Ok, I've already made an ATMega88-based chronograph for measuring the speed of the BBs from my airsoft guns.
The programming and the circuit around the processor was easily enough made, but, as always for me, the mechanical stuff ended up terrible.
By "mechanical stuff" I'm referring to the tube which the user must shoot through, which has two points where the BB must break an IR light beam which is sensed by a group of phototransistors... Long story short, this arrangement isn't very stable in the long run and really hard to "manufacture" easily. So...
It looks like this:
What I was imagining was to replace the tube with a vero-board with two parallel lines of light sensing components (phototransistors or photo diodes).
The idea would then be that once every .5 seconds or so, the processor reads the light level from both and use it as a reference value. When the BB is then shot, the shadow it casts on the photosensors as it passes each of the lines would be compared to the reference value and trigger the timer to start/stop.
My question is: Based on the collective experiences of people on this forum, do you guys think this idea is viable? Are the photo sensing components sensitive enough, and is it possible to perform A/D-conversions fast enough to trigger on the passing BB with any decent accuracy?
Also, as a bonus question, do any of you have suggestions to what the photo sensing and amplifying circuit could look like?
I imagine the two parallel lines of sensors to each consist of 3-10 transistors/diodes and some sort of amplification before the signal enters the A/D circuit in the AVR...
Note: The BBs are expected to move 50-250 m/s (that's 150-850 fps, roughly).