Making Power measurements from the Sunlight?

[Roger Breton]

Can I use a radiometer like a Tektronix J16 to make power measurements from
the sunlight? In Watts?

Roger Breton

 

[Victor Roberts]

Can I use a radiometer like a Tektronix J16 to make power measurements from
the sunlight? In Watts?

Roger Breton

What aspect of sunlight are you trying to measure? Most of the probes
for the J16 are photopically corrected so they will give lumens or
candelas per M^2. The J6502-A/J6512-A probe is an irradiance probe
that reads in mw/M^2. However, it is flat from 450nm to 950nm and down
50% at 400nm and 1030nm, so it will not read all the energy emitted
by the sun. The upper limit without use of a neutral density filter
seems to be 2mW/cm^2.

I don't think the J16 is the right instrument for measuring the power
density of sunlight at the surface of the earth, if that is what you
are trying to do. I would suggest that you use a broadband radiometer.

 

[Roger Breton]

Can I use a radiometer like a Tektronix J16 to make power measurements from

What aspect of sunlight are you trying to measure? Most of the probes
for the J16 are photopically corrected so they will give lumens or
candelas per M^2. The J6502-A/J6512-A probe is an irradiance probe
that reads in mw/M^2. However, it is flat from 450nm to 950nm and down
50% at 400nm and 1030nm, so it will not read all the energy emitted
by the sun. The upper limit without use of a neutral density filter
seems to be 2mW/cm^2.

I don't think the J16 is the right instrument for measuring the power
density of sunlight at the surface of the earth, if that is what you
are trying to do. I would suggest that you use a broadband radiometer.

What I am trying to do, as usual, is to push the boundaries of my ignorance.
As I am studying radiometry these days, I read up on climatology and how the
sun is radiating its power to the earth (I had no idea there was such a
thing called 'the solar constant'!). I figure if I can understand how to
measure the sun radiating power then I'll be able to generalize the concepts
to other sources of radiation, namely artificial light sources. All colors
start with light and to study light I need to understand the concept of
radiant energy (eventually spectral radiant energy). These are not trivial
physical concepts to grasp. I figure my best bet, right now, is to study how
to measure radiant energy (flux), because once I understand that much then
moving to photometry (and later coloriimetry) is just crossing into V-lamba
territory -- I'm probably oversimplifying to death (?). I often see these
Tektronix J16 instruments come up for sale, used, on eBay for dirt cheap. I
am not interested in their photometric measuring capabilities but in their
radiometric capabilities which I did not know existed until yesterday. Hence
my post. To adequately measure the sun radianting power I probably need and
irradiance measuring device which will give me Watts/m2 or microWatts/cm2.
But I am not interested in the measurement of how much power fall per
unitary area at the stage of learning radiometry I am (I know I eventually
will) but how to measure *raw* radiating power from a source, in Watts, and
how much power per unit time this source radiates, in Watts/sec. How can I
get my feet, and those of my students, in this area whithout breaking my
bank (which is already half broken by now)?

Regards,

Roger Breton

 

[Roger Breton]

The J6502-A/J6512-A probe is an irradiance probe

that reads in mw/M^2. However, it is flat from 450nm to 950nm and down
50% at 400nm and 1030nm, so it will not read all the energy emitted
by the sun.

By the way, what is wrong with it 'being flat'?

Roger Breton

 

[Victor Roberts]

By the way, what is wrong with it 'being flat'?

Roger Breton

Flat is good. I should have said "... flat ONLY from 450nm to 950nm."
I am not an expert on solar radiation, but according to one reference
I found on the Web, the solar spectrum at the surface of the earth
ranges from 305nm to 2500nm. (This same reference tabulates the solar
radiation, after absorption by the atmosphere, out to 4045nm, but
there is very little energy between 2500nm and 4045nm.)

You need a radiometer that is has equal response to all wavelengths
from about 305nm to 2500nm to correctly measure the solar radiation at
the surface of the earth. The Tektronix probe I listed above does not
measure energy over this full wavelength range and will therefore
measure only a potion of the solar spectrum.

