Ignition Coils

Okay, so what on earth is an ignition coil? Well if you know a little about cars, you'll know that a 4 stroke engine goes: SUCK, SQUEEZE, BANG, BLOW. Literally.

The BANG bit utilizes an ignition coil to produce a spark across the spark plug via what's known as a distributor, initiating the power stroke of the engine.

Ignition coils are essentially pulse transformers - transformers that work using pulses of current, instead of standard alternating current. These pulses that are applied to the primary coil are magnified and the resultant voltage appearing from the secondary coil can be very high.

Image courtesy of Magnet Lab

I managed to get my ignition coil for $5 from the local car wreckers. There is limited demand for these things, so the wreckers usually sell them for quite a bargain. Alternatively, you could strip one out of your neighbour's car - easy!

Incidentally, old cars also have other goodies in them, like speedometers in the dashboard, or money in the glove box (haven't come across this one yet...).

On the whole, scavenging through old cars can be quite fun for any Parts Geek, and makes for some interesting photo opportunites. Of course, getting a Parts Geek to leave after scavenging for old cars and car parts, now that can be the real challenge!

To run the ignition coil outside of the car, specialised circuitry is required that produces pulses of voltage to the primary coil. I took a very common design (a 555 timer driving a MOSFET) and built it. Well, more like lashed it up... as seen in photo.

In the photo, it shows the initial setup with two ignition coils. The sparks in the photos below are from one ignition coil only however. I consider the circuitry used here quite unsophisticated, and I have since fine tuned certain components in the design.

Here is a poor circuit board being electrified. Notice the corona at the top of the wire. This is due to the 'roughness' of the copper - very small sharp kinks in the wire lead to the build up of charges, and when this charge builds up past the dielectric breakdown threshold of air, it shoots off a minute stream of particles into the air, also known as corona.

The sparks tend towards the nearest grounded objects, in this case the tops of those electrolytic capacitors.

Those sparks are only about 4cm long - an estimated 44kV (1.1kV/mm).

This is a quick demonstration of what happens in a lightning cloud. The charges build up on the tip of the wire (cloud) and discharges back to ground, creating a spark (lightning) and associated 'snap' noise (thunder).

Notice how the spark isn't anywhere near straight, despite electricity's constant desire to return to ground via the shortest possible path. In practice, the spark path also depends on the temperature, humidity, wind draft and other properties of certain spots in the air.

Needless to say these properties, much like the 'butterfly effect', influence the spark in a chaotic manner and the exact path of the spark is thus rarely deterministic.

This spark is estimated at around 50kV.

As the wire is brought closer to ground, sparks are produced much more frequently, as the amount of charge build up required to traverse the air gap is reduced.

These are quite intense and the camera has been saturated.

Sparks can be incredibly noisy and this is one example where that occurs!

When the wire is brought very close (this is about 1.5cm), the sparks are so frequent that a smooth arc forms. The charges no longer have to build up very much, and immediately jump onto the circuit board through the plasma channel.

The temperature of the arc probably exceeds a few thousand degrees Celsius, and in a matter of seconds, the copper wire has heated enough to start melting off its insulation.

Notice that stray spark catching the other corner of the metal can. This rogue spark probably formed due to wind drafts.

In this similar photo, you can see that the air around the arc is heated super hot. In fact, the metal can is untouchable after a few seconds of arcing.

This is what happens during arc welding, except that the arc here is actually much longer. However it lacks the power and amperage that is forced through the arc in welding applications.

This is a demonstration of the ineffectiveness of normal plastic insulation at very high voltages. The sparks simply track along the insulating plastic until it reaches the grounded terminal at the end. This effect is known as surface tracking, and can be a real nuisance in many HV applications.

In this picture, the high voltage terminal is in the center of an insulated wire at ground potential, causing most of the sparks to track right around the insulating plastic. Some sparks however, go right through the plastic, and this leads to a very burnt, crispy, and generally... well... unpleasant smell from the leftovers of what used to be plastic.

Insulators break down at a certain potential known as the dielectric breakdown voltage, which is typically specified in MV/m. I guess the lesson learned here is that plastic can be rather futile as an insulator!