Thanks guys, this is making good sense. I see what you mean, Mac, about the ambient light sensors in series, and with some diffused led back lights, it seems simpler and more reliable than the IR setup. A .223 at top velocity is a reasonable upward/lower bound for the project. A 6mm bb at 156 m/s (Canadian legal limit) would generate a `37 usec pulse. Would it be at all advisable to run all the sensors into the same comparator (if it’s done in series, the V1 role could be played by the other gate) or is it better to keep each node separate to reduce noise? I’m also thinking about ditching the old array in favor of this new layout.
If I and use a wired-or configuration for the two sensors, would I have to ensure that one pulse ends before the other pulse begins to get accurate readings? I imagine this would be done with a balance of sensor spacing and amplifier/comparator design. In the event that adjusting for it becomes unreasonable, there are other interrupt pins that can be used. I suppose this will depend on how the sensors turn out.
In regards to specific parts, digikey spat out this one as the most likely candidate, a 1125-1009-ND photo diode. At $8.62, it’s a fairly reasonable choice (next to the $60 units.) Looking around for 100khz comparators, I’m finding a lot of $10-20 chips, and then the INA148UA. If anybody has any other suggestions, I’m all ears.
I can imagine dealing with the distance calibration in three ways: 1) a very accurate ruler and a steady hand, 2) if space allows, designing a PCB and letting the machine mark the distances (this may not be practicable for the velocities at play,) or 3) incorporating some sort of proximity sensor to allow the unit to self-correct (seems expensive/overly complex.)
My soldering skills are just on the happy side of terrible, so depending on the complexity of the circuit I may go so far as to do the PCB/soldering oven thing. I’ve never designed a PCB before, so that would be an interesting project. Maybe later down the road.
Re: ambient light vs IR sensors … the chronys I’ve used and seen were all the former type. They all seemed to work w/o problems, assuming you had enough light. Adding some backlighting can’t hurt. I’ve read the IR type also work but you do need IR illuminators, aka IR LEDs. In either case you should make the FOV as small as practical so the shadow is as large as possible. Yet a small FOV means a small measurement area. Maximizing sensitivity up front will pay off later, when you figure how much noise you have to contend with. And that may mean an array of small FOV sensors, all just overlapping side-to-side, perhaps with some lens over them, to get a usefully wide shooting area and good sensitivity.
Have you figured out how much the sensor signal will change w/a bullet ? Or a BB ? I’d be concentrating on this as I think everything else will work after this part is figured out.
As for OR’ing front and back sensors … it certainly makes any delay added by any following circuitry
the same, certainly a good thing accuracy-wise. I guess it depends on how you intend to measure the time. Will 2 pulses on the same pin work ? Measure leading edge to leading edge ? Doing so would seem to exclude measuring something long, like an arrow, as the pulses would overlap.
FWIW a basic LM139/339 comparator has a response time <1 usec for any “large” signal. Digi has them for <$2 ea. Newer, better designs may be available at the same price.
Mee_n_Mac:
As for OR’ing front and back sensors … it certainly makes any delay added by any following circuitry
the same, certainly a good thing accuracy-wise. I guess it depends on how you intend to measure the time. Will 2 pulses on the same pin work ? Measure leading edge to leading edge ? Doing so would seem to exclude measuring something long, like an arrow, as the pulses would overlap.
Just a quick reply, my coding adventures with the atmega on the previous page of this thread lead me to believe that I can do this with 1 pin, with the drawback of only one gate being set off at a time. There is an external interrupt built into the mcu that will copy a 16 bit hardware timer value to a register and fire off an ISR. This interrupt is connected to an internal comparator, which I don’t think can be used directly with the sensors I’m looking at (I’m pretty sure.) I can also connect other external interrupts to the 16 bit timer if the single pin thing doesn’t work out.
I will spend some time figuring out what the signal change from various projectiles might look like, and do some work from there. Night!
It looks as though the largest pulse I can expect out these diodes will be ~400mv.
Reading the atmega datasheet again, I think I may be able to use the internal comparator instead of an external one. Comparing schematics, it seems as though the wired-or setup could be done with only a few caps/resistors hooked into the MCU.
It’s always fun to learn new things about a chip you’ve had for years!
