Realization of the NIST total Spectral Radiant Flux Scale - PMOD/WRC

Realization of the NIST Total Spectral Radiant Flux Scale
Y. Ohno and Y. Zong
National Institute of Standards and Technology, Gaithersburg, Maryland USA
Abstract. The total spectral radiant flux scale has been
realized at National Institute of Standards and Technology
using a goniospectroradiometer for the 360 nm to 800 nm
region. The construction of the goniospectroradiometer
and results of the scale realization as well as the
uncertainty budget are presented.
Introduction. There are increasing needs for total
spectral radiant flux standards, which are required to
calibrate integrating sphere systems employing a
spectroradiometer. Such integrating sphere systems are
increasingly used for measurement of color and total
luminous flux of light sources such as discharge lamps and
light-emitting diodes (LEDs) as well as for measurement
of total radiant flux of ultraviolet (UV) sources.
A realization of total spectral radiant flux in the visible
region using a gonio-colorimeter was reported (Goodman,
1991), which was based on measurement of the correlated
color temperature of a test lamp in many directions, and
the average spectral distribution of the lamp was derived
based on Planck’s equation. No further developments have
been reported. Since then, new methods and technologies
have become available to realize the scale more directly,
possibly with lower uncertainties.
To address the strong needs for total spectral flux
standards in the U.S. (CORM, 2001), a project has started
at National Institute of Standards and Technology (NIST)
to realize the total spectral radiant flux scale from the UV
to visible region. Two independent methods are employed
to realize the scale, allowing cross-check of the realized
scales. The first method uses a goniospectroradiometer
designed and built at NIST. The second method uses the
NIST 2.5 m integrating sphere, applying the principles of
the Absolute Integrating Sphere (AIS) method, which is
successfully used for the realization of the lumen since
1995 at NIST (Ohno, 1996).
For the first phase of the project, NIST total spectral
radiant flux scale has been realized in the 360 nm to
800 nm region using the goniospectroradiometer. This
paper presents the principles of the methods of realization,
the construction of the goniospectroradiometer, and the
results of the realization of the scale as well as its
uncertainty budget.
Methods for Scale Realization
Goniospectroradiometric method
The spectral radiant intensity of a light source is measured
in many directions (θ, φ) over 4π steradian using a
goniospectroradiometer. The total spectral radiant flux
Φ λ (λ) of the light source is given by
2π π
Φ λ (λ) = ∫ ∫ I λ (λ,θ,φ) sinθ dθ dφ , (1)
φ=0
θ =0
where I λ (λ,θ,φ) is the spectral radiant intensity
distribution of the source on spherical coordinates (θ, φ).
If the goniospectroradiometer measures spectral irradiance,
the total spectral radiant flux Φ λ (λ) of the light source is
given by
Φ λ (λ) = r 2 2π π
∫ ∫ E λ (λ,θ,φ) sinθ dθ dφ , (2)
φ=0 θ =0
where E λ (λ,θ,φ) is the spectral irradiance distribution on
the spherical surface with radius r around the light source
being measured.
The realization of the scale can be done more easily if the
absolute scale is brought from the luminous flux unit. Then
the spectral irradiance measurements can be done
relatively. In such a method, the total spectral radiant flux
of a lamp is given by
and
Φ λ (λ) = k scale
∫
2π
φ=0
k scale =
∞
K m V(λ)
∫
λ=0
∫
π
∫ S(λ,θ,φ) sinθ dθ dφ , (3)
θ =0
2π
φ=0
∫
Φ v
π
θ =0
S(λ,θ,φ) sinθ dθ dφ dλ
(4)
where S(λ,θ,φ) is the relative spectral and spatial
distribution of the lamp as given by
S(λ,θ,φ) = k ⋅ E λ (λ,θ,φ)
(k : an arbitrary constant) , (5)
and Φ v is the total luminous flux (lumen) of the lamp,
determined using other methods. With this relative
method, errors in the rotation radius r of the
goniospectroradiometer, the positioning of lamp and
detector, stray light from surrounding walls, and the
absolute scale calibration of the spectroradiometer are
mostly not relevant.
Absolute Integrating Sphere Method
The AIS method is applied spectrally. Figure 1 shows the
arrangement of an integrating sphere system for this
application. The flux from the spectral irradiance
standard lamp is introduced through a calibrated aperture
placed in front of the opening. The internal source, a lamp
to be calibrated, is mounted in the center of the sphere.
The external source and the internal source are operated
alternately. The flux Φ λ,ref (λ) introduced from the
external source is given by,
Φ λ,ref (λ) = A ⋅ E λ (λ) , (6)
where E λ (λ) is the average spectral irradiance from the
external source over the limiting aperture of known area A.
