Luminous Intensity Final Report: CCPR Key Comparison CCPR-K3.2014
Luminous Intensity Final Report: CCPR Key Comparison CCPR-K3.2014
Luminous Intensity Final Report: CCPR Key Comparison CCPR-K3.2014
Luminous Intensity
Final Report
Arnold Gaertner1, Éric Côté1, Joaquin Campos2, Gaël Obein3, Peter Blattner4,
Reto Schafer4, Liu Hui5, Jiang Xiaomei5, Cameron Miller6, Yuqin Zong6,
Errol Atkinson7, Erik Thorvaldson7, Kenichi Kinoshita8, Rheinhardt Sieberhagen9,
Irma Rabe9, Teresa Goodman10, Barry Scott10, Armin Sperling11, Detlef Lindner11,
Boris Khlevnoy12, Evgeniy Ivashin12.
Contents
1. Introduction ................................................................................................................................... 3
2. Organization .................................................................................................................................. 4
2.1. Participants, selection .......................................................................................................4
2.2. Participants, contact information ......................................................................................4
2.3. Task Group, selection .......................................................................................................6
2.4. Task Group, duties ............................................................................................................6
2.5. Comparison artifacts, selection .........................................................................................6
2.6. Comparison measurement and analysis components ........................................................6
3. Comparison Procedures and Timetable ..................................................................................... 9
3.1. Comparison Protocol ........................................................................................................9
3.2. Lamp Shipment to Pilot ....................................................................................................9
3.3. Lamp Measurement at Pilot ............................................................................................10
3.4. Lamp Re-measurement by Participants and Report of Results ......................................10
3.5. Pre-Draft-A Process 1: Verification of Reported Results ...............................................10
3.6. Pre-Draft-A Process 2: Review of Uncertainty Budgets ................................................10
3.7. Pre-Draft-A Process 3: Review of Relative Data ...........................................................11
3.8. Pre-Draft-A Process 4: Identification of Outliers and Consistency Check.....................11
3.9. Draft A ............................................................................................................................12
3.10. Comparison Timetable ....................................................................................................14
4. Measurement Data and Analysis ............................................................................................... 15
4.1. Uncertainty Analysis .......................................................................................................15
4.2. Participant Lamp Data ....................................................................................................16
4.3. Measurements at Pilot .....................................................................................................17
4.4. Calculation of the KCRV and the DOE ..........................................................................23
5. Comparison with 1999 CCPR-K3.a Key Comparison of Luminous Intensity...................... 27
6. Summary ...................................................................................................................................... 29
7. Acronyms ..................................................................................................................................... 32
8. References .................................................................................................................................... 32
9. Appendices ................................................................................................................................... 33
Appendix A NMI Reports
Appendix B Review of Uncertainty Budgets
Appendix C Summary of Participant Lamp Luminous Intensity Values
Appendix D Summary of Pilot Measurements of Participant Lamps
Appendix E Calculation of the KCRV and the Unilateral DOE
Appendix F Calculation of the Bilateral DOE
1. Introduction
1.1 The metrological equivalence of national measurement standards will be determined by a set
of key comparisons chosen and organized by the Consultative Committees of the CIPM
working closely with the Regional Metrology Organizations (RMOs).
1.2 At the 14th meeting of the Consultative Committee for Photometry and Radiometry (CCPR)
held on 1997-June-10 and 11, several key comparisons in the field of optical radiation
metrology were identified. In particular, it decided that luminous intensity/responsivity be
considered a Key Comparison (KC) and that the comparisons being piloted by PTB (K3.a
Luminous Intensity of lamps) and the BIPM (K3.b Luminous Responsivity of photometers) at
that time be treated as Key Comparisons. These first KCs of luminous intensity/responsivity
were completed in 1999 [1]. At the 20th meeting of the CCPR (2009-September-17, 18), it was
decided that a second round of key comparison CCPR-K3 be commenced [2]. The CCPR
approved “that for this next round there will be only one CCPR-K3 comparison, called
luminous intensity, and the details of the comparison (use of lamps or photometers) should be
decided by the task group carrying out the comparison. The task group will be established by
the WG-KC and its proposal of comparison artifacts shall be submitted to CCPR for
approval.”1 The National Research Council of Canada (NRC) was chosen to pilot this
comparison, with the intention that measurements would start in 2012.
