Need to produce fake sunlight: power question

[Paul Ciszek]

I need to produce fake sunlight over a small area--maybe a square 20cm
on a side--on a low budget. I have been told that ELH projector bulbs
make a good first approximation to sunlight on the cheap. I am trying
to make an estimate of how many of them I need to achieve sunlight
intesnsity. I tried to work it out based on the following numbers:

Sunlight = 100,000 lumens/m^2
Halogen bulbs yield 16 to 24 lumens/W
That's 6250 to 4167 W to produce 1m^2 of sunlight
250 to 167 W to illuminate my 20cm square

That just seems awfully low; am I missing something?

 

[[email protected]]

I need to produce fake sunlight over a small area--maybe a square 20cm
on a side--on a low budget. I have been told that ELH projector bulbs
make a good first approximation to sunlight on the cheap. I am trying
to make an estimate of how many of them I need to achieve sunlight
intesnsity. I tried to work it out based on the following numbers:

Sunlight = 100,000 lumens/m^2
Halogen bulbs yield 16 to 24 lumens/W
That's 6250 to 4167 W to produce 1m^2 of sunlight
250 to 167 W to illuminate my 20cm square

That just seems awfully low; am I missing something?

What is the "fake sunlight" for? Have you checked the prices on
those special bulbs for growing plants indoors?

 

[Morris Dovey]

Paul Ciszek wrote:
| I need to produce fake sunlight over a small area--maybe a square
| 20cm on a side--on a low budget. I have been told that ELH
| projector bulbs make a good first approximation to sunlight on the
| cheap. I am trying to make an estimate of how many of them I need
| to achieve sunlight intesnsity. I tried to work it out based on
| the following numbers:
|
| Sunlight = 100,000 lumens/m^2
| Halogen bulbs yield 16 to 24 lumens/W
| That's 6250 to 4167 W to produce 1m^2 of sunlight
| 250 to 167 W to illuminate my 20cm square
|
| That just seems awfully low; am I missing something?

Maybe. Why not use the real thing?

 

[Paul Ciszek]

Paul Ciszek wrote:
| I need to produce fake sunlight over a small area--maybe a square
| 20cm on a side--on a low budget. I have been told that ELH
| projector bulbs make a good first approximation to sunlight on the
| cheap. I am trying to make an estimate of how many of them I need
| to achieve sunlight intesnsity. I tried to work it out based on
| the following numbers:
|
| Sunlight = 100,000 lumens/m^2
| Halogen bulbs yield 16 to 24 lumens/W
| That's 6250 to 4167 W to produce 1m^2 of sunlight
| 250 to 167 W to illuminate my 20cm square
|
| That just seems awfully low; am I missing something?

Maybe. Why not use the real thing?

The easiest way to bring sunlight into the lab, on demand, where and
when I want it, is to fake it.

Sure, Colorado has 300+ days of sunshine a year, so there are ample
opportunities to test stuff outdoors. On the other hand, the late
afternoon thundershowers can be a real @#$%^&*!.

For *serious* testing, there are labs that do this sort of thing, but
for development purposes I just need "close enough" fake sunlight.

 

[Don Klipstein]

The easiest way to bring sunlight into the lab, on demand, where and
when I want it, is to fake it.

Sure, Colorado has 300+ days of sunshine a year, so there are ample
opportunities to test stuff outdoors. On the other hand, the late
afternoon thundershowers can be a real @#$%^&*!.

For *serious* testing, there are labs that do this sort of thing, but
for development purposes I just need "close enough" fake sunlight.

I ask that you say what you need the "fake sunlight" for or state
specific requirements of the "fake sunlight". That way some of us here
may have some ideas as to how to make fake sunlight that is "close
enough".

- Don Klipstein ([email protected])

 

[Paul Ciszek]

I ask that you say what you need the "fake sunlight" for or state
specific requirements of the "fake sunlight". That way some of us here
may have some ideas as to how to make fake sunlight that is "close
enough".

I need to measure the transmission (and possibly later, reflection)
of certain materials at those wavelengths that account for most of
the power in sunlight--near UV to about 2um. I don't need the source
to mimic the spectrum of sunlight per se, so long as there is enough
light present at all these wavelengths to measure transmission fraction
at each wavelength. Originally, I didn't see any need to reach the
intensity of sunlight, since I was just taking the ratio of two
measurements, and I would have been happy with a halogen work light
as my broad spectrum source. But now they want the tests to include
things that actually respond to sunlight, and so I may need to match
the actual intensity of sunlight. I still figure that a black body
with a hot enough color temperature should be good enough as far as
matching the spectral distribution goes, and the references I have
consulted seem to indicate that glass is still transparent at 2um.
I have not had any luck getting reference spectra for light bulbs
that goes outside the visible range; does anyone know if ELH bulbs
emit that far into the IR? We do not yet have the IR spectrometer
that would allow me to check that directly; I want to take visible
spectra in the meantime and be ready to start taking the IR data
as soon as we get the spectrometer, so I would rather get the light
source right the first time around.

I have heard that some halogen bulbs are coated to reflect the IR
inward, keep the filament hotter with less electricity, and therefore
improve the visible lumens per watt. While this is a good idea from
a lighting standpoint, it would throw a wrench into my plan for a
cheap artificial sun.