Sei sulla pagina 1di 14

Physics 227L

Optical Inverse Square Law Experiment

Purpose: To measure and confirm the inverse square law fall off of light intensity with distance
from a small, point like source of light over at least two orders of magnitude in optical power
(approximately 15 cm to 2 meters in distance). To become familiar with Excel log-log plots and
using a least square power law fit to data in order to extract the exponent. To become familiar
with making sensitive, high accuracy light power measurements over several orders of
magnitude using a frequency counter technique. To investigate and understand the origins of
experimental deviations from the inverse square law and greatly reduce or eliminate them.

Equipment: In order to be able to measure light intensity over several orders of magnitude with
low cost apparatus, we use a light to frequency converter chip, the TSL 237, mounted to an
optical post, combined with a digital multimeter with frequency counter (GW Instek GDM-396)
to function as a wide dynamic range optical power meter. This allows light power measurements
over more than 5 orders of magnitude, from 1Hz to 500 KHz, corresponding to light powers of
approximately 10-11 Watts (10 pW, picoWatts) to 5x10-6 Watts (5 W, microWatts). The chip is
calibrated directly in intensity, with a conversion factor of 2.3 kHz/(W/cm2) at 524nm. This
assumes the small lens on the chip of radius 0.90 mm is illuminated uniformly over its cross
sectional area. The typical frequency with no light present, the so-called dark frequency is 0.1
Hz at 25C. A light source with small aperture, a sufficiently large piece of black velvet to drape
as a hood over the Pasco light source to eliminate sources of stray light, an optical rail with
sliding mount for the optical sensor head, a meter stick, translucent scotch tape, tinfoil, toothpick
or other sharp point for making pinholes, and dark colored index cards are also needed. Photos
of the apparatus follow.

Left: Inverse square law apparatus showing light source and sliding optical sensor head on
optical rail. Sensor head connected to DMM used in frequency counter mode (GW Instek GDM-
396). Right: Closeup of optical sensor head with TSL237 chip centered on circular light output
from source.

KW S2011
Physics 227L

Assuming a measurement conversion factor of 2.3kHz/(W/cm2), and that the lightbulb radiates
as an isotropic point source, calculate the total optical power radiated by the light source. Is this
number reasonable given the rated (electrical) wattage of the light bulb (Pasco light source is a
Phillips 10 Watt, 12 Volt Halogen incandescent bulb)? (i.e. how can you account for the
discrepancy?) Hint: Pick a distance and corresponding frequency, convert the frequency to
intensity and then use the formula for the intensity in terms of the total power. Assume the peak
of the light bulb is at 524 nm. Real incandescent lightbulbs radiate more like a blackbody, with a
broad spectrum of wavelengths. Since the TSL237 response varies strongly as a function of
wavelength, one must integrate the product of the sensor response and the spectral intensity of
the light source in order to get an accurate measurement of the light intensity. This is done in the
Appendix.

Note also pitfall of measuring 60Hz time varying light sources, causes jitter in output, which
coupled with sampling times that are not a multiple of 60Hz, will appear as aliased noise output
on meter. (i.e. meter sampling time needs to get the same number of 60 Hz. Cycles each
sampling interval. Slows down and speeds up freq. 60 times per second. A real problem.
Closeup of TSL 237 light to frequency chip centered on conical output of light source.

Pasco light source showing recessed bulb. Centerline of bulb is recessed 2.75 cm back from the
front surface.

KW S2011
Physics 227L

Examination of light source emission pattern for spatial non-uniformities that change with
distance from the light source.

KW S2011
Physics 227L

Examination of light source emission pattern for spatial non-uniformities that change with
distance from source due to shadowing and focusing effects from filament.

Backlit DMM is important so that readings can be made in the dark with minimal disturbance to
the background light levels.

Note the switch on the optical power sensor head. It is important to remember to switch this off
after making measurements to avoid draining the battery.

Procedure: Since it only takes 15-25 minutes to take a set of 15-25 data points, you should take
more than one set of data and compare the plots of the different data sets, especially noting the
differences in the exponent for the power law fit.

Since the intensity decreases rapidly for the first few set of distances (i.e. 1/102, 1/202, 1/302,
etc.), you should take data spaced apart by about 5 cm for distances between 10cm up to about
50 cm. For distances larger than 50cm, you can take data points about 10 cm apart out to the end
of the optical rail measurement, 120cm on the marking tape. In order to get two full orders of
magnitude in the distance values, it is necessary to take data at a distance of 2 meters away by
placing a meter (or 2 meter) stick on the ruled tape at the end of the optical rail and having one
person hold the meter stick and another hold the optical sensor head on it. Be sure that this 2
meter data point is taken along the line of the optical rail.

