I have project to build object (like baseball, tennisball etc) , velocity measurement. Planned setup contains IR start gate and end gate, velocity is calculated based on timestamps. Nothing fancy, very basic implementation with IR Leds (eg. TSAL6100, angle of half intensity=10 deg) and IR Receivers (TSSP4038).
We need several trip wires -> posts needs to be about 80cm high, because one cannot define exact position of ball / moving object. The gap between IR leds / receivers are now 20mm because need to get detection to triggered with smaller throwed objects as well.
I have done prototype and it works fine with 4 "tripwires", but when increasing number of IR leds and receivers, IR Leds starts to interference other IR receivers because of IR led beams looks like cone. Have tried with different kind of If values and pwm pulse frequencies (to adjust "sensitivity") without luck
So I cannot pulse modulate all infrared leds at once, because receivers won't detect objects good enough because of interference of other "trip wires".
Object speed passing trough "IR gates" would be between in "worst case" situation 0-55m/s what comes to requirement for triggering speed. I have used ESP32 for prototype..
Any idea how to solve this issue?
Have thought few things:
Would it work if try to narrow IR led beam with 3d printed covers with very small shaped holes (IR led power will be more than enough)
Second option is to drive IR-Led / Receivers in smaller groups (maybe group of 4) to avoid IR-led cross interferences... and run detection sequency from bottom to up etc. My only concern is that in worst case scenario setup should be able to scan whole area in 1.4ms timeframe if trying to catch targetted objects flying with max speed.
how about mix led and receivers so that there are "led-receiver-led-receiver..." pairs and in other post "receiver-led-receiver-led..." sequency... I think that this will still fail because of wider IR cone beams...
How about change whole desing so that there are only 1 post containing Leds and receivers and try to identify receiver IR via reflection of object that pass trough the sensing "gate"
Any other IR led / sensor combination that might work? I know that eg. bullet velocity chronometers can easily detect 2Mach objects, but of course whole setup is a bit different
Or totally other approach: Are there any other viable ways to implement this kind of velocity/speed measurement? Any practical ideas?
Sounds similar to the problems I faced on this project. I got improved spacing by having tubes over the transmitter and receiver. I also got an improvement by scanning the LEDs so that only one was on at any time. And you will see that I also alternated TX and RX LEDs.
I need to be able to detect balls / objects that are abt 20mm thick so as far as I understand receivers need to be aligned based on this requirement
Do you have all your emitters on one side, and all the receivers on the other?
Yes I have at the moment. Reducing overlapping might work if the diameter of cone is "manageable". Based on calculations IR Beam spot diameter with these leds will be 35cm if distance between IR Led and receiver is 2m... so separation of 4cm instead of 2cm is not probably enough (gap between receivers / leds now 2cm)
How about removing TSSP receivers and replace them with IR phototransistors and OP-amps? Would it be possible to detect intensity changes in photoresistors even when there are overlapping cones and trigger MCU interrupt based on these?
Practically more or less utilizing concept of this:
It would be interesting to understand basic functional logic of these "light curtains"... How they avoid interrupting IR led beams. Lenses? Sequential switching and analyzing if any ray is broken... something else?
I have sweeped trough IR Led vendors and it seems that there are not too much narrower available. I have found one viable with 3 degree ( VSLY5940)... but still spot size will be 20cm (2m distance). This approach would need certainly some extra lenses....
3°, that's going to require some precise alignment.
It you are off by 3° in the horizontal plane the IR receiver will be in the less than half power area and the power drops off quickly. Besides that, aligning an IR beam you can't see is not easy
Just thinking out loud: Would matters improve any if sender/receiver pairs of differing frequencies were interleaved? This would increase the separation between pairs of the same frequency and might skirt the dispersion problem.
One thing I'd verify is that you can detect one object at the highest anticipated speed.
This is good question.Based on TSSP40xx datasheets relative sensitivity drops dramatically when going down 100nm of wavelength....Only issue is that would need several different frequencies to get enough isolation.
I have done now some basic testing and it seems that IR Sender that is located 110mm above the baseleve will illuminate enough to get IR Receiver at height of 10mm (distance between rx and tx is 1.2m) to detect interefering IR rays...
Have now done tests to cover RX side with 30-50mm (length) tubes. Those tubes collects still too much interferencing IR rays from other IR Leds than the one in "sight"
