All too often in photography, you simply don't have enough light. The camera's internal flash provides lots of bulk photons, but at what is possibly the absolute worst angle. Wouldn't it be great if you could move the flash off the camera? Or use more than one light per scene? With the help of my guide, you can do just this. I'll show you how to build a slave flash that never needs batteries and recharges in a fraction of a second. The whole thing fits nicely inside a standard light bulb.
Let's start by collecting everything we need
Behold: my diamond light bulb cutting jig. Made from scrap wood held together with clamps, the jig holds a rotary cutting tool at just the right height for slicing open the bulb. The light fixture holds the bulb steady, allowing for easy rotation.
The real secret is the diamond cutting disk. It cuts through glass with ease, and only occasianal small cracks. I find it works best to cut the bulb in a single pass, rotating it in the opposite direction of the cutter. Hold on to the top of the bulb near the end, or the remaining uncut glass might crack under the weight.
A bit of soapy water and a scrub brush make short work of the diffusive coating. Now here's the annoying part: after all that work of cutting the bulb open, you have about a 50:50 chance of the bulb being useable. Inside the glass electrode support is a tiny fuse which may or may not have blown when the tungsten filament broke. It's pretty easy to see if the fuse is intact, but check with a meter if you're unsure.
You can bypass the fuse by running a new wire through the center column and soldering it to the tip of the bulb, but I think it's easier to find a new bulb and start over. Make sure everything is completely dry before powering it up.
Here is the circuit I designed. A simple voltage doubler charges the main 160 μF capacitor to about 330 volts from regular household 120v AC. The protection fuse is already built into the light bulb, and pair of bleeder resistors ensure the capacitors discharge when the power is disconnected.
When the flash fires, the capacitor voltage drops near zero in a few milliseconds, extinguishing the zenon arc. The design of the voltage doubler requires several complete AC cycles to fully recharge the main capacitor, effectively forcing a quarter second delay between flashes. This may seem insignificant, or even annoying, but it is absolutely vital. If the flashbulb were supplied with 330 volts continuously, there would be nothing to quench the flash. The arc would burn continuouslty until the flashtube melted, possibly exploding.
The flashbulb triggering circuit and neon indicator light are identical to the unmodified flasher.
Due to the shape of the bulb, an untraditional construction method is needed to utilize the available space. Most components are dead bugged. The following images outline my connection method.
Begin by soldering the 160 kΩ bleeder resistors between the leads of the two electrolytic capacitors.
The diodes connect the capacitors as shown. The neon indicator light is soldered to the terminals of the 160 μF capacitor from the original flash unit. A tiny resistor taken from the board and mounted in series prevents the light from burning out instantly.
As most of the camera flash circuitry was used to generate 330 volts from a battery, the size of the board can be significantly reduced. Cut along the line as shown. You may have a different model, so be wary of any differences.
The newly miniaturized circuit board sits nicely on top of the capacitors.
Solder the wires of the main capacitor to either side of the flash tube, observing the correct polarity. A bit of electrical tape on the back of the reflector prevents the capacitor from shorting out.
To create a more diffuse light, I frosted the top of the glass by sanding it with a very coarse grit sandpaper. It may take a while without a power sander, but it softens the flash quite nicely.
I'll get around to finishing this writeup as soon as I can.