Connect with us

need LARGE light meter

Discussion in 'Electronic Design' started by Rileyesi, Jan 30, 2004.

Scroll to continue with content
  1. Rileyesi

    Rileyesi Guest

    How about simply using a solar panel? I don't know if that will give you the
    frequency response you need. If it does, you can simply measure the voltage
    generated by the panel(s). That would solve your "averaging" problem as well.

    Just a thought.
  2. Chris Carlen

    Chris Carlen Guest

    One way to meter light over large areas is to image the area onto a
    smaller sensor, like a camera.
  3. I think what Jim was suggesting (below) is some optics that will collect
    light over your entire test area at once and focus it onto a small

    You could set up a white card of the test size you require and then some
    imaging optics (aka a camera) with an aperature that would only 'see'
    the surface of the card. The sensor would be placed at the image plane
    of the optics (where the film goes in a camera) and would be of a size
    related to the test area by the focal length of the optics.

    You can sample an area that is limited only by the size of white card
    you can aquire.

    The folks over on could probably help you figure out
    exactly what you'll need.
  4. A solar panel definitely integrates the light over the area.
    Currentwise only though, not voltagewise. So cells in series
    are not doing what you intend them.

  5. Hi Jim,
    There are several things you need to consider. First, your sensor will
    have some sort of 'field of view', i.e. the angle that it can see.
    Second, of course, is that it will have some sort of spectral response.

    To establish your sensor's field of view, put it in the center of your
    diffuser, and measure it's output with no additional illumination.
    Next, take a flashlight or other small source, and aim it at various
    places on the dome. Have the flashlight the same distance from the dome
    for each measurement, and measure the output from the sensor. From
    this, you will get a measure of angle versus intensity, and you will see
    what your sensor's actual field of view is. You will probably find that
    the sensor has a gaussian field of view, being more sensitive to the
    light directly in from of it.

    You can widen the field of view by using a lens to gather and focus the
    light from an area. You might think of using a box with a large, square
    diffusing panel, like from a flourescent light fixture, and a lens
    imaging this panel onto your sensor. Make sure that the intensity will
    not burn the sensor out. This lets you have the diffusing surface at a
    specific distance, and lets you focus the image of this diffuse light
    source directly on your sensor. You can also place small filters
    between the lens and your sensor if you need to correct your spectrum.

    Edmondson Engineering
    Unique Solutions to Unusual Problems
  6. Tim Shoppa

    Tim Shoppa Guest

    Often these meters come with two integrating filters... one is flat, the other
    is a hemisphere. The purpose of the hemisphere one is to give an
    appropriately weighted (over angles) reading for incident light metering.
    The flat one is a different weighting, with the predominant angle being
    forward instead of to the side, I think this is used for differnt (e.g.
    "flat") subject matter.

    An idea: you don't really need to integrate over all the whole square foot,
    do you? Sixteen or twenty-five sensors in a grid may be more appropriate
    for you... you could either do the averaging in software or hardware
    (software would let you vary the weighting for whatever your purposes are).
    At some point you have to stop adding sensors because it gets too expensive.
    Yes, you can add filters to force flatness as you wish. If you can
    find a set of Wratten filters on E-bay or similar, you'll be all set
    for experimenting.

  7. Jim

    Jim Guest

    I need to build a light meter that will average the intensity if visible
    light over a large area.....possibly a square foot or so.

    I have a number of light meters that use what I think is a "blue enhanced"
    sensor behind a white plastic translucent filter of about 2 inches or so in
    diameter. As far as I can tell the main purpose of this white filter is to
    provide an average of the light intensity that may be hitting different
    parts of it's surface. Since the sensors in all of these I own has an
    active area of around 2.5 x 2.5 mm and the sensor is located below and in
    the center of this white filter, I guess that the way it works is that the
    sensor "see's" the whole of the bottom surface of the white disk and
    responds to the diffuse light coming through, bouncing around, and whatever.