 

[Victor Roberts]

But I am not interested in the measurement of how much power fall per
unitary area at the stage of learning radiometry I am (I know I eventually
will) but how to measure *raw* radiating power from a source, in Watts, and
how much power per unit time this source radiates, in Watts/sec. How can I
get my feet, and those of my students, in this area whithout breaking my
bank (which is already half broken by now)?

Well, you can't measure the "raw" radiating power of the sun, at least
not with any instrument made today.

First of all, to measure the raw solar radiation you would have to
place your radiometer above the atmosphere. So, unless you have good
connections at NASA or the Russian, European of Japanese space
programs, you have to settle for measuring the sun through the earth's
atmospheric filter.

Second, even if you could get above the atmosphere you would need to
completely surround the sun with detectors to DIRECTLY measure the
total raw radiation of the sun. This is rather impractical, to say the
least, and is even impractical for most conventional light sources.
For the sun, you can assume that the radiation is isotropic, measure
the power per unit area at your detector, and then multiply that
result by the area of the virtual sphere defined by your
sun-to-detector distance.

To measure the total output of conventional light sources they are
most often placed inside an integrating sphere that takes the
non-uniform output of the source, and through multiple diffuse
reflections, distributes this energy uniformly over the inner surface
of the sphere. A measurement is made with a small detector that can
then be converted to total radiation by multiplying the energy
received by the ratio of the sphere area to the detector area.

In practice, however, the measurement is not made this way due to the
fact that the surface of the sphere is not a perfect reflector. The
sphere is usually calibrated using a lamp that has been measured and
certified by the National Institute of Standards and Technology (NIST)
and then the output of the subject lamp is determined by comparing the
reading of the detector when the NIST calibration lamp is operating to
the reading when the subject lamp is operating.

Or ... instead of using an integrating sphere, you can measure the
light at each position around the subject lamp using a goniometer (in
a black room) and the summing all the readings.

 

[Bill Kaszeta / Photovoltaic Resources]

What I am trying to do, as usual, is to push the boundaries of my ignorance.
As I am studying radiometry these days, I read up on climatology and how the
sun is radiating its power to the earth (I had no idea there was such a
thing called 'the solar constant'!). I figure if I can understand how to
measure the sun radiating power then I'll be able to generalize the concepts
to other sources of radiation, namely artificial light sources. All colors
start with light and to study light I need to understand the concept of
radiant energy (eventually spectral radiant energy). These are not trivial
physical concepts to grasp. I figure my best bet, right now, is to study how
to measure radiant energy (flux), because once I understand that much then
moving to photometry (and later coloriimetry) is just crossing into V-lamba
territory -- I'm probably oversimplifying to death (?). I often see these
Tektronix J16 instruments come up for sale, used, on eBay for dirt cheap. I
am not interested in their photometric measuring capabilities but in their
radiometric capabilities which I did not know existed until yesterday. Hence
my post. To adequately measure the sun radianting power I probably need and
irradiance measuring device which will give me Watts/m2 or microWatts/cm2.
But I am not interested in the measurement of how much power fall per
unitary area at the stage of learning radiometry I am (I know I eventually
will) but how to measure *raw* radiating power from a source, in Watts, and
how much power per unit time this source radiates, in Watts/sec. How can I
get my feet, and those of my students, in this area whithout breaking my
bank (which is already half broken by now)?

Regards,

Roger Breton

The measurement of solar radiation is undertaken by several organizations.
Environment Canada used to have an extensive program, but I have not
heard much about it in recent years.

Some starting links:

http://www.eere.energy.gov/pv/pvmenu.cgi?site=pv&idx=1&body=lightsun.html

http://rredc.nrel.gov/solar/

http://solardata.uoregon.edu/

Bill Kaszeta
Photovoltaic Resources Int'l
Tempe Arizona USA
[email protected]

 

[Roger Breton]

Victor-

Flat is good. I should have said "... flat ONLY from 450nm to 950nm."
OK.

