0
\$\begingroup\$

A previous student built a measurement setup, which involves dropping a small sphere (2.4 mm) onto various surfaces to measure rebound characteristics.

As it falls it will (partially) interrupt a red laser beam pointed at a photodetector, thus triggering a high speed camera via micro controller. The photodetector output is "converted" to 5 V TTL by means of a comparator op-amp circuit.

However, the detection accuracy is very limited for fast objects, sometimes it will trigger sometimes not. The op amp can be tuned by potentiometer to be more sensitive, but at some point, the ambient light is enough to trigger the circuit.

Can someone suggest a better way of detecting the drop, either circuit wise or other way of approaching the problem. (We can't mount detectors to/under the target surface.)

thanks

EDIT: I forgot to mention: when the camera triggers a very bright light is turned on for ~3 seconds. During this time + 2 seconds after the light is off, the system is set to ignore any input to prevent any false detections.
The photodetector has an integrated amplifier, supplied by a 12V battery. We noticed, that it never reaches such high voltages and will produce ~ 3V while the laser is active. The Zener diode is for the protection of the circuit in case the detector would output higher voltages.

circuit diagram by the student:
circuit diagram

Setup, the photodetector is far back to not catch as much light from the top light source: setup

\$\endgroup\$
10
  • \$\begingroup\$ "comparator?" or "op amp?" Can you show us the detector circuit diagram? \$\endgroup\$
    – glen_geek
    Commented Jun 13, 2023 at 22:33
  • 1
    \$\begingroup\$ Do you have the electronics skills to drive the laser with a square wave, and put a tuned filter and detector in front of the comparator? This could give much greater rejection of ambient light. It's why TV remote controls work in ambient light, by modulating the IR beam at 38 kHz. \$\endgroup\$ Commented Jun 13, 2023 at 22:49
  • \$\begingroup\$ Raphael, do you need to start the high speed camera before the strike and subsequent bouncing so that the velocity just before impact can be also gleaned from the recorded images, as part of the post-processing analysis? Also, what's the physical setup (dimensions, etc.)? \$\endgroup\$ Commented Jun 13, 2023 at 22:57
  • \$\begingroup\$ Switch to an IR laser and photodiode with an IR filter. A lens can also be used to focus the photodiode field of view towards the laser, limiting background signals. \$\endgroup\$ Commented Jun 13, 2023 at 23:10
  • \$\begingroup\$ Raphael, is there some range of area within which the ball falls though? Is that part of the reason there is some difficulty? Or do you know exactly and very precisely the line along which every ball path must fall and can align the laser beam so that it always intersects a falling ball at some point on its surface (not necessarily dead center, though?) \$\endgroup\$ Commented Jun 13, 2023 at 23:15

3 Answers 3

2
\$\begingroup\$

One possibility, which I have used, is to use a microphone to detect the sound of the ball hitting the surface and use that signal to trigger a camera. I used a sound level meter for convenience since it contains the microphone and amplifier in one. You can use a comparator circuit like you already have to generate a trigger signal.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ That will only work if the camera starts recording before the ball drops, and stops at the impact. \$\endgroup\$ Commented Jun 13, 2023 at 23:37
  • \$\begingroup\$ the camera can record before and after the trigger by a set amount, so this would work. Thanks for the idea! I might try this. \$\endgroup\$
    – Raphael
    Commented Jun 14, 2023 at 9:39
2
\$\begingroup\$
  1. Be sure the trigger is AC-, not DC-coupled, so that steady ambient light has no effect.

  2. Exclude all flickering light, such as from fluorescent or LED lamps. Most incandescent lamps have less flicker, and it can be eliminated by using DC power. Also keep computer displays, which flicker, far enough away to prevent accidental triggering.

  3. Consider dropping the ball through a light-scattering detector.

    The spread laser beam goes through the detection area from 0° - 180°, and exits into a light trap, e.g., a rough cone covered in Vantablack.

    The detecting optics are at right angles, 90° - 270°, also looking into a light trap.

    Any particle falling through the junction scatters light to the side, creating a sharp signal with little noise, easier to detect.

\$\endgroup\$
1
\$\begingroup\$

However, the detection accuracy is very limited for fast objects, sometimes it will trigger sometimes not. The op amp can be tuned by potentiometer to be more sensitive, but at some point, the ambient light is enough to trigger the circuit.

This statement suggests the possibility that detector circuit is too slow to catch a fast laser-light pulse.

From the photo, photodiode appears to run through long wires to the comparator circuit. These wires add capacitance, which adversely impacts speed. These wires also allow ambient electrical noise to make triggering erratic. The comparator circuit should be placed very close to the photodiode, with very short wire lengths. You can run long power-supply wires and digital_out wire to the comparator/photodiode circuit, but add a bypass capacitor to DC supply lines right at the comparator DC supply pins.
The purpose of the zener diode is unclear - it adds capacitance which impacts response-speed.

Photodiode operates in photovoltaic-mode. This makes comparator bias currents critical to circuit operation - some CMOS comparators with low bias currents might work fine. Comparators with bias currents of wrong polarity won't work in this circuit without a load resistor.
In any case, with no load resistor, pulse response time differs between rising edge and falling edge. A loading resistor helps to make rise/fall more equal, but reduces sensitivity.


Driving a laser directly from an I/O pin seems risky. Typically, a small milliwatt-type laser runs close to two volts, perhaps 20-30 mA. Laser threshold current may not be much less - That ESP WROOM 32 module might be starving or over-driving the laser, since it dials back I/O current depending on other loads. This crude laser-drive method may be a matter of good luck.
A more-constant laser current source may help stabilize photodetector threshold. A more-constant laser supply might simply consist of a series-resistor to +3.3V supply (or to GND), in the 50-ohm ballpark.

\$\endgroup\$
2
  • \$\begingroup\$ The photodetector has an integrated amplifier, supplied by a 12V battery. We noticed, that it never reaches such high voltages and will produce ~ 3V while the laser is active. The Zener diode is for the protection of the circuit in case the detector would output higher voltages. \$\endgroup\$
    – Raphael
    Commented Jun 14, 2023 at 14:32
  • \$\begingroup\$ But I can try moving the comparator closer. \$\endgroup\$
    – Raphael
    Commented Jun 14, 2023 at 14:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.