It is important that the light source be warmed up for at least 10 minutes in order to stabilize its
temperature and thus its light intensity. If the source is still warming up, the light intensity
values will change. Thus it is wise to repeat intensity measurements several minutes apart to see
how reproducible they are. How do you expect the intensity to change as the bulb is warming
up?

KW S2011
Physics 227L

Analysis of data: Plot the data as a scatter plot with no line connecting the data, and change the
scale on both axis to log so it will be a log-log plot. Add a power law trend line and display the
fit equation and R2 values on the graph, formatting both the exponent and R2 value to 4 decimal
places. The graph should have a title, and the axis with units should be labeled. Both major and
minor tick marks should be displayed on the inside of both the horizontal and vertical axis.
Make sure your graph looks nice and that the date and your names are displayed on it. (see
following sample graph and data). The correlation coefficient or coefficient of determination is a
measure of the goodness of the fit, the closer the value is to 1.000, the better the data fit the
model function, in this case, a power law function. Report your exponent and its uncertainty
according to the values on the Excel graph.

Discussion: Inverse square laws have been extremely important in the historical development of
physics. Newtons law of universal gravitational attraction for the force F between two spherical
or point-like masses m1 and m2 separated by a distance r between centers is F = Gm 1 m 2 /r2,
where G is the universal constant of gravitation. Coulombs law for the force between two point
charges or two charged spheres with charges q1 and q2 at a sufficient distance apart is F =
kq 1 q 2 /r2, where k is the Coulombs law constant.

For the case at hand, the fall off of intensity or irradiance from a point or spherical source of light
is I = P/(4r2), where P is the total (optical) power emitted by the spherical source. The question
is, where in nature or practice can we find a spherical source of light? The sun comes to mind,
and it is a good approximation to a spherical source of light. What about a light bulb?
Unfortunately, light bulbs produce a non-uniformity of the source that changes with distance due
to the shadowing of the hot glowing filament. This shows up as darker and brighter regions in
the light field, which change as a function of the distance from the source.

. For distances that are large compared to the dimensions of the source, the time averaged light
intensity or irradiance I (SI units of Watts/m^2) can be written as

I = B/r^2, where B is a constant (dont use A, confused with area). This is true in general,
regardless of the angular emission pattern of the light source, provided we are far away from the
source compared to its dimensions. For a spherically symmetric light source, (which can be very
difficult to achieve in practice), I = P/(4Pi*r^2), where P is the total power radiated. For uniform
conical, I=P/( r^2), where =2*Pi*(1-Cos(thetamax)) is the solid angle into which the cone
radiates (see diagram of aperture, measure cone thetamax, note that max=4*Pi, corresponding
to a sphere). It turns out that for incoherent excitation of individual elementary dipole radiators,
which normally radiate in a sin^theta pattern (see figure), these dipoles are equally excited along
all three axis, and thus radiate isotropically (reference Wanser AJP paper). For a uniform conical
emission pattern with half angle theta (see cone diagram), the intensity is related to the total
power. It is important to realize that the optical Intensity or Irradiance is (check this out) the
Poynting vector component that is parallel to the area in which the measurement is being made.
In simple cases, the magnitude of S is the intensity (not just the magnitude of S in general). S
dot n da is the flux of power thru a surface, in particular, a detector surface.

KW S2011
Physics 227L

Cross section of vertically oriented dipole radiation emission pattern, so called butterfly
emission.

Dipole emission pattern oriented along x axis. Sum of y and z oriented dipole emission patterns.

KW S2011
Physics 227L

Superposition of 3 independent dipole patterns along x, y, and z produces an isotropic emission


pattern.

For the case at hand, the fall off of intensity from a point or spherical source of light is I =
P/(4Pi*r^2), where P is the total power emitted by the spherical source. The question is, where
in nature or practice can we find an spherical source of light? The sun comes to mind, and it is a
good approximation to a spherical source of light. What about a light bulb? Unfortunately, light
bulbs produce a non-uniformity of the source that changes with distance due to the shadowing of
the hot glowing filament. This shows up as dark regions in the light field, which change as a
function of the distance from the source. A candle with short filament is a reasonable
approximation to a spherical source, but the best spherical source are individual dipole radiators
excited incoherently, which produce isotropic radiation due to equal oscillation along all three
coordinate axis.