    What I need though is the equivelent that can do the same averaging trick
    over a much larger area. So far I have made a large diffuse filter by
    cutting a frosted white glass light fixture globe so that it is a true
    hemi-sphere. I think the diameter is around 10 inches. It's frosted on the
    inside, but I couldn't find one frosted on the outside which I would have
    preffered. Unless I find something better, this will be my diffuse filter.

    I am wondering about sensors now. I have a couple of types of blue enhanced
    sensors which have a voltage output, which are suppose to somewhat mimic the
    spectral response of the human eye. This is a bell shaped curve from
    approximately 400 to 700 nm. Two questions on this; first, since the
    primary purpose of this light meter it to measure the amount of sunlight
    that is reaching the plants on a forest floor (the light filtering through
    the canopy above), I would prefer that the spectral response of my sensor
    to be much more flat from 400 to 700 nm....., the photosynthesis active
    radiation there a sensor that is available that can do this?
    How about adding glass filters to a blue enhanced filter?

    Secondly, does one sensor mounted in the center really do the trick in
    "seeing" the whole of the inside of a hemi-sphere this size, or do I need to
    consider an additional circular "array" of 5 or 6 sensors mounted flat
    around the center one. If I do this with voltage output sensors, how do I
    connect them together? Does wiring them in parallel provide an "average"
    output of what all are sensing?

    I can buy a light meter on ebay or somewhere that I can salvage the
    electronics and the display from, and call the displayed readings "relative"
    and either create a lookup table or calibrate it with another light meter
    somehow, but for now this is a separate problem.

    Can someone give some ideas or suggections on the sensor issues and any
    other issues?

  8. Jim

    Jim Guest

    Well, I thought of that, and I even have a nice 7" circular panel that fits
    nicely under my frosted dome. Trouble is that it responds too well to
    radiation ouside the 400-700 nm range, and the response curve is far from
    flat, not even a decent bell curve.

    Also, would it REALLY provide an average?? I experimented with a couple in
    a dark room and a small 2 AA cell flash light illuminating "spots" of the
    panel from various distances.....can't say I was too impressed with the
    results. All the cells in mine are in series....maybe that's the

  9. Jim

    Jim Guest

    Right, but in talking with botanists who regularly do this, the many small
    area reading is less than desirable. For one thing they are very difficult
    to duplicate even at the same time at the same place, especailly when you
    are talking about uneven, subdued and dappled light patterns such as a
    forest floor. Get's very tedious.

    Even if I can get this 95 square inch metered area idea to work, there are
    still multiple measurements to be made........move and measure, move and
    measure....I guess d'd set up some sort of imaginary grid to cover any
    area. All that statistical stuff, yu know!

  10. Jim

    Jim Guest

    I have tried to get data sheets for every sensor that I have or am
    considering. Luckily, if they can be trusted the 1/2 angle is usually given
    in the specs, as well as a graphic showing what I take to be a "field of
    view" (roughly). I have wondered about this. In the case where only one
    sensor is used, and placed in the center of the hemi-sphere it should be
    about equidistant from all points on the sphere. I have a bunch of small
    (about 1/2") fisheye lenses that I salvaged from old smoke detectors that
    focus almost a complete 360 degree spherical view onto a sensor. They are
    plastic, and I am unable to determine if they filter any particular
    wavelengths. Is focusing the light on the sensor this way capitalizing on
    the gaussian field of view aspect of the sensor?

  11. Jim

    Jim Guest

    O.K., I think I get the idea. The white card would be placed flat on the
    ground, and assuming you could get the camera to see it without a person
    interfering with the incident light, the card's image would be compressed to
    the sensor (or film) size. Sound's like you could also just take a photo
    and have the advantage of post-processing with software if you wanted.
    Would a flat card be representative of a spherical view of the light, or
    would it be weighted too much toward that light coming from one general

    Another idea was suggested that sounds somewhat similar........put a fisheye
    lens (do they make fisheyes with 360 degree views?) on the camera, take a
    B&W photo and let Photoshop or such count pixels and do the processing
    (averaging or whatever). In use I guess the camera would be put on it's
    back on the ground pointing straight up, with the self timer set. Maybe
    even put a frosted dome over the fisheye??

    Ever heard of anything like that?

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