Connect with us

Wanted: Simple Sensitive Light Detector Circuit

Discussion in 'Electronic Design' started by Radio Man, Oct 8, 2004.

Scroll to continue with content
  1. Radio Man

    Radio Man Guest

    I need a sensitive light detector that i can place in front of the eyepiece
    a binocular. This detector should sound a "beep" whenever it detects light.
    I would also like it to be powered by 6 vdc or less.
  2. Ken Smith

    Ken Smith Guest

    Unless you super cool the binoculars, you could just wire the battery
    straight to the buzzer. There will always be some infrared light.

    What color light?
    How bright?
    Is it a focused image or a wide area you care about?
    Do you only care about a change in light level?
  3. Radio Man

    Radio Man Guest

    My interest is detecting "Starlight"
  4. Jim Miller

    Jim Miller Guest

    a very bright star such as vega well focused on a silicon photodetector only
    produces a few picoamps of signal.

    not an easy task...


    My interest is detecting "Starlight"
  5. Ken Smith

    Ken Smith Guest

    Is it starlight as in you are pointing it at a star?
    Are you looking for a variable star?
    Is it one specific star at a time or a general average?

    The light detector you want sounds quite simple.

    Chances are you will need some type of PIN diode detector, a
    transimpedance amplifier and a comparitor.

    A PIN diode can be had as a very small photovoltaic cell for quite a low
    price. For more money, you can get a better one.

    A transimpedance amplifier sounds scary but it isn't really. More on this
    a bit later.

    The comparitor can be a slight misuse of an op-amp.

    Go look up the LM324 or TL074 op-amp. Many quad op-amps all share the
    same pin-out. If you use a socket, you can change your mind about the
    op-amp by just plugging a different one in.

    Lets imagine we are going cheap. You can improve it later.
    Lets also say we use the TL074.

    First we need some voltages like this:

    / 100K
    +------!+\ U1A
    ! ! ----- V1
    \ --!-/ !
    / ! !
    \ -------
    / 5K
    +------!+\ U1B
    ! ! ----- V2
    \ --!-/ !
    / ! !
    \ -------
    / 100K

    Now the transimpedance section:

    ! 100K !
    ! !
    ----+--!-\ U1C !
    ! ! -------+-- Bright
    --- V1-!+/
    ^ D1

    Note: The photodiode D1 has a very small back bias on it. If D1 is one
    that is intended for photoconductive use, you can ground the end of it

    If light hits D1 "Bright" will go to a voltage greater than V1. The more
    light the higher the voltage.

    You may want to raise the 100K to 1Meg to get a higher gain.

    Now the comparitor:

    Bright ---/\/\/---!+\
    ! ---- BuzEn
    V(+) -----!-/
    ! !
    \ !
    / !
    100K \ !
    / !
    ! !
    / !
    10K \<---
    10T / This is a 10 turn pot.

    When "bright" goes above where you've set the pot, BuzEn goes positive.
    Now we just need to drive the buzzer. A power MOSFET is likely to be all
    you need here.
  6. Ken Smith

    Ken Smith Guest

    I assume he is just after "there are stars up there".

    If he is trying to detect just one star, matters are even worse than the
    few picoamps you suggest. The collection area he has is a very small part
    of a square meter.

    If he is gathering from a significant fraction of the sky, he may be able
    to do it with an op-amp like the TL074 so long as he is willing to adjust
    out the bias error and keep the temperature fairly stable.
  7. Jim Miller

    Jim Miller Guest

    my picoamp number was one calculated recently for vega and a 75mm collection
    area, 4pa actually for a detector matched to the visual response curve

    vega is a magnitude 0 star and there aren't many of those around. much more
    visible are second magnitude stars like polaris. each magnitude is a factor
    of ~2.5 dimmer than the previous.

    i think all he is going to detect with a wider field of view is light
    polution rather than stars.

  8. Ken Smith wrote...
    [ snip rest of interesting post with schematics ]

    Radio Man, chances are you'll need a detector that's much more
    sensitive than a PIN diode, such as either a photo-multiplier
    tube (PMT) or a low-dark-count avalanche photodiode (APD).

    APDs are very expensive, with their specialized power supplies.

    Photo-multiplier tubes are also potentially expensive, but you
    can get them for low prices on eBay. You'll also need a high-
    voltage power supply for the PMT, or you can make it yourself.

    In either case you may want to explore photon counting, rather
    than current measurement, as a way to capture the signal.
  9. Rich Grise

    Rich Grise Guest

    I agree. I live about 50 miles from downtown LA, CA, and even when
    visibility is unlimited, I can see about a dozen stars, and it usually
    turns out that 3 or 4 of them are airplanes. You have to drive about another
    50 miles to get a good night sky. Or maybe on top of Mt. Wilson or

  10. Steven Swift

    Steven Swift Guest

    One of the most sensitive (but still cheap) circuits is to put a
    phototransistor as one-half of a differential long-tailed pair. With
    an op-amp as bias stabiliziation, you can build an extremely
    sensitive circuit.
  11. Steven Swift wrote...
    That idea is so crazy, it just might work! :<)
  12. JeffM

    JeffM Guest

    I live about 50 miles from downtown LA
    Yeah. To see Halley's Comet
    most folks went below the cliffs of the Palos Verde Peninsula or to Big Bear.
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