• Measuring sunlight strength throughout the day with low-angle sensor

    From malna.polya@gmail.com@21:1/5 to All on Tue Oct 3 18:55:36 2017
    Hi All,

    I am working on a project: we have a light sensor with only around 40 degrees field of view. Our goal is to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different
    angle. We want to keep the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the rays but I am stuck, since one lens would only focus the parallel rays of different angles to different points in the focal plane - and our sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with up to 10 degrees which is not enough, since we want to measure from -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From JTS@21:1/5 to malna.polya@gmail.com on Wed Oct 4 10:29:51 2017
    On 2017-10-04 03:55, malna.polya@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40 degrees field of view. Our goal is to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different
    angle. We want to keep the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the rays but I am stuck, since one lens would only focus the parallel rays of different angles to different points in the focal plane - and our sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with up to 10 degrees which is not enough, since we want to measure from -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!



    Other posters have more experience and might comment on this, but I have
    got a feeling that the theorem of conservation of etendue makes what you
    want impossible. If you want to decrease the angular spread of a source
    (and this is what you want to do: a source emitting at different angles
    for different times is the same as a source emitting at different angles
    at the same time for what regards the theorem) then you have to increase
    the area which the light crosses. I think this is what is happening in
    the solution with a set of lenses.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From ggherold@gmail.com@21:1/5 to malna...@gmail.com on Wed Oct 4 06:16:47 2017
    On Tuesday, October 3, 2017 at 9:55:37 PM UTC-4, malna...@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40 degrees field of view. Our goal is to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different
    angle. We want to keep the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the rays but I am stuck, since one lens would only focus the parallel rays of different angles to different points in the focal plane - and our sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with up to 10 degrees which is not enough, since we want to measure from -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!

    Forest Mims has a whole bunch of stuff using led's to measure sky light. http://www.forrestmims.org/scientificresearch.html
    (I don't know if that's what you're trying to do.)

    How about a number of sensor's pointing in different directions,
    then keep track of the time of day. And do some math.

    Why don't you want to use a solar tracker?

    George H.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Phil Hobbs@21:1/5 to JTS on Wed Oct 4 10:28:26 2017
    On 10/04/2017 04:29 AM, JTS wrote:
    On 2017-10-04 03:55, malna.polya@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40
    degrees field of view. Our goal is to measure the strength of the
    sunlight without introducing too much of a bias based on the time of
    the day - when the sun shines at a different angle. We want to keep
    the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the
    rays but I am stuck, since one lens would only focus the parallel rays
    of different angles to different points in the focal plane - and our
    sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with
    up to 10 degrees which is not enough, since we want to measure from
    -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of
    the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!



    Other posters have more experience and might comment on this, but I have
    got a feeling that the theorem of conservation of etendue makes what you
    want impossible. If you want to decrease the angular spread of a source
    (and this is what you want to do: a source emitting at different angles
    for different times is the same as a source emitting at different angles
    at the same time for what regards the theorem) then you have to increase
    the area which the light crosses. I think this is what is happening in
    the solution with a set of lenses.

    Lenses make this difficult, because the image intensity wants to fall
    off like some power of the cosine of the field angle.

    The OP can do pretty well with a non-imaging concentrator run backwards,
    i.e. small end towards the sky. Of course it has to be smallish in
    order for the big end to fit the available detector, and there's still a
    cosine rolloff due to obliquity.

    Were it I doing it, I'd probably use a hexagonal cylinder made out of
    solar cells, with the axis tilted to take care of solar elevation.
    Those cells are pretty cheap, and the hexagonal shape introduces only a
    few percent error. Octagonal would be somewhat better.

    Shadow masks can fix up the remainder, at least in azimuth.

    The fun method is a quad cell, a shadow mask, and two RC airplane servos. ;)

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs
    Principal Consultant
    ElectroOptical Innovations LLC
    Optics, Electro-optics, Photonics, Analog Electronics

    160 North State Road #203
    Briarcliff Manor NY 10510

    hobbs at electrooptical dot net
    http://electrooptical.net

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From JTS@21:1/5 to Phil Hobbs on Wed Oct 4 22:38:27 2017
    On 2017-10-04 16:28, Phil Hobbs wrote:
    On 10/04/2017 04:29 AM, JTS wrote:
    On 2017-10-04 03:55, malna.polya@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40
    degrees field of view. Our goal is to measure the strength of the
    sunlight without introducing too much of a bias based on the time of
    the day - when the sun shines at a different angle. We want to keep
    the sensor stationary.


    (cut)


    Another concern is to find a solution which doesn't filter out much of
    the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!



    Other posters have more experience and might comment on this, but I have
    got a feeling that the theorem of conservation of etendue makes what you
    want impossible. If you want to decrease the angular spread of a source
    (and this is what you want to do: a source emitting at different angles
    for different times is the same as a source emitting at different angles
    at the same time for what regards the theorem) then you have to increase
    the area which the light crosses. I think this is what is happening in
    the solution with a set of lenses.



    The OP can do pretty well with a non-imaging concentrator run backwards,
    i.e. small end towards the sky. Of course it has to be smallish in
    order for the big end to fit the available detector, and there's still a cosine rolloff due to obliquity.



    Yes, excellent: you need to decrease the area from where you collect the
    light and you will be able to get light with the reduced angular
    distribution at the detector.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Mike S@21:1/5 to ggherold@gmail.com on Fri Oct 6 19:28:45 2017
    On 10/4/2017 6:16 AM, ggherold@gmail.com wrote:
    On Tuesday, October 3, 2017 at 9:55:37 PM UTC-4, malna...@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40 degrees field of view. Our goal is to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different
    angle. We want to keep the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the rays but I am stuck, since one lens would only focus the parallel rays of different angles to different points in the focal plane - and our sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with up to 10 degrees which is not enough, since we want to measure from -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!