Measure the light cone half angle of your light source. This can be done by measuring the
diameter of the light cone at 2 different distances from the source on a colored 3x5 card. The
change in diameter of the cone divided by the change in distance between the two measurements
is the tangent of the light cone half angle. Compute the solid angle omega of your light source,
and the power radiated into this cone. Also calculate the light power radiated, assuming the solid
angle is 4pi and compare the two numbers, which is the ratio of your solid angle compared to
4pi.

Comment on the closeness of your exponent to the ideal value and the correlation coefficient of
your data. Is this value reasonable given the measurement accuracy of your data. Estimate the
accuracy of your measurements.

Issues and Pitfalls: Though it is easy to discuss a point source of light, in practice it can be
difficult to obtain an accurate 1/r2 fall off of intensity with distance from the source. Stray light
due to reflections off of nearby objects, such as computer screens, walls, tables, nearby
experimenters, etc. can cause deviations of the exponent from the value of 2 for a point light
source. This is especially true with the Pasco light source, which has light coming out the top,
the bottom, and both the front and back apertures. A sufficiently large piece of black velvet
employed as a hood draped over the back and top of the Pasco light source suppresses most of
the sources of deviation from pure inverse square behavior. Non-uniform source intensity in the
forward cone of light due to filament focusing and shadowing effects can also cause deviations
in the exponent. These should be examined with a colored 3 x 5 card and their effect made
negligible by making sure the photodetector is slightly off to the side of the non-uniformity in the
beam (slight adjustment of the direction of the light source can be made by adjusting the
mounting bracket). The background light intensity in the room should be much less than the
measured light intensity with the point source light on. Typical background intensity in the room

KW S2011
Physics 227L

should be less than 100 Hz. Saturation of the light sensor may occur for frequencies greater than
500 KHz. Usual maximum (saturation) values are around 560 KHz. This saturation effect will
be apparent in the plots as a curving over to the left of the trendline thru the data. May have to
use pinhole fabricated from dark construction paper, index card, or tinfoil, in order to better
approximate a point source for data closer than 10cm. Saturation up close may be a problem,
limit light power values to 500KHz or more. Values higher than 500 KHz may be nonlinear in
power. Use of diffuser or pinhole on aperture effectively moves the location of the light source
forward to the front of the Pasco holder. This distance is about 0.75in and should be accounted
for in the distance ( r ) values when recording measurements and making plots of the intensity as
a function of distance.

Angular effects. The TSL 237 has a very large angular field of view, at least +/-45 in both the
horizontal and vertical planes, as can be seen from the following graphs. This makes it
particularly sensitive to stray light incident from a variety of angles, and if sources of stray light
can not be eliminated to a sufficient degree, a small aperture placed in front of the TSL 237 chip
should be used to limit the angular field of view of the sensor, so that it looks primarily at the
Pasco light source, and excludes sources of light coming from large angles. This limiting
aperture can be an aluminum beverage can cut in half and with a hole poked in it, placed over the
sensor holder post. The cut, razor sharp edge of the beverage can should be taped in order to
prevent it from causing injury.

KW S2011
Physics 227L

KW S2011
Physics 227L

Spectrum of Pasco Light Source

Pasco OS-8517 Light Source


(Phillips 10Watt, 12 Volt Halogen bulb #23269)
vs. 2935 & 2850 Kelvin Blackbody Radiators
1.20E-09

1.00E-09

8.00E-10
Power (Watts)

6.00E-10

4.00E-10
Keith Wanser
2.00E-10 2/24/2011 data
10nm resolution
0.00E+00
400 600 800 1000 1200 1400 1600
Wavelength (nm)

1.2 x .3 in. dimensions


31mm tall, 9mm diam.
Note that the majority of the power radiated by the light source is in the infrared beyond the end
of the visible spectrum of about 690 nm. The spectra for blackbody radiators at 2935 and 2850
Kelvin are also displayed, and show the general shape of the Pasco spectrum, however, they do
not account for the wavelength dependent emissivity of the tungsten filament, nor the absorption
or emission lines of the fill gas, in this case a halogen such as Iodine or Bromine (usually
Bromine). The source of light inside the light box is a 10-Watt 12-Volt halogen bulb (Philips
Order Code: 13284; Philips product number 232629, Philips Description: 10W/12V/Capsule,
PASCO part number 526-035). The manufacturer lists the color temperature as 2850K. Inside
the light box is somewhat hotter than in free space, thus the color temperature is actually a bit
higher, 2935 K fitting the peak of the spectrum better than 2850 K. 140 Lumens rated output.

