Shop OBEX P1 Docs P2 Docs Learn Events
Laser-Free Reflector Sensor Uses IR — Parallax Forums

Laser-Free Reflector Sensor Uses IR

ercoerco Posts: 20,256
edited 2014-11-08 22:45 in General Discussion
I still love my Hamamatsu S6986 laser sensors, but I've been playing with a safer, laser-free alternative. The Sharp IS471F uses a simple IR LED as a short range proximity sensor (3-4 inches). I found that by adding a small lens to collimate the IR beam, I could detect Scotchlite reflective tape from 3-4 feet away. I have more experimenting to do using bigger lenses (ordered) to possibly increase the range. The sensors are simple to use, cheap ($3) and in abundant supply from Junun. Video is self-explanatory. I'll have a Boebot navigating between reflectors soon.

For comparison, here's my Terminator robot navigating and targeting using the laser sensor: https://www.youtube.com/watch?v=dZTfIyrRne0

Comments

  • Martin_HMartin_H Posts: 4,051
    edited 2014-09-14 15:12
    Wait just a minute here. You convince us to bulk buy lasers and now you're trying laser alternatives?! Plus tips about safety from a man who's Boe-bot is outfitted with a flame thrower?

    I only want to see gratuitous laser projects from now on.
  • ercoerco Posts: 20,256
    edited 2014-09-14 15:47
    It's the old Bait & Switch, Baby Boy. :)
  • TtailspinTtailspin Posts: 1,326
    edited 2014-09-14 17:54
    Where did you get the lens? is it adjustable? is it just a hollow tube with a lens on top? will it chop lettuce and dice tomatoes? Do tell.

    The whole assembly looks like it could be made water resistant too. :thumb:





    -Tommy
  • ercoerco Posts: 20,256
    edited 2014-09-14 18:40
    I had several small convex lenses laying about, so I used those. No adjustment, the focal length dictates the tube length. I built & tested with a 5mm red LED to optimize focus and then swapped in the IR LED.

    A bigger diameter lens might increase range, so I have ordered a few to try. Once I arrive at a favorite lens & geometry combination, I'll share.

    It may chop lettuce and dice tomatoes, but I will have to test that after the current round of cold fusion and perpetual motion tests, which the sensor module is passing with flying colors. :)
  • ercoerco Posts: 20,256
    edited 2014-09-22 13:46
    Here's an inexpensive Ebay proximity sensor with some tracking & targeting potential. It uses 38 kHz, which is good and bad. You could also use this to pick up IR remote signals (good), and any 38 kHz IR remote signals will give false proximity readings (bad). Nonetheless, cheap and easy to get. Lenses might further improve range & directionality.

    I might have to get into the sensor business. I like this stuff.

    $1.90 now: http://www.ebay.com/itm/For-Arduino-New1pc-Infrared-IR-Sensor-Obstacle-Avoidance-Sensor-Smart-Car-/131202170633
  • YanomaniYanomani Posts: 1,524
    edited 2014-09-22 20:15
    Hi erco

    The specs shown at Ebay's product page are kind of a terrible translation from chinese to english. At least far worse than mine, from portuguese (sincerely hoping that's the case).:nerd:

    As for my current understanding, reading thru the specs, the enable signal (EN) could be used (after removing the green jumper cap), to control overall sensor activity. Just about (EN) operation, I'm curious to know the following:

    - Could the driver circuit, that flashes the IR transmitter, be turned off by controlling (EN), without disturbing its receiver passing sensor info to the connector?

    If this is the case, and by the fact they are so inexpensive, you could use two at a time, mounted atop or aside each other.

    Tweaking the transmitter output frequency by adjusting the trimpot, you could replace one of the detectors in a pair by another one, of a different frequency.

    Since they have an enable signal and if it works the way I believe it does, you could control which one, none or both are active each time, without disturbing the receptors, thus allowing the processor to detect incoming spurious signals, and select the one that is experimenting less interference, if any at all.

    This could be used to defeat IR remotes from the "AVOID USING NEARBY" list.

    Yanomani
  • ercoerco Posts: 20,256
    edited 2014-09-22 20:38
    Finally somebody "gets" me! :)

    I believe what you have said about the enable function is true. I'll test tomorrow and get back to you.

    I have tried 56 kHz IR receivers, but they still get blasted by 38 kHz IR TV remotes (very strong signal output). This is a much weaker signal, so it might be possible to mix frequencies. Nice idea.
  • YanomaniYanomani Posts: 1,524
    edited 2014-09-22 21:15
    I'll be here!

