*small* spark generator

May I suggest an electronic gas stove igniter?

The most common type runs fro a i.5V cell using a single transistor blocking oscillator to generate about 250V - that charges a capacitor which is dumped into a HV pulse transformer a few times a second by aa gas discharge tube - replace the tube with a thyristor and you can control it.

If you're lucky, you might find a handheld igniter, alternatively go to recycling and look on the backs of scrap gas stoves.

Alternatively; look for a tiny SMPSU (DVB-T tet top box type thing) with a little re-arrangement of the blocking oscillator wiring you can make a 5V or

Have you looked at a Tazer?

Potato gun?

ScottM,

You might take a look at this item and its associated video:

Spark Gap Igniter - 4.8V -- $15

It's not completely clear whether their "gap

The text say 3A draw, the vid says 20A draw. ?????

Also, be aware Sparkfun, being a small company, doesn't have the financial/legal horsepower to fight for yer privacy. They recently got into trouble cuz some dirtbags used one of their items to make illegal credit card readers and the govt came down on them, demanding lists of all their customers. A deal was worked out and sparkfun released only the customers that had recently purchased the offending device, not the entire customer list. Not saying you shouldn't deal with sparkfun, jes that you should know what's what.

nb

It's usually a good if the OP states the objective. For a potato gun most use a BBQ igniter.

Fair enough. It's to make a prop for a play involving mad science. The spark will be created inside a small brass funnel, and I'd like it to be bright enough and loud enough to be noticeable at a distance, which is why I want a longish spark. I'm not igniting anything with it. I'd prefer not to use a tazer circuit, on the grounds that those are meant to knock people over; I want something less likely to cause lawsuits if there's an accidental shocking.

I hadn't considered a continuous arc, on the grounds that I figured that would eat a lot of energy, but I have to admit that would look impressive, if possible...

I don't think the arc will be heard any distance in the magnitudes that you are describing. If it's for "Mad Science" why not incorporate a Jacob's ladder, much more impressive?

Tom

With a motorcycle ign coil (twin HT lead "wasted spark" type is best) and a

After playing with the experiment for a while, my ears were left ringing!

This sounds promising, but I've been digging into this, and I need

I probably should clarify the problem, since people are talking about AC power sources and jacob's ladders.

The prop is a handheld device, about 6-8" long, made mostly of 1.5" ID PVC pipe. Maybe 3" inside that pipe are available for electronics. On the front is a small brass funnel, with an insulated wire inside it. I want sparks to jump from the wire tip, to a point on the inside surface of the funnel. The funnel will point at people (the audience when possible) and I want the spark to be plainly visible at a distance. (I figure the acoustics of the funnel will make sure it is audible.) The power source is a small 12v battery, probably in a backpack or pocket. (A 9v battery would be great, but I figure if I want a few thousand sparks, that's impossible.)

My guess is, this is a hard problem, which is why I'm here. If I wanted a jacob's ladder, I can find designs for those myself, online. If I had AC power available, I could probably tinker up something for myself and eventually get it right. It's the combination of small space, long spark and small battery that makes this something I don't want to just tinker up. It needs to work, it needs to be efficient, and for safety reasons I want the circuit to fit entirely within the PVC. So please, think SMALL. Thanks!

Not to be rude but, I think it maybe just out side your abilities?

You can use a Cockcroft Walton voltage multiplier..

Follow the guide lines on the actual rectifier chain, you can keep adding to the string..

but for your circuit to start off with.. You can get a HV Transistor of a MOSFET or NPN and what you do is use a high frequency pulse source like a 555 timer to switch this on/off. This transistor will pull current through an inductor and when the transistor switches off, the release energy is many times higher in voltage. This output is connected to the first stage of the Cockcroft Walton multiplier.

P.S.

Look up a MARX generator.

Hope that helps.

It might. I'd ruled out a Marx generator because it looked like the size of the spark gaps was critical, and it didn't look easy to pack it into a small space without distorting the wiring and messing up the gaps. But the CW multiplier looks interesting.

I do have a couple questions. Since I need to get into the neighborhood of 45kV, the last few stages are going to be handling serious voltage. I've never worked with diodes that can handle that voltage - is this likely to fit into my space constraints? Can you recommend specific parts?

Secondly, I was hoping to get sparks at a regular, controlled interval. (The goal is to have a 555 drive a 4017 at 10-20Hz, with each output of the 4017 controlling an LED, except for the last, which triggers a spark. Hence flash-flash-flash-flash-flash-flash-flash- flash-flash-SPARK!) The CW circuit is clearly going to have some recharge time, and if it's a few milliseconds that's fine, but I suspect it's a lot longer. How is it calculated?

(At some point I'm expecting someone to look at this, and say something like "12v to 45kV, 2 sparks a second, in 5.25 cubic inches - give up, it's not possible".)

The sparkfun thing is small but needs three amps of input so you'll have to be careful with the batteries - probably need at least a couple of D cell sets in parallel to prevent them from getting hot.

You'll also have to take into account that battery voltage varies over time and also with temperature and current draw. Getting exactly five volts from a battery without extra hardware isn't easy.

Battery voltage varies over time like this:

eg. You might have to use four D cells which have been partially discharged to work with that spark gap igniter from Sparkfun (ie. get them down to about 1.3 volts each before using them otherwise you're looking at 6V of input and it might burn out).

How long does it have to run for? If it's only short bursts then it might not matter much if the batteries warm up a bit. For continuous use it can be a problem.

The sparkfun thing definitely looks like a good starting point. Get hold of a couple and play...

There's seven of them in the video.

Then look for a very small SMPSU transformer as I've alreadyy suggested - you can run the blocking oscillator from 4 or 5 AA NiMh.

Anything over about 20kHz you can get an apparently continuous arc with way less than the voltages you're on about.

If you want individual sparks the BO is easy to control with a single additional transistor.

Just a follow up:

I was ordering some Arduino stuff from Sparkfun so I slipped a couple of their spark gap generators into the order.

They're perfect for any kind of mad scientist machine. You can get decent sparks from a single AA battery.

This is with three AA batteries:

Disclaimer: 3xAA is not the way to do it for more than a couple of seconds. The batteries are overloaded, they'll heat up, bad things will happen...

Can you point to a non-private video? I'd like to have a way to generate single sparks cheaply to collect audio impulse responses.

-- Les Cargill