Blackbody spectral function used

KW S2011
Physics 227L

A
I=
( ) where l is the wavelength, h Plancks constant, c the vacuum speed of
5 e hc / k B T 1
light, kB Boltzmans constant, and T the absolute temperature. A is a proportionality constant
adjusted to give the peak power detected, since the (fiber coupled) optical coupling efficiency is
far less than one. The fiber was simply inserted into the opening of the Pasco light source and
brought near to the bulb. Standard form for power/unit area/unit wavelength into all solid angles
C1
is I =
(
5 eC 2 / T 1), where C1 and C2 are the first and second radiation constants respectively,

given by C1 = 3.74177 x 10-16 Watt-m2, C2 = 1.43877 x 10-2 m-K. Thus I has the units of
Watts/m3, as it should since it is power per unit area per unit wavelength.

KW S2011
Physics 227L

http://zeiss-campus.magnet.fsu.edu/articles/lightsources/tungstenhalogen.html

KW S2011
Physics 227L

Normalized and absolute spectral responsivity of TSL 237 chip (from Mathematica interpolation
of numerous specification sheet graph points).

TSL 237
2
2.3kHz/(Watt/cm ) @ 524nm calibration point

Wavelength Normalized Responsivity Wavelength Normalized Responsivity


2 2
(nm) Responsivity kHz/(Watt/cm ) (nm) Responsivity kHz/(Watt/cm )
300 0.0000 0.0000 710 0.9932 2.8583
310 0.1116 0.3212 720 0.9808 2.8227
320 0.1762 0.5070 730 0.9635 2.7729
330 0.2191 0.6305 740 0.9425 2.7125
340 0.2658 0.7648 750 0.9170 2.6391
350 0.3206 0.9226 760 0.8884 2.5568
360 0.3581 1.0306 770 0.8582 2.4698
370 0.3930 1.1309 780 0.8285 2.3844
380 0.4305 1.2389 790 0.7978 2.2959
390 0.4686 1.3485 800 0.7642 2.1993
400 0.5053 1.4541 810 0.7285 2.0965
410 0.5385 1.5497 820 0.6905 1.9871
420 0.5693 1.6385 830 0.6504 1.8718
430 0.5979 1.7207 840 0.6105 1.7569
440 0.6229 1.7928 850 0.5738 1.6513
450 0.6457 1.8582 860 0.5388 1.5505
460 0.6672 1.9203 870 0.5045 1.4518
470 0.6884 1.9811 880 0.4697 1.3517
480 0.7108 2.0455 890 0.4346 1.2509
490 0.7340 2.1125 900 0.3991 1.1486
500 0.7557 2.1747 910 0.3617 1.0408
510 0.7741 2.2278 920 0.3248 0.9347
520 0.7907 2.2756 930 0.2903 0.8354
524 0.7970 2.2936 940 0.2596 0.7471
530 0.8060 2.3196 950 0.2311 0.6650
540 0.8205 2.3614 960 0.2044 0.5882
550 0.8344 2.4012 970 0.1785 0.5138
560 0.8478 2.4400 980 0.1540 0.4433
570 0.8615 2.4794 990 0.1309 0.3768
580 0.8758 2.5204 1000 0.1092 0.3142
590 0.8900 2.5612 1010 0.0886 0.2549
600 0.9040 2.6017 1020 0.0702 0.2019
610 0.9181 2.6422 1030 0.0554 0.1594
620 0.9318 2.6816 1040 0.0434 0.1248
630 0.9449 2.7193 1050 0.0334 0.0962
640 0.9572 2.7548 1060 0.0246 0.0707
650 0.9697 2.7907 1070 0.0174 0.0501
660 0.9812 2.8239 1080 0.0125 0.0359
670 0.9906 2.8509 1090 0.0101 0.0290
680 0.9971 2.8697 1100 0.0104 0.0300
690 1.0005 2.8792
700 1.0000 2.8778

KW S2011
Physics 227L

JUNK BELOW!

It is suggested that data be taken with the light source in different conditions, the unmodified
aperture, the aperture with a few pieces of tape at angles to each other acting as a diffuser to
homogenize the intensity and eliminate the shadow of the light bulb filament, which appears at
different distances from the light source. A pinhole placed over the light source aperture to limit
the effect of light source spatial non-uniformities that change with distance from the source.
Note that the bulb is powered by a DC power supply. Look at the light intensity of a 60Hz light
bulb, for example the light you turn on to see what your distances are and to record them when it
is dark in the room. What do you notice about the frequency readings on the meter compared to
those from the DC powered light source? Explain your observations.

KW S2011

Potrebbero piacerti anche