Sooner or later you’ll come across a camera that is "capping". One side of the frame is exposed more than the other. What causes this is the two shutter curtains traveling at different speeds.

If you’ve built the shutter speed tester you have a good idea of what is needed already. Only this one has greater capabilities (can test curtain speed) for even less money. Same old catch though... takes even more work and requires a computer with a soundboard. But you have a computer or you wouldn’t be reading this. And I bet you have a soundboard with a stereo input port.

Theory: Computers have very fast built in clocks. If analog inputs are graphed along a time axis events can be measured and compared.

First download an audio manipulating program like CoolEdit 2000 and learn to use it. If you already have a similar program, try it. It’ll probably work. Basically these programs show sound as a wave along a horizontal time axis. We’ll fool the computer into thinking it is hearing light...

Next build up the simple circuit shown. Your tester should look similar to the pictures of the one I built. The only really critical thing is the 18mm that the phototransistors must be spaced apart. This dimension should be held very close. I drilled my mounting holes on a vertical mill to maintain a .001" tolerance. This is going a bit overboard but do try to get it accurate. It will pay off in test accuracy and ease of the head math. The phototransistor is lightly press fit into the hole. If the holes turn out slightly too large, where the phototransistor falls through, just glue it in.

A 35mm photo frame is 24mm tall by 36mm wide. 18mm is exactly half of 36mm and three-quarters of 24mm. So if the factory spec of the curtain speed is 10 ms the graph of the curtain speed should measure 5ms on the time axis for a horizontal shutter and 7.5ms for the vertical shutter. Just multiply the time measured in CoolEdit by two to get the curtain speed of a horizontal focal plane shutter. With a vertical focal plane shutter multiply by 1.33

Use in the same manner as the previous shutter speed tester only with this tester light intensity and distance are not as important. The phototransistors are only being used as switches. Turn off (mute) all other sound inputs...like the microphone, etc.... to get a smoother line between the pulses produced by the curtains uncovering and covering the phototransistors.

I used 3/16" thick light baffle material on the tester surface that contacts the cameras focal plane. All extraneous light must be blocked.

This tester isn’t as useful on lens shutters, for obvious reasons: a lens shutter doesn’t have two curtains. It will not tell the lens shutters blade speed. It’ll still give shutter speeds but the curves generally won’t be as nice as with focal plane shutters. This is because of using a phototransistor as a switch. The lens shutter acts like a moving aperture going from full dark to full light. This "confuses" a phototransistor because it doesn’t see the instant bright/dark that it sees on a focal plane shutter. This results in a hard to decipher graph... less steep up/down spikes.

Before I built the "stereo" unit described here, I built a "mono" unit. This unit, of course, will not give curtain speeds but does shutter speeds fine. The only circuit difference is that instead of two phototransistors in parallel there is just one phototransistor.

The biggest drawback to using a tester like the two just described here is the difficulty reading the graphs. The phototransistors don’t switch "on and off" instantly. There is a "rise time" and a "fall time". This makes the fast shutter speeds especially hard to read. But I have had luck reading up to 1/2000 shutter speeds with the phototransistor mentioned. Granted I had to make a big effort to understand what the different areas of the curves represented.

I’m working on ways of making the graphs easier to read. I’ve tried adding various size capacitors to the circuit but this seems to help very little. My next trial will be to add a comparator and Schottky. I’ll describe the results here after I try it.