit] Battery

A battery is slightly more complex, however, substantially safer, as you can use car ignition coils or TV flyback transformers, which are designed for the voltages they are transforming. The only issue is that you need to get an AC current from somewere. I do this using a simple NE555 oscilator and a 2n3055 transistor.

As the maximum CE current is 15A, and i'm running of a 6v motorcycle batter, it means the maximum possible current pushed into the coils is 90w.

In reality, its much less than this, as the coils have a restance of about 0.4 ohm at DC, and more at frequencies. the transistor drops 0.7v, The NE555 only outputs about 60ma into the base resistor, which as the DC current gain of the transistor is only 40, which give Ic at 2.4A, leaves the wattage at about 12w.

Therefore after transformation to 20,000v, the current is about 0.6ma, which is well below the threshold of Lethality, although you will still feel it, and it will still burn.

Also, the oscilator is of high enough frequency that it will tend to run along the skin rather than penetrating and touching vital organs.

However, just because you probably wont die, doesnt mean you shouldnt still treat the equipment with respect. The skin is where lots of nerves end up, and it still hurts like hell.

[edit] The ones i have built

Neither of the ones i have built are actually complete. (As in has its complete set of safety things)

The only thing both want doing is the replacement of the resistive cables with copper cables. Resistave cables are used in modern cars to reduce any arcing. This is of course complete the opposite to what I actually want. Due to the physics of it, its not a problem when it fires, however, when the arc starts, the wires become a significant part of the resistance, and therefore the spark isnt as powerful or as bright as it could be.

In both cases, i decided to totally box off all the HT terminals for safey.

[edit] The original one

This one uses a single coil. The relay and control system are in the lid.

Mostly, this one is ok, but, spares have been nicked for the second one. Its missing:

• power transistor.
• I would like to replace the resistive cables with copper cables - Probably make more of a difference with this one.

[edit] My second attempt

My second attempt uses a reciperacal pair of coils to double the voltate at the prongs.

This one is missing

• tilt switch.

The circuit is slightly different, In addition to the extra coil, the switch has been moved out of the box, and the relay is now only used to switch the power and not to provide the feedback to keep the coil closed. This means the circuit must be held on, which is safer.

Note, the ne555 circuit is on a seperate board mounted on top again. This board is smaller than the previous one. The NE555 components where taken straight from the datasheet and it it set to oscilate at approximately 1kHz.

The basis for this calculation is that its in the expected firing frequency the coils would experience in a car, so they should work at their most efficient. This may be utter bunk however.

Note that the 2N3055 is the limiting aspect for this circuit. Its gain is not that good, and it could probably do with an additional transistor before it to amplify the signal from the Oscilator to increase the switched power. However another aspect of the circuit is that i would rather not accidentally electrocute someone if their curiosity exceeds their common sense, so lower power does have its advantages.

An additional modification is to wire a Condensor across the poles of the coil. this has the effect of making the coils oscilate slightly when the transistor switches off, and seems to sustain a slightly brighter gap. (Which should hopefully mean less stress on the transistor).

As with my first attempt, the prong outputs are covered in a wooden box so they cant arc between themselves or to the electronics underneath.

The control handle.	And again	The closed box	The innards
closeup of the coils, Note the consensor (Not on the circuit) and the reed switch at the bottom left.	The top lid with the protection removed. Also note the magnet for the reed switch.

[edit] Running

This is a picture of it running. Unfortunately, it doesnt look very impressive. I need to take a movie of it really. In good weather, it can jump about 5cm, although 3cm makes for more reliable firing. Unfortunately cos its so damp, and the battery needs charging, i'm using it at 7mm.