Hi everyone, im after some guidance in regard to s project im doing.
Looking at using ~20 horizontal IR leds to detect a specific sized object as it passes through a field of beams.
They will span approximately 50cm vertically, and i need to know which beams are broken at exact points in time. (Sampling approx 2kHz)
The beams cover a 1.5m distance and have Tx / Rx alternating on each side.
I want to make the beams basicly laser accuracy without having to align the Tx and Rx, so want to modulate the LED’s such that only the corellating Rx can detect the Tx it matches with on the other side of the beam. (As LEDs have a beam angle that would be received by many Rx on the opposing side.
Can i get away with just modulating at ~5 different frequencies and pattern them? would this require hardware? or would i need to use 5 different types of modulation and is that even possible on something like an arduino uno / mega. (cheap)
First microcontroller project and i’ve not chosen something easy.
Can i get away with just modulating at ~5 different frequencies and pattern them?
That might work, but sounds like a complex solution to a simple problem.
You might consider turning each IR LED in your array on briefly (While modulated at whatever frequency your receivers are looking for) and then moving onto the next. Then, have an array of receivers that also are each listening only when the corresponding IR LED is on and looking to see if it’s getting a signal or not. That eliminates the need to exactly target each receiver but maintains the ability to filter out what beam is broken since only one beam is active at any given time.
LbHawtspur:
No, i had this idea alos, but the height being surveyed is much larger than 2x the diameter of the ball. more like 20x
I don’t understand what you mean here. Are the balls bouncing?
Anyway, it sounds sort of like a ‘light curtain’ application. While most are used for safety applications, those are binary: either the curtain is clear or there’s an object in its field of vision. Some can perform measurements, though. Here’s one: