1
\$\begingroup\$

I'm talking about this kind of sensor:

enter image description here

Question and then long description: can I do something to improve their blindness in front of window? At least for detection of object in front of the robot, like 10cm for example? not to detect distance but only prevent crash. Is there a sense to equip robot with them at all?

When I've ordered these IR receivers and leds I didn't think they will be so slow and more than that useless during day even in my room. My bad, must have googled more. They are too slow for signal modulation. I was going to wrap them around my robot, but now I don't know should I even use them or not. Just for obstacle avoidance at least. During night they are great for this.. Here's my configuration with foil shielding.

schematic

simulate this circuit – Schematic created using CircuitLab

enter image description here

Green signal is from IR leds ON part and blue when IR leds OFF Sensor under the table directed horizontal. I wave my hand in front of sensor. enter image description here

And this directed horizontal to window on level of floor. And today is very cloudy. Wave much closer. Around 4-5cm.

enter image description here

And this same as previous only it directed opposite from the window in front of window. enter image description here

Also when pointing to the side of window (not on it or the opposite) then signal is a little better. And all this when almost there is no sun..

Here's maximum frequency of modulation. 12mS or 83Hz.. IR led inside that foil tube. enter image description here

\$\endgroup\$
  • 3
    \$\begingroup\$ The basic trick is to make the LED flash, make the foto diode AC-coupled and then try to detect the LED flashes from the fotodiode. Your poor man's circuit will need a few extra components though. \$\endgroup\$ – jippie Jun 14 '14 at 13:06
  • \$\begingroup\$ @jippie IR receiver is too way slow get flashes. I wait 25 mS after led ON to sensor settle down. \$\endgroup\$ – a_vasilkov Jun 14 '14 at 13:10
  • 1
    \$\begingroup\$ This appnote discusses basic operation of an IR proximity detector pdfserv.maximintegrated.com/en/an/AN4622.pdf \$\endgroup\$ – jippie Jun 14 '14 at 13:34
  • 1
    \$\begingroup\$ Just us an IR receiver from an broken VCR or TV. They will filter out any unwanted noise, and only give a signal when a 38kHz IR signal is detected. Just have the IR flash at 38kHz. \$\endgroup\$ – Gerben Jun 14 '14 at 16:10
  • 1
    \$\begingroup\$ A window made from a piece of floppy disk will filter some ambient light, and pass IR. \$\endgroup\$ – Optionparty Jun 14 '14 at 21:48
5
\$\begingroup\$

A photo diode should be used in reverse bias mode for better speed, and AC coupling should be used to reduce interference from most ambient light sources.

schematic

simulate this circuit – Schematic created using CircuitLab

Some added design notes on the above circuit:

L1 lets most of the DC current pass to ground but blocks most of the higher AC current.

The op-amp gain needs to be fairly high (maybe 50 or more) as the diode's reverse current signal is very small.

The two high value resistors bias the signal at 0v so a maximum output (wider swing) signal can be obtained (and easily rectified if needed). Bias at a voltage other then near 0v could overload the op-amp's output when using a high gain.

The bias resistors are very high values because the diode current is very small, (the effective diode output impedance is also very high). Low value resistors would swamp out the diodes signal. The higher values here also help to keep the bandwidth wide.

The diode does not operate as a simple diode, it puts out a fairly linear reverse current related to the intensity of light hitting it.

Using AC pulses allows for an easy separation of the ambient light signals (at or near DC) and the desired LED light signal (high frequency pulses).

The coding of the output signal is such that a stronger reflection coming from the LED (as when moving closer to an object) gives a higher AC signal at the output. The signal could be rectified to get a DC voltage related to the reflected light level.

Some additional photo diode reference: http://en.wikipedia.org/wiki/Photodiode

\$\endgroup\$
  • \$\begingroup\$ How does the right-hand portion of the top schematic work (the receiver)? Under normal conditions the diode will be reverse biased and the voltage across L1/R3 will be zero. When a light pulse is exposed to the diode then a voltage > V1 will be necessary for conduction. Where does this voltage come from? \$\endgroup\$ – sherrellbc Feb 14 '15 at 18:17
  • 1
    \$\begingroup\$ Because it is not a standard diode. A photo diode (in the photoconductive mode) leaks in reverse bias when hit with light. See: en.wikipedia.org/wiki/Photodiode \$\endgroup\$ – Nedd Feb 15 '15 at 13:28
  • \$\begingroup\$ Ah, thank you for the link. What is the purpose of L1? Also, what would you suggest as an appropriate gain for the amplifier? It seems like you suggest to bias the in input voltage (V+) with half the supply, why is this? The diode will either be conducting or not, so it seems that you would either get nominal output voltage (from the bias) or some higher voltage. Why would you not use a simple comparator stage such that you have either +Vcc or GND output? It would seem pretty simple at that point to measure the time of each pulse and decode the information. \$\endgroup\$ – sherrellbc Feb 15 '15 at 17:30
  • \$\begingroup\$ The comment field this is not a good place to create new questions. See notes added below the schematic. \$\endgroup\$ – Nedd Feb 16 '15 at 11:17
  • \$\begingroup\$ Feel free to up vote the answer, (or the question) if you think it is worth while... \$\endgroup\$ – Nedd Feb 16 '15 at 11:42
1
\$\begingroup\$

Most decent-quality IR Beam-Break detectors do a bunch of things to improve their operation

1) Good Quality IR filter in front of the detector. These block visible light but pass IR through to the detector.

2) High-gain ac-coupled amplifier with AGC and band-pass filter that matches . . .

3) modulated light source

4) physical light-pipe. Basically, mount the sensor at the end of a pipe or tube so as to allow light from only the front to reach the detector.

5) Lens and parabolic reflector to improve both the range of the light source as well as the sensitivity of the detector.

We used to build high-speed photo beam-break systems intended for use on roads having low under-pass bridge decks or tunnels. The beam-break detector was situated several hundred yards before the exit that allowed an alternate route for vehicles having a high load, along with the flashing warning lights. The spec that we had to meet was to detect the wide edge of a 2x4 stick moving at highway speeds (110 Km/h). We used all of the above tricks that allowed this to work in all weather conditions except blinding snow storms or blizzards. Those weather conditions caused our beam-break sensors to indicate that a high load was present but there wasn't anything that we could do about it.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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