You need a radiometer that is has equal response to all wavelengths
from about 305nm to 2500nm

Is the equal response capabilities of the radiometer itself or the detector?

Roger Breton

 

[Roger Breton]

First of all, to measure the raw solar radiation you would have to

place your radiometer above the atmosphere. So, unless you have good
connections at NASA or the Russian, European of Japanese space
programs, you have to settle for measuring the sun through the earth's
atmospheric filter.

No, I am afraid I don't have such useful connections

Second, even if you could get above the atmosphere you would need to
completely surround the sun with detectors to DIRECTLY measure the
total raw radiation of the sun. This is rather impractical, to say the
least, and is even impractical for most conventional light sources.

How about, theoretically, enclosing the sun in an integrating sphere: would
I then be able to measure through a port like you describe below?

For the sun, you can assume that the radiation is isotropic,

Isotropic as in 'equal intensity in all directions'? I've many times
encountered that word but I never clearly understood its meaning.

 measure
the power per unit area at your detector,

Please correct me if I am wrong but doesn't measure the 'power per unit area
(of the detector)' equate to measure the irradiance? E.g. Watts/m2

 and then multiply that
result by the area of the virtual sphere defined by your
sun-to-detector distance.

The area of the virtual sphere? You mean to infer the sun's irradiance from
a single power measurement and then extrapolate that measurement to the
whole surface of the sphere to find the total power of the sun irradiance
within a certain radius around it?

To measure the total output of conventional light sources they are
most often placed inside an integrating sphere that takes the
non-uniform output of the source, and through multiple diffuse
reflections, distributes this energy uniformly over the inner surface
of the sphere. A measurement is made with a small detector that can
then be converted to total radiation by multiplying the energy
received by the ratio of the sphere area to the detector area.

Well, that seems exactly like what I was trying to describe above, just more
elegantly. So the ratio of the sphere area to the detector area multiplied
by the power measured on the detector area gives total irradiance on any
sphere surface? That is in theory, I gather from reading below.

In practice, however, the measurement is not made this way due to the
fact that the surface of the sphere is not a perfect reflector. The
sphere is usually calibrated using a lamp that has been measured and
certified by the National Institute of Standards and Technology (NIST)
and then the output of the subject lamp is determined by comparing the
reading of the detector when the NIST calibration lamp is operating to
the reading when the subject lamp is operating.

OK, I understand the measurement is not made 'this way' because no sphere
are perfect reflector (even with Spectralon or other special material?).
Then, some other reference is used to infer the true measurement of power
from a given sphere. So this is an indirect method of measurement but one
that uses a 'standard'. But I wonder how NIST can determine their lamp
calibration in the first place if they too are stuck with sphere that are
not perfect reflectors?

Or ... instead of using an integrating sphere, you can measure the
light at each position around the subject lamp using a goniometer (in
a black room) and the summing all the readings.

Yeah, I once saw such an instrument at the National Institute of Standards
and Measurements in Ottawa, Canada.

Roger Breton

 

[Victor Roberts]

No, I am afraid I don't have such useful connections


How about, theoretically, enclosing the sun in an integrating sphere: would
I then be able to measure through a port like you describe below?

That is still not a direct measurement of the total energy, as I
define it. You are still measuring a fraction of the total energy and
then need to multiply by the ratio of the area of the sphere to the
area of your detector.

Isotropic as in 'equal intensity in all directions'? I've many times
encountered that word but I never clearly understood its meaning.
Yes.


Please correct me if I am wrong but doesn't measure the 'power per unit area
(of the detector)' equate to measure the irradiance? E.g. Watts/m2

Yes. "Irradiance, E † the density of radiant flux (power) incident on
a surface." from the IES Handbook, 9th Edition, 2000.