The total spectral radiant flux Φ λ,test (λ) of the test lamp
is obtained by comparison to the external radiant flux:
Φ λ,test (λ) = y test (λ)
y ref (λ) ⋅ k cor (λ) ⋅Φ λ,ref (λ), (7)

where y test (λ) is the detector signal of the
spectroradiometer for the test lamp at wavelength λ, and
y ref (λ) is that for the introduced flux at wavelength λ.
k cor (λ) is a correction factor for spatial nonuniformity of
the integrating sphere system and the ratio of diffuse
reflectance of the sphere coating at 0° and 45° incidence.
Similar to the case with the goniospectroradiometric
method, the measurement can be done for relative total
spectral radiant flux, with the absolute scale determined by
the luminous flux unit.
Instrumentation
The mechanical design of the goniospectroradiometer
developed at NIST is shown in Fig. 1. The detector arm
and lamp holder are rotated by servo motors, and the actual
angles are measured by rotary encoders. Measurements are
taken, typically at 5° or 10° intervals, and the motor stops
during measurement. The arm holding the lamp holder can
also be rotated to change the burning position of the lamp.
The spectral irradiance of the test lamp is measured with
an array spectroradiometer employing a fiber optic input.
The fiber bundle is connected to the spectroradiometer
through a rotation coupler, which allows free rotation of
the arm without twisting the fiber. The constancy of the
spectral transmittance of the coupler was tested with a
white LED mounted on the irradiance head. The variations
in measured spectra depending on the arm angle was found
to be less than 2 %, and is corrected in actual
measurements. From the rotation coupler to the irradiance
measuring head, a fiber bundle of 8 mm diameter is used.
The fiber bundle on the spectroradiometer side is 5 mm
diameter.
The array spectroradiometer covering the UV and visible
region is calibrated by a spectral irradiance standard lamp
(1000 W FEL type quartz halogen lamp) traceable to the
NIST spectral irradiance scale [7]. The stray light of the
array spectrometer has been corrected (Zong, 2005). For
the consistency with the total luminous flux unit, the total
spectral radiant flux scale has been realized using Eqs.
Servo
motor
Stepping
motor
Irradiance
head
1.25 m
φ
Light
trap
Encoder
Laser
Servo
motor
Fiber
bundle
Figure 1. Construction of the NIST
goniospectroradiometer.
θ
Spectroradiometer
Rotation
coupling
Figure 2. NIST 2.5 m integrating sphere configured
for the total spectral radiant flux measurement.
(3)-(5). The uncertainty budget is to be reported at
presentation and the full paper.
Figure 2 shows the arrangement of the NIST 2.5 m sphere
configured for total spectral radiant flux measurement.
The sphere is equipped with a high-sensitivity
spectroradiometer employing a back-lit CCD array. The
external source is a 1000 W FEL type quartz halogen lamp
calibrated for spectral irradiance. The aperture is 50 mm in
diameter. The spectroradiometer is an array
spectroradiometer, which is corrected for stray light (Zong,
2005). Linearity of the spectroradiometer has also been
determined and corrected. The integrating sphere response
has been mapped spectrally by rotating a beam scanner.
Angular correction factor for the coating has also been
measured spectrally. An experimental realization of total
spectral radiant flux scale in the near UV region (360 nm –
450 nm) has been successfully made at NIST using the 2.5
m integrating sphere to calibrate the total radiant flux of
deep blue and UV LEDs (Zong, 2004). Work is underway
to realize the scale in the full visible region.
Conclusion
The total spectral radiant flux scale has been established at
NIST using a goniospectroradiometer for the 360 nm to
800 nm region. The AIS method, now under development,
will enable realization of the scale with much simpler
instrumentation and fast measurements.
References
Goodman T. M., et al, The Establishment of a New National
Scale of Spectral Total Flux, Proc., 22 nd Session of CIE,
Melbourne 1991, Vol. 1, Part 1, 50-53 (1991).
CORM 7 th Report 2001–Pressing Problems and Projected
National Needs in Optical Radiation Measurements, December
2001.
Ohno Y., Realization of NIST 1995 Luminous Flux Scale using
Integrating Sphere Method, J. IES, 25-1, 13-22 (1996).
Zong, Y., et al, A Simple Stray-light Correction Matrix for Array
Spectrometers, to be presented at NEWRAD 2005
Zong, Y., Miller, C.C., Lykke, K., Ohno, Y., Measurement of
Total Radiant Flux of UV LEDs, Proceeding of the CIE Expert
Symposium on LED Light Sources, June 2004, Tokyo,
107-110 (2004)