1.3 The technical protocol was drawn up by the eight-member Task Group (TG) of the participants
of the CCPR-K3.2014 key comparison (see Section 2.3.), approved by all the participants, and
approved by the WG-KC.
1.4 This is the Final Report of the CCPR-K3.2014 Key Comparison. Draft A was reviewed and
approved by the participants in 2020-October. The draft B report was reviewed by CCPR WG-
KC from 2020-October to 2021-April, including one revision. The Draft B-2 was approved by
the CCPR WG-KC on 2021-April-02 and by the CCPR on 2021-November-30.
1.5 This report describes the comparison organisation (Section 2), the measurement methods and
uncertainties achieved at all the participants and at the pilot (Sections 3 and 4), and the method
for analysis and the results of the comparison according to this method (Section 4). It includes
a comparison of the results of this comparison with the 1999 comparison [1] (Section 5).
Section 6 presents a summary of the comparison.
2. Organization
2.1. Participants, selection
2.1.1 The invitation to participate in this comparison was prepared by the pilot laboratory and the
WG-KC, and then sent to all CCPR members by Michael Stock, Executive Secretary of the
CCPR.
2.1.2 The selection process for the participants was guided by the following criteria [4]:
1. The participant must be a member of CCPR.
2. The participant must be willing to serve as a link laboratory to their RMO.
3. The participant must have an independent realization of the unit or scale of the
comparison quantity.
4. The participant’s measurement capability of the comparison quantity, over the full range
of the comparison (e.g., full spectral range), must be listed in the CMC table published
at the time of the call for participants.
2.1.3 Since the number of applications exceeded the maximum of 12, the RMO Groups were
requested by the pilot to select the maximum number of participants in accordance with the
following Table One [4]:
Table One
RMO
RMO Group Members Maximum Number of Participants
Group
Group 1 EURAMET+COOMET 6
Group 2 APMP+AFRIMETS 4
Group 3 SIM 2
Table Two
intent of the pilot laboratory to measure all lamps from all participants under as identical
conditions as possible. To this end, the measurements at the pilot did not commence until all
NMI travelling standard lamps had been received at the pilot laboratory, and all lamps were
then measured sequentially using the same measurement set-up, over a time period of
approximately 2 months. The NMI lamps were measured upon at least two occasions for all
NMIs and several lamp sets were measured three times.
The quantity compared using this setup was the photometer signal produced by the optical
radiation of each lamp. This procedure does not compare the lamps to any photometric scale
of the pilot laboratory, so that the lamps from the pilot NMI may be considered on an equal
basis to all NMI participants. Since near-identical measurement conditions were used for
each lamp, the photometer signal gives a direct comparison of all NMI lamps. Two
photometers were used, sequentially at each measurement, to provide a measurement and
equipment check. The measurement results from the two photometers over the two-month
time period also provided information concerning the stability of the comparison reference
scale. Additional information concerning the stability of this scale was determined from extra
repeat measurement of the pilot lamps, and the repeat measurements of the NMI lamps also
gave an indication of the scale stability.
2.6.2 Comparison analysis: The fundamental outcomes of a key comparison are the Key
Comparison Reference Value (KCRV), the unilateral Degrees of Equivalence (DOEs)
between each NMI and the KCRV, and the bilateral DOEs between pairs of NMIs.
The measurement procedure presented above results in a photometer responsivity 𝑅𝑖,𝑗,𝑚 in
units of (cd/V) for each measurement. In this symbol, i is the NMI number (i=1 to 12), j is
the NMI lamp number (j=1 to number of lamps submitted by the NMI), and m is the
measurement number of that lamp at the pilot (m=1 to number of times the lamp was
measured at the pilot laboratory). The candela values are the values for each lamp as obtained
from the measurements submitted by the NMIs.