The area of the virtual sphere? You mean to infer the sun's irradiance from
a single power measurement and then extrapolate that measurement to the
whole surface of the sphere to find the total power of the sun irradiance
within a certain radius around it?

I don't understand what you have said

You measure the irradiance (watts/M^2) with your detector. Then, with
the assumption that the sun's radiation is isotropic, you calculate
the area of a sphere that has a radius equal to the distance of your
detector from the center of the sun. When you multiple your irradiance
measurement in watts/M^2 by the area of the sphere in M^2 you will get
watts of radiation from the sun, not irradiance. In the end the total
radiation will be independent of the place you decide to place your
detector. If you are further away, the irradiance will be lower but
the sphere will be larger and the result will be the same number of
watts, as it must be.

Well, that seems exactly like what I was trying to describe above, just more
elegantly. So the ratio of the sphere area to the detector area multiplied
by the power measured on the detector area gives total irradiance on any
sphere surface? That is in theory, I gather from reading below.

Not exactly. The ratio of the sphere area to the detector area
multiplied by the power measured on the detector area gives the total
POWER radiated by the source. Carry the units through and you will
see that the result is watts, not watts per unit area.

OK, I understand the measurement is not made 'this way' because no sphere
are perfect reflector (even with Spectralon or other special material?).
Then, some other reference is used to infer the true measurement of power
from a given sphere. So this is an indirect method of measurement but one
that uses a 'standard'. But I wonder how NIST can determine their lamp
calibration in the first place if they too are stuck with sphere that are
not perfect reflectors?

Good point. I am not sure, but I think they use a goniometer with a
calibrated detector, so there is no sphere reflection error. Perhaps
others can comment here.

 

[Victor Roberts]

Victor-

Is the equal response capabilities of the radiometer itself or the detector?

Roger Breton

If you have only a simple detector it must have equal response at all
wavelengths. If you have a detector connected to the output of a
spectrometer, then the sensitivity of the detector can be corrected
for each wavelength, assuming you know the detector sensitivity and
the spectrometer throughput as a function of wavelength.

 

[Petri Kärhä]

How about, theoretically, enclosing the sun in an integrating sphere: would
I then be able to measure through a port like you describe below?

There is a much simpler way to do this, not that accurate but much
cheaper. What I think you try to measure is related to "solar constant".
If you google that you find e.g. page
http://www.solarserver.de/lexikon/solarkonstante-e.html
that says that just outside the atmosphere the total irradiance produced
by sun is 1.37 kW / m². Now you can assume that the sun radiates with
equal power to all directions. Find somewhere the distance from sun to
earth and calculate the area of the sphere that this radius produces.
Then simply multiply the area with the given irradiance and you get the
total radiant power of the sun.

If you want more accuracy, you have to take into account sun-spots and
such things that change the geometrical distribution of the solar
radiation. You might try to find papers where they describe trends of
the solar constant. If you take the average of 11 years, this might give
a good average value.

As already pointed out, you can't measure the solar constant without
satellites, because the atmosphere blocks a lot of radiation. There are
however satellites that are monitoring the solar constant routinely and
trying to estimate trends in solar radiation. The data produced is not
top secret. The guys making the measurements are scientists and thus try
to publish their results regularly. Articles should be pretty
straightforward to find. By googling e.g. "fröhlich solar constant" you
can find some articles. (C. Fhröhlich is one guy I know is involved in
measurements of solar constant). The following article:
http://remotesensing.oma.be/solarconstant/articles/article2.html
describes one satellite project and gives a quite accurate value for the
solar constant.

If you still want to measure the solar constant you need a flat
radiometer. A pyroelectric radiometer might be a good choice. If you
can't measure in space, you need to take into account the transmittance
of the atmosphere. It may be complicated but not impossible. You should
find out the spectral behaviour of the blocking of the atmosphere. Then
you need to take into account the clouds. Well it gets complicated.
Calling NASA and asking them would maybe be the simplest solution.

Hope this helps.

Pete