Connect with us

Re: Garage Door Safety Sensors

Discussion in 'Electronic Design' started by Spehro Pefhany, Nov 27, 2011.

Scroll to continue with content
  1. I think they're just simple modulated IR beams with integrated IR
    demodulators (similar to IR remote control receivers), such as this
    one:

    http://www.biltek.tubitak.gov.tr/gelisim/elektronik/dosyalar/3/IR_sfh506.pdf

    All the rest of the smarts resides in a microcontroller.


    Best regards,
    Spehro Pefhany
     
  2. Winston

    Winston Guest

    Spehro Pefhany wrote:

    (...)


    Short answer: It's like a 'one wire' protocol.

    Long answer:
    The 'transmitter' and 'receiver' are parallel-connected to
    a medium-pullup 5.5V DC source. The transmitter pulses an
    IRLED at a 154 Hz rate, 7.7% duty cycle.

    The receiver (probably just an IR filtered phototransistor)
    receives the pulses and pulls the DC source down to nearly
    zero volts in response to each received
    IR pulse. The microcontroller in the opener decides if it
    is getting the pulses sourced by the transmitter.

    Blockage of the light will result in a continuous 5.5 V
    present on the two wires. Continuous 0 V or 5.5 V is
    considered physical path interference and will prevent
    the door from closing unless the hard switch in the garage
    is held closed continuously.

    Clever!

    --Winston
     
  3. I think you'll find the transmitter supplies a pulsed IR light carrier
    in the tens of kHz range to match the reciever.

    It would be interesting to see if the carrier is pulsed or not without
    the receiver- too bad he didn't hook a photodiode to his 'scope. Maybe
    someone has this type of garage door opener and can take a
    measurement?

    The receiver has an integrated receiver module inside. It pulls its
    own power supply down (and the transmitter's power supply) when the
    carrier is detected.

    (but presumably there's a diode + cap in each module to keep at least
    the receiver alive for a few msec. There has to be current limiting in
    the base unit- maybe a zener shunt regulator and pullup to +24 or
    whatever.



    Best regards,
    Spehro Pefhany
     
  4. Okay. patents.

    US5465033 illustrates an early higher voltage implementation that has
    the parallel transmitter/receiver configuration.

    Looks like US6906307 is pretty much what is used now- the transmitter
    circuitry is in one module and the photo detector, amplifier, AGC,
    detector is in another.

    Looks like the transmitter sends continously so long as power is
    applied (and the beam is not received in the parallel configuration).

    It _is_ a clever arrangement- sort of a windowed watchdog timer that
    depends on the modulated optical path for feedback. Very, very
    unlikely to accidentally fail in a state which indicates an unblocked
    door, despite using cheapie parts in most spots (and a few
    redundant-for-reliability parts around the micro).



    Best regards,
    Spehro Pefhany
     
  5. Winston

    Winston Guest

    I imagine that Chamberlain and Genie, etc. use different frequencies.
    The patent you cite (US5465033) indicates 1400 Hz in one embodiment.
    (Column 5, line 39, 46) So apparently they split the difference. :)

    I *do* see 30 KHz tossed out as a possible frequency in US6906307,
    however. The pulse repetition rate does not matter, as long as
    the respective circuits are adjusted to accommodate it, IMHO.

    I'm interested, too.

    'Hardwired single-frequency RFID' is how I think of it.

    The diode and cap you mention popped into my head as well.
    I see they use the phrase 'one wire' to describe the
    design. Where did I hear that before? :)
    I suspect we will find that the transmitter sends it's
    pulses regardless of whether the pulses are received.
    At the risk of revealing the blindingly obvious,
    a broken wire to either sensor puts the unit in fail-
    safe mode and forces it to act as if an obstruction
    is in place. That is nifty IMHO.

    Back atcha, Speff.

    --Winston
     
  6. Joerg

    Joerg Guest

    Western solution: Bolt a 2-by-4 onto the floor, with the upper edges
    rounded off. Done :)

    [...]
     
  7. Joerg

    Joerg Guest

    Paint it fire engine red and write on it "Watch your step - Ver su paso".

    Easy. Drill a couple of new holes, re-arrange, patch old holes :)
     
  8. Winston

    Winston Guest

    Jim Thompson wrote:

    (...)
    You meant to say 'unlimited moxie' yes? :)
    I agree that would heat things up a bit.

    There is a 'constant moxie controller' powering
    the transmitter. In US5465033 it is set up to limit
    at 62.5 milli moxies.

    'Works great!

    --Winston
     
  9. Guest

    Hang a fishing practice weight (rubber plumb bob looking thing) on
    monofilament fishing line from the ceiling so it lines up with the center of
    the windshield (mirror mount is a nice target).
     
  10. Careful- I think there are at least two kinds of 2-wire sensors. One
    seems to run on high voltage and one on ~6V.

    Replacement parts refer to those with red and those with green LEDs
    being incompatible.


    Best regards,
    Spehro Pefhany
     
  11. Yup, you nailed it!
    That is possible, but it would not be as good, IMHO. More complex and
    it would mean that the millisecond pulses are generated inside the
    receiver so an internal failure after the demodulator could cause a
    false "no obstruction" signal. Easy enough to test if someone had the
    bits- just use independent current-limited* supplies for each half.
    Yup, only two connections to the box and no combination of reversed,
    shorted, broken or damaged sensors can plausibly cause a false
    positive (or cause any damage or even much stress to the bits thanks
    to the current-limited supply and diodes).
    * I'd build a little unbypassed current/voltage-limited supply rather
    than use a bench supply neat- there's always(?) a largish e-cap across
    the output terminals in the latter, and that could kill the receiver
    even if the supply limits the steady-state current. A resistor and
    6.2V zener should suffice for bench testing.


    Best regards,
    Spehro Pefhany
     
  12. legg

    legg Guest

    My solution was a ping pong ball suspended fro the roof. If it hit the
    windshield, she's a little closer than she prefers to be.

    No batteries, no wiring.

    RL
     
  13. Guest

    Or a tennis ball. A bright yellow one.
     
  14. Just what I am guessing from the patents I previously referenced and
    the video. One type is (from one patent and the video) six-ish volts,
    but I'm pretty sure there's a higher voltage type out there.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-