An average (weighted mean) value 𝑅𝑖 is determined for each NMI. This ensures that each
NMI is treated equitably and that the results do not depend upon the number of lamps
submitted by the NMI, nor the number of times the lamps were measured at the pilot
laboratory. However, the uncertainties associated with the final 𝑅𝑖 for each NMI will depend
upon both the number of lamps and the number of repeat measurements since the
uncorrelated (random) aspects will be affected.
The KCRV is then determined from these 12 values of 𝑅𝑖 . In this comparison, the luminous
intensity scale of one of the NMIs (NMISA) was not their own independent realization.
Consequently, a tentative KCRV was determined from the remaining 11 NMIs. The DOEs
can be determined for all 12 participants. The KCRV determined from the 𝑅𝑖 is the
responsivity 𝑅𝐾𝐶𝑅𝑉 (cd/V) of the pilot photometer as determined by the measurement of a
virtual KCRV Luminous Intensity lamp measured under the same conditions as the NMI
lamps.
The uncertainties in the determination of this KCRV are based upon the combination of three
basic uncertainties applied to the 𝑅𝑖,𝑗 measurements of each lamp: (1) the uncertainties in the
luminous intensity calibration of each NMI travelling standard as determined from the
measurements of each NMI, (2) the uncertainties of the comparison measurements made at
the pilot laboratory, and (3) an estimate of each lamp’s repeatability as determined from the
measurements at the pilot laboratory. These uncertainties are combined to produce the
uncertainties in the weighted mean 𝑅𝑖 for each participant.
The determination of the initial tentative KCRV for this comparison was made according to
the CCPR guidelines [6], using only the data for the 11 NMIs as indicated above. It is based
upon a weighted mean with ‘cut-off’. The weights are determined based upon the NMI
reported uncertainties adjusted by the cut-off, combined with the transfer uncertainty of the
comparison and the uncertainty caused by the estimated lamp reproducibility observed
during the measurements at the pilot lab. The cut-off value for the NMI uncertainty is
determined as the average of the uncertainty values of those participants that reported
uncertainties smaller than or equal to the median of all the participants.
This initial KCRV was then tested for statistical consistency with the measurement results
using two criteria [6]: testing for statistical ‘outliers’, and testing for statistical indications of
under-estimated uncertainties.
As indicated in the CCPR-G2 guidelines [6], Pre-Draft-A Process 4, an ‘obvious outlier’ was
defined as participant results whose 𝑅𝑖 deviated from the KCRV by more than 3 times its
associated expanded (k=2) uncertainty. There were no ‘obvious outlier’ participants.
The CCPR-G2 guidelines [6] defined the Chi-Square (=0.05) test as the statistical
indication of under-estimated uncertainties. This test failed for the measurement data. It was
observed that a large contribution to the Chi-Square (observed) for the measurements was
due to the results from one participant, whose contribution to the Chi-square (observed) was
more than one-half the total. In addition, although not an outlier, their 𝑅𝑖 deviation from the
KCRV was close to 3 times its’ associated expanded (k=2) uncertainty. The comparison
participants agreed that the results from this NMI would not be used to determine the KCRV,
although the DOE for their results would be determined.
A new tentative KCRV was then calculated from the results of the remaining 10 participants
and the statistical tests were re-applied. The Chi-square (observed) test again failed and the
Mandel-Paule method was applied as suggested in the CCPR-G2 guidelines [6]. The
additional uncertainty required to enable the Chi-square (observed) to pass the test was
determined to be reasonable and the participants agreed to use this procedure to determine
the KCRV for this comparison.