    Forest Mims has a whole bunch of stuff using led's to measure sky light. http://www.forrestmims.org/scientificresearch.html
    (I don't know if that's what you're trying to do.)

    How about a number of sensor's pointing in different directions,
    then keep track of the time of day. And do some math.

    Why don't you want to use a solar tracker?

    George H.

    Rotate the sensor once every n minutes, plot the maximum readings over
    time, feasible?

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Helpful person@21:1/5 to malna...@gmail.com on Sat Oct 7 18:32:47 2017
    On Tuesday, October 3, 2017 at 9:55:37 PM UTC-4, malna...@gmail.com wrote:
    Hi All,

    I am working on a project: we have a light sensor with only around 40 degrees field of view. Our goal is to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different
    angle. We want to keep the sensor stationary.

    I was thinking about using a set of lenses to change the angle of the rays but I am stuck, since one lens would only focus the parallel rays of different angles to different points in the focal plane - and our sensor has a fairly small surface area.

    Some of the diffusers I have found were only changing the angles with up to 10 degrees which is not enough, since we want to measure from -75 to 75 degrees.

    Another concern is to find a solution which doesn't filter out much of the light spectrum.

    Does anyone have any ideas on how to solve this challenge?

    Thank you in advance!

    A simple solution is to use a series of fiber optics, each pointed at a different angle, say 15 degrees apart. They can then be bent to be parallel at their distal end where a lens can then image them onto your detector. The NA of the collection lens
    must be greater than the NA of the fibers.

    The best solution would be to use a single thick fiber. The system will need to be calibrated.

    http://www.richardfisher.com

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From whit3rd@21:1/5 to malna...@gmail.com on Mon Oct 9 16:38:18 2017
    On Tuesday, October 3, 2017 at 6:55:37 PM UTC-7, malna...@gmail.com wrote:

    I am working...to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different angle. We want to keep the sensor stationary.


    Well, the classical approach is to use a crystal ball, and a carved wooden screen.

    Really:

    <http://www.kuriositas.com/2011/02/campbellstokes-sunshine-recorder.html>

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Mike S@21:1/5 to All on Mon Oct 9 19:04:14 2017
    On 10/9/2017 4:38 PM, whit3rd wrote:
    On Tuesday, October 3, 2017 at 6:55:37 PM UTC-7, malna...@gmail.com wrote:

    I am working...to measure the strength of the sunlight without introducing too much of a bias based on the time of the day - when the sun shines at a different angle. We want to keep the sensor stationary.


    Well, the classical approach is to use a crystal ball, and a carved wooden screen.
    Really: <http://www.kuriositas.com/2011/02/campbellstokes-sunshine-recorder.html>

    Great post, thanks.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From glen walpert@21:1/5 to All on Tue Oct 10 22:45:43 2017
    On Mon, 09 Oct 2017 16:38:18 -0700, whit3rd wrote:

    On Tuesday, October 3, 2017 at 6:55:37 PM UTC-7, malna...@gmail.com
    wrote:

    I am working...to measure the strength of the sunlight without
    introducing too much of a bias based on the time of the day - when the
    sun shines at a different angle. We want to keep the sensor stationary.


    Well, the classical approach is to use a crystal ball, and a carved
    wooden screen.

    Really:

    <http://www.kuriositas.com/2011/02/campbellstokes-sunshine-
    recorder.html>

    Interesting approach, but for simple electronic monitoring with no moving
    parts how about the method used by photographic lightmeters for
    omnidirectional measurement, a spherical diffuser placed over the
    detector, such as:

    http://www.ebay.com/itm/Kenko-K-KFM-300-Spherical-Diffuser

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From whit3rd@21:1/5 to glen walpert on Wed Oct 11 04:40:20 2017
    On Tuesday, October 10, 2017 at 3:45:39 PM UTC-7, glen walpert wrote:
    On Mon, 09 Oct 2017 16:38:18 -0700, whit3rd wrote:

    On Tuesday, October 3, 2017 at 6:55:37 PM UTC-7, malna...@gmail.com
    wrote:

    I am working...to measure the strength of the sunlight without
    introducing too much of a bias based on the time of the day - when the
    sun shines at a different angle

    Well, the classical approach is to use a crystal ball, and a carved
    wooden screen.

    Interesting approach, but for simple electronic monitoring with no moving parts how about the method used by photographic lightmeters for omnidirectional measurement, a spherical diffuser placed over the
    detector,

    A good approach too, is an integrating sphere; basically, a white sphere
    with a porthole, or (to get fancy) two spheres, with a porthole into the
    first, and an aperture from the first to the second, with the sensor
    in the second.

    The weakness, is that the direction of the illumination can be
    edge-on to the circular hole, it isn't really as omnidirectional
    as you'd like. The hemisphere diffusers subtend pi *R*R as lit
    from their axis of rotation, but only pi * R*R/2 if the source
    is 90 degrees to that...

    The angle dependence for a flat window is as cos(theta), so you
    get 1% diminution after 8 degrees; a belt of hemispheres
    at 16 degree intervals (22 sensors) and a choose-whichever-is-brightest
    rule might do a good measurement job. If you
    need not only the plane of the ecliptic but the full 4pi steradians
    covered, it's up to a couple of hundred sensors.

    More practical, would be a couple of hundred light fibers, aimed all different directions,
    with end B coupled to a single, nonspherical plate of a sensor.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)