    I'm also curious about the SMD IC staying atop of the circuit board.

    Oh GOD! A simple schematics worth a thousand fried fingers and a bacon smelly shack!

    Yanomani
  • ercoerco Posts: 20,256
    edited 2014-09-23 00:50
    Confirmed, the signal LED lights when an external IR source is detected, even when the ENable jumper is removed. That apparently shuts down the onboard LED oscillator, possibly a power saving option.

    Looks like the IR receiver is unusually low gain, though. Probably OK to use with an IR remote, but it only picked up my low-output IR beacons from about 5 feet away.

    Edit: the IC on the board a the venerable NE555. How many other IC's from 1971 are still in everyday use?
  • YanomaniYanomani Posts: 1,524
    edited 2014-09-23 19:19
    Hi erco

    Hmmmm, five feet? This means two robots could gather in a comms link, says about more that 10 feet apart, considering the reflective material losses?

    Were those measurements done with or without the use of the lenses you have mentioned before?

    As for the 555 and many other grey haired ones, they are almost everywhere, if not explicitly, like in the IR led flasher circuit, at least embodied as IP (or not, who knows for sure) you use to deal with, everyday.

    One more question:
    If one of the trimpots is used to adjust the center frequency of the transmitter, can't the other one be used to control driver's strength?

    Also occurs to me that, if the 555 could withstand the extra current, you could parallel another IR emitter, at the expense of spreading a bit the output signal, due to inherent difficulties of alignment, but since power is summed, although not doubled at the receiving end, it could lead to a better overall performance.

    To spare some power, after you verify its performance, you could remove both visible leds and let its output to be monitored by the target cpu.
    As a final hack, you could scratch the frequency adjust trimpot and part of the oscillator circuit, letting to the cpu the burden of frequency control, at the cost of one more pin to connect, but, gaining the extras of fine tunning the system for maximum sensor response, and sure, you could craft a beautiful PWM buffer around 555's circuit and let the cpu do all the adjustments for you.

    Yanomani
  • rosco_pcrosco_pc Posts: 464
    edited 2014-09-23 23:21
    Yanomani wrote: »
    Oh GOD! A simple schematics worth a thousand fried fingers and a bacon smelly shack!

    These ebay modules seem to be based on this one: http://www.instructables.com/id/Reliable-and-low-cost-IR-Proximity-sensor/
  • YanomaniYanomani Posts: 1,524
    edited 2014-09-24 15:21
    rosco_pc

    Much thanks! I've followed the link and found the schematic.
    They had reworked it a bit, to include the frequency adjustment trimpot, the (EN) control end sure, the ubiquitous PLED.

    Yanomani
  • ercoerco Posts: 20,256
    edited 2014-09-27 11:29
    You can certainly make your own per if you like. Just be aware that the article specifies a rare and obsolete TSOP1738 IR sensor which is continuous signal compatible. Most other sensors are not, and will not work in this application. That's why I made a big deal about stocking up on TSOP4038 sensors here.
  • ercoerco Posts: 20,256
    edited 2014-11-08 22:45
    Factory rejects on Ebay! More precisely, misassembled sensors. Whooda thunkit?

    I have bought eight of the sensors I demoed in post #6. Two can be adjusted to work fine, but the others don't, no matter how I twirl the two adjustment pots. Could not get the output frequency anywhere near 38 kHz. Just today I took a closer look and the two pots are switched on all of my sensors units as compared to photos at http://www.dx.com/p/ir-infrared-sens...7#.VF6j-TwtBIc and http://www.amazon.com/NEXT-Infrared-.../dp/B00NPZ6RCC . The 103 pot goes by the GND pin and the 202 pot goes by the EN pin. Swapping them back (obviously) restores proper function. Printed on the PCB is "10K" where the 103 pot belongs, and "5K" where the 202 belongs. Check yours before dismissing and tossing!

    The thumbnail photo of my video shows the incorrect parts placement in post #6.

    The 103 pot controls the 555's output frequency to drive the IR LED, which should be adjusted to ~37.5 kHz to match the IR receiver module frequency. The 202 pot is a power output control for the IR LED which can be adjusted to trigger at different ranges. Once you swap pots and adjust, it's a nice little sensor.

    As I mentioned previously, the IR receiver module is unusual in that it is continuous signal compatible and low gain for this application.
Sign In or Register to comment.