After the publication of Draft A v1.0, LNE-CNAM pointed out that the results for their
transfer lamps deviated much farther from the KCRV than could be expected based upon
their previous experience and measurements. It was noted that their results were for the
participant identified above with 𝑅𝑖 deviation from the KCRV close to 3 times its’ associated
expanded (k=2) uncertainty. It was also noted that for the shipment of lamps from LNE-
CNAM, two lamps of the Polaron-type were received broken at NRC, and that one of the
Osram-type lamps indicated significant changes when re-measured at LNE-CNAM after the
pilot measurements, indicating that the entire shipment of their lamps had experienced a
severe shock during the shipment from LNE-CNAM to the pilot. They considered it highly
likely that the remaining 3 lamps would have suffered a short-term instability that would not
have been noticed during the comparison measurement sequence, and requested that the
results for their lamps be withdrawn from the comparison. This was accepted by the
participants. This request did not affect the calculations for the KCRV or the DOE, since, as
discussed above, the results from LNE-CNAM were considered ‘outlier’ data and not
included in these calculations.
Table Three
NMI Number of Lamps
Osram Polaron Final Transportation Receipt Returned
1
One of the Polaron lamps from IO-CSIC failed during measurements at the Pilot lab.
2
Two of the Polaron lamps shipped from LNE-CNAM were received broken at the Pilot lab. As a
result the remaining lamps were hand carried by LNE-CNAM for the return to LNE-CNAM.
Upon their return measurements, they observed a large change in one of the Osram lamps and
asked to have it removed from the comparison. After the publication for Draft A v1.0 it was
agreed to remove all the LNE-CNAM lamps from the comparison. See Section 2.6.2.
red type). It also includes the reason for the changes in VNIIOFI luminous intensity and
uncertainty values. PDA Process 2 was completed by 2016-July-11.
3.7. Pre-Draft-A Process 3: Review of Relative Data
The pilot lab prepared the “Relative Data” of each participant, which are the data reduced to
show only the stability of transfer standards for each participant before (Round #1) and after
travel (Round #2) and the internal consistency of all the transfer standards measured at each
participant lab. The “Relative Data” for all participants was distributed to all participants
without identifying any of the participants.
As a response to this data, there were three requests to remove a lamp from the comparison,
one request to change the reported Luminous Intensity of a lamp, and one request to use only
data from the first set (Round #1) of measurements by the NMI.
1. NMIJ requested removal of their lamp #69 due to the large change in value between
their Round #1 and Round #2 measurements.
2. NMIA requested removal of their lamp S14 due to the large change in value between
their Round #1 and Round #2 measurements.
3. After some discussion among the participants, NIM requested removal of their lamp
G-1071 due to the large difference in the relative value of this lamp between the
Pilot and NIM. It was concluded that this difference was due to a difference in the
construction of this lamp from their other lamps that caused the measurement set-
ups at the two laboratories (Pilot and NIM) to produce different results.
4. IO-CSIC requested a change in the Luminous Intensity value given to their lamp A454
for the Round #2 measurements. They traced this to be caused by the use of the
incorrect value for the resistance of their standard resistor used to measure the lamp
current (see Appendix A).
5. NPL compared the relative data and their repeat (R#1, R#2) measurements and
concluded that 2 lamps (Wotan lamps 877 and 890) had changed after the pilot
measurements. Thus the R#1 data could still be used, but the R#2 data removed (see
Appendix A).
After some discussion, these changes were accepted by all the participants (2017-February-
09). This explains the values in the ‘Number of Lamps/Final’ column in Table Three above.
The values for LNE-CNAM are explained in Section 2.6.2 above.
3.8. Pre-Draft-A Process 4: Identification of Outliers and Consistency Check
3.8.1 Pre-Draft-A Process 4: Identification of outliers:
A tentative KCRV was calculated by the pilot using the procedure described above in Section
2.6.2. There were no ‘obvious outliers’, as pre-defined in the CCPR-G2 Guideline [6], whose
average 𝑅𝑖 deviated from the tentative KCRV by as much as 6 times (k=1) the associated
uncertainty for 𝑅𝑖 .
3.8.2 Pre-Draft-A Process 4: Consistency check:
The consistency check of the data with the initial tentative weighted mean KCRV, using the
2
Chi-square (=0.05) test defined in the CCPR-G2 Guideline [6], resulted in 𝜒𝑜𝑏𝑠 ≈ 45.4.
2
This was larger than the 𝜒0.05 (𝜈 = 10) ≈ 18.3 consistency check indicated in the example in
Appendix B of the CCPR-G2 Guideline [6]. Since the data fails this Chi-square test, the
Guideline then suggests the use of the Mandel-Paule method, applying an additional “inter-
laboratory variance” s2 to all the participant uncertainties that will force the data set to pass
2
the Chi-square test. However, it was noted that more than one-half of this 𝜒𝑜𝑏𝑠 was due to the
results of the measurements on the lamps from one NMI. It was also noted that the
measurements of these particular lamps showed a deviation from the tentative KCRV that
were close to the 6 times (k=1) their associated uncertainty for 𝑅𝑖 .
Considering the possibility of removing this data from the calculation of the KCRV, the data
was reanalyzed for a tentative KCRV based upon the remaining 10-participant data. This
2 2
resulted in 𝜒𝑜𝑏𝑠 ≈ 18.3., which is only slightly higher than the 𝜒0.05 (𝜈 = 09) ≈ 16.9 given
in the table in the example Appendix B of the CCPR-G2 Guideline [6].
The pilot then conveyed this data to the participants (2019-March-11), and suggested that the
comparison KCRV be based upon the measurements of the 10 participants whose data
2
resulted in the 𝜒𝑜𝑏𝑠 close to the CCPR-G2 guideline [6]. Replies were received from 4 of the
participants, all of whom approved the determination of the KCRV from the data of the
remaining 10 participants. Several replies recommended, noting the requirement of the
2
CCPR-G2 guideline [6], that since the 𝜒𝑜𝑏𝑠 was still higher than 16.9, that the Mandel-Paule
2 2
method be applied to reduce the 𝜒𝑜𝑏𝑠 to the 𝜒0.05 (𝜈 = 09) ≈ 16.9 given in the table in the
example Appendix B of the CCPR-G2 Guideline [6].
The pilot then performed the analysis to include the Mandel-Paule uncertainty. It was
2
determined that the additional uncertainty required to reduce the 𝜒𝑜𝑏𝑠 , from approximately
18.3 to approximately 16.9, was small, approximately 0.06%, and that the changes in the
KCRV and the unilateral degrees of equivalence were also small, approximately 0.0014%.
This uncertainty could be explained as an adjustment due to ‘uncertainties’ in calculating the
uncertainties. The pilot then proposed that the data analysis for the Draft A comparison
report would include the Mandel-Paule adjustment of the uncertainties of the comparison.
The participants agreed (2019-May-24) to this procedure to determine the KCRV and DOEs
for this comparison.
3.8.3 Pre-Draft-A Process 4: Inconsistent Data Issues:
At this point it can be noted that the inconsistent data were from the measurements of the
lamps submitted by LNE-CNAM. As indicated above in Section 3.2, two of the Polaron
lamps shipped from LNE-CNAM were received broken at the Pilot laboratory. In addition, as
indicated above in Section 3.4, LNE-CNAM observed, upon their return measurements, a
large change in one of their Osram lamps between their pre-shipment and after return
measurements, and requested that this lamp be removed from the comparison.
The return measurements at LNE-CNAM for the remaining 3 lamps did not indicate a change
in these lamps larger than they would expect from the uncertainties in their measurements.
Consequently, this comparison cannot choose between an actual difference in the luminous
intensity scale at LNE-CNAM with respect to the KCRV, and the possibility that the lamps
were damaged during their transit to the pilot laboratory and subsequently ‘annealed’ during
the measurements performed at the pilot and then at LNE-CNAM upon their return, which
was done by hand-carrying the lamps.
3.9. Draft A
3.9.1 Draft A v1.0
Draft A v1.0 was prepared and sent on 2019-July-24 to all participants for review. As a
results of comments received the next version of Draft A was prepared.
3.9.2 Draft A v2.0
Draft A v2.0 was prepared as described below and sent on 2020-January-09 to all
participants for their review.
3.9.2.1 Removal of LNE-CNAM results:
As discussed in Section 2.6.2 above, after the publication of Draft A v1.0, LNE-CNAM
pointed out that the results for their transfer lamps deviated much farther from the KCRV
than could be expected based upon their previous experience and measurements. Noting the
shock that their shipment of lamps had sustained upon shipment to the pilot, causing the
breakage of some lamps, they requested that the results for all their lamps be removed from
the comparison. This was agreed to by the participants. This request did not affect the
calculations for the KCRV or the DOE, since, as discussed above, the results from LNE-
CNAM were considered ‘outlier’ data and not included in these calculations.
3.9.2.2 Adjustment of Weighting Factors used for Weighted Means of Participant Data:
Participant measurement data is combined at two stages: i) the combination of Round #1 and
Round #2 values for each lamp, and ii) the combination of the results for all lamps used by a
participant into a final value for the participant. The weights used in Draft A v1.0 for each of
these combinations was the usual statistical inverse square of the (absolute) uncertainties of
the luminous intensities. It was noted that this caused an issue with the combination of results
of all lamps for participants that used both the Osram-type lamps and the Polaron-type lamps.
If the lamps all have the same fractional luminous intensity uncertainties, but quite different
luminous intensities, the weights for the higher intensity (Type Polaron) lamps are
considerably smaller than the weights for the Osram-type lamps if absolute uncertainties are
used. It was requested that fractional uncertainties be used for these calculations in order to
provide a more equal weighting of all lamps used by the participant. This will have an almost
negligible effect for participants that only sent one type of lamp, but will affect those that
sent both types of lamps. For Draft A v2.0, fractional uncertainties are used for the
combination of the results for all lamps used by a participant.
3.9.2.3 Uncertainty Analysis:
The uncertainty analysis, presented in Section 4, combines the many uncertainty components
of the participant and pilot measurements. Concern was expressed that the use of the terms
Type A, Type B, uncorrelated and correlated were inconsistently and incorrectly applied by
both the participants and the pilot in the uncertainty analysis. The explanation and analysis in
Section 4 has been reworked to provide more clarity in the procedures used in the preparation
of this report. In particular, the use of the terms Type A and Type B has been avoided since
they describe the origins/evaluations of uncertainties, whereas the combination of
uncertainties requires the use of the correlations or non-correlations between variables. The
predominant changes in calculations were made in the combination of participant
uncertainties for each lamp into a final participant uncertainty (Section 4.2.3), and in the
calculation of the pilot transfer uncertainty (Section 4.3.4). These changes had very little
effect on the DOEs and their uncertainties from the values presented in Draft A v1.0.
3.9.3 Draft A v2.1
As a results of comments upon Draft A v2.0, a slightly modified version, Draft A v2.1, was
prepared. This version corrected the mathematical application of the split of the uncorrelated
component of each participant lamp into correlated and uncorrelated components as
described in Section 4.2.3. If a fraction 𝑓 of the original uncorrelated component is taken as
the final uncorrelated component, then a fraction √(1 − 𝑓 2 ) of the original uncorrelated
component must be taken as the additional correlated component in order to keep the total
final uncertainty for each lamp the same before and after the split. This correction (𝑓=0.5)
had a negligible effect upon the results.
As a result of further comments concerning the preparation of participant uncertainty
analyses and the separation of uncertainties into correlated and uncorrelated components at
various stages of uncertainty combination, an additional paragraph was added to the
Summary.
3.10. Comparison Timetable
Table Four
Comparison Timetable
𝑓 = ∑𝑛𝑖=1 𝜔𝑖 ∙ 𝑥𝑖 (1)
𝜕𝑓
= 𝜔𝑖 (2)
𝜕𝑥𝑖
𝜕𝑓 2
2 (𝑓)
𝑢𝑢𝑐 = ∑𝑛𝑖=1 (𝜕𝑥 ) ∙ 𝑢𝑢𝑐
2 (𝑥 ) 𝑛 2 2
𝑖 = ∑𝑖=1 𝜔𝑖 ∙ 𝑢𝑢𝑐 (𝑥𝑖 ) (3)
𝑖
𝜕𝑓 2
𝑢𝑐2 (𝑓) = [∑𝑛𝑖=1 (𝜕𝑥 ) ∙ 𝑢𝑐 (𝑥𝑖 )] = [∑𝑛𝑖=1 𝜔𝑖 ∙ 𝑢𝑐 (𝑥𝑖 )]2 (4)
𝑖
The subscripts uc and c indicate uncorrelated and correlated, respectively. These uncorrelated
and correlated uncertainties can be carried into subsequent calculations as appropriate. The
final total uncertainty for the function 𝑓 is given by the combination:
𝑢2𝑇 (𝑓) = {(∑𝑛𝑖=1 𝜔𝑖2 ∙ 𝑢𝑢𝑐
2
(𝑥𝑖 )) + ([∑𝑛𝑖=1 𝜔𝑖 ∙ 𝑢𝑐 (𝑥𝑖 )]2 )} (5)
Z
SideView X
Optical Axis
Optical Axis
X
Y
Top View
Telescope
Figure One
Schematic of NRC Measurement Configuration
LIS = Luminous Intensity Standard (Osram lamp shown)
measurements by three photometers sequentially for each lamp measurement. There were
three cycles of measurement for each lamp lighting. Each cycle consisted of the
measurements:
Lamp current, lamp voltage, photometer #1, photometer #2, photometer #3, lamp current
lamp voltage. Since there were shutters in front of each photometer, photometer zero
measurements were taken before and after the photometer signal measurements. Each of
all these (13) measurements was a sequence of 5 voltmeter readings.
The use of three photometers gave one measure of the stability of the measurement
configuration during the comparison. The data from one photometer was discarded due to
drift that was traced to a mechanical instability in the mounting of the photometer upon the
linear table. The remaining two photometers gave almost identical results. Consequently, the
data from only one of these photometers was used for the final analysis.
Further discussion of the NRC measurements is given in the following uncertainty analysis
for the comparison measurements.
4.3.2 Uncertainty of Comparison Measurements at Pilot (NRC)
A summary of the uncertainties is presented in Table Five. The column ‘Combination Type’
indicates the correlation between the variables for the purpose of combining participant lamp
measurements.
Table Five
Pilot Measurement Uncertainty Budget
Relative Standard
Combination Uncertainty in
Number Source of Uncertainty
Type Luminous
Intensity
NRC Optical Coordinate System
1 Alignment of Z-axis correlated 0.0000002
Alignment of Y-axis (Telescope optical
2 correlated 0.0000025
axis)
NRC Photometer
3 Spectral Mismatch Error correlated 0.0001
4 Responsivity Drift uncorrelated 0.0005
5 Signal Noise/fluctuations uncorrelated 0.000001
6 Alignment to optical axis (Y-Z center) correlated 0.000000
7 Alignment to optical axis (Y-Z angular) correlated 0.000000
Participant Lamps
Electrical
8 Standard resistor calibration correlated 0.0000042
DVM voltage calibration (lamp
9 correlated 0.00007
current)
10 Lamp current setting uncorrelated 0.0000498
11 Lamp current fluctuations uncorrelated 0.0000616
Optical
12 Vertical filament plane (Z-axis) uncorrelated 0.000140
13 Vertical filament plane (Y-axis) uncorrelated 0.000035
14 Lamp to Photometer distance uncorrelated 0.0003
Photometric
Lamp output fluctuations 0.0000262 /
15 uncorrelated
(Osram/Polaron) 0.0002322
SQRT(SUMSQ) Uncorrelated:
0.061% / 0.065%
(Osram/Polaron)
SQRT(SUMSQ) Correlated:
0.012% / 0.012%
(Osram/Polaron)
Uncertainty #10: Lamp current setting: Since we measured the lamp currents for each lamp
(see Section 4.3.1), we can compare the difference between the measured value and the NMI
set value. The average difference was -0.00021% with scatter of 0.00068%. Instead of
correcting for the shift, a larger uncertainty may be calculated by combining the two
[Reference (8), equation (22)]:
𝑢2 = (0.0000021)2 + (0.0000068)2 => 𝑢 = 0.0000071
This is an uncertainty in current, so the commensurate uncertainty in lamp output is scaled by
a factor of 7 as we did for Uncertainties 8 and 9.
Uncertainty #11: Lamp current Fluctuations: Since we measured the lamp current 30 times at
each lighting, we can calculate the fractional standard deviation for each lighting, and then
calculate an average value for all the lamps and lightings. This gave a result of 0.0000088.
This is an uncertainty in current, so the commensurate uncertainty in lamp output will be
scaled by a factor of 7 as we did for Uncertainties 8 and 9.
4.3.2.5 Participant Lamps - Optical
Uncertainty #12: Vertical Filament Plane (Z-axis): This is the uncertainty in luminous
intensity caused by the uncertainty in aligning the filament plane parallel to the Z-axis for a
rotation about the Y-axis. For the case of lamps with center filament support where only one-
half of the filament is visible, estimate an uncertainty of 0.2 mm in the 8 mm visible. Using
the equations from Section 4.3.2.1, we obtain the estimated uncertainty of 0.000140.
Uncertainty #13: Vertical Filament Plane (Y-axis): This is the uncertainty in luminous
intensity caused by the uncertainty in aligning the filament plane parallel to the Y-axis for a
rotation about the Z-axis. Estimate an uncertainty of 0.2 mm in the 16 mm of the filament
plane. Using the equations from Section 4.3.2.1, we obtain the estimated uncertainty of
0.000035.
Uncertainty #14: Lamp to Photometer Distance: This is the uncertainty in luminous intensity
caused by the uncertainty in setting the lamp filament plane to the telescope crosshair focus
point. Estimate 0.5 mm, which results in a relative standard uncertainty in luminous intensity
of:
0.5
𝑢 =2∗ = 0.0003
3200
4.3.2.6 Participant Lamps - Photometric
Uncertainty #15: Lamp Output Fluctuations: This was estimated from the average fractional
standard deviation in all the photometer measurements of all the lamps. It was different for
the two types of lamps (Osram and Polaron).
4.3.3 Transfer Lamp Reproducibility at Pilot
Most participant lamps were measured at the pilot only two times. The reproducibility of
each transfer lamp was estimated as the standard deviation of the mean of all the (m)
measurements of the lamp at the pilot. While this value may contain effects of photometer
drift (Section 4.3.2.3 above), and is based on only a few (m) measurements, it gives
information concerning the scatter of the (m) pilot measurements about their mean value.
4.3.4 Pilot Measurement Data
The Pilot measurement data for each participant lamp is summarised in Appendix D. The
NMI Lamp Luminous Intensity values are taken from Appendix C (Section 4.2). The column
R(i,j) is the average photometer responsivity 𝑅𝑖,𝑗 (see Section 2.6.2) as determined from all
the (m) measurements of the lamp j of the participant i at the pilot laboratory. The Pilot
uncertainties uuncorr and ucorr are taken from Table Five above. The Pilot uncertainty uuncorr has
CCPR-K3.2014: Luminous Intensity, Final Report Page 22 of 33
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
been reduced by the factor 1⁄√𝑚 where m is the number of measurements of the lamp j. The
column uuncorr(lamp) is the transfer lamp reproducibility as described in Section 4.3.3 above.
These 5 uncertainty components (NMI uuncorr and ucorr, Pilot uuncorr and ucorr, and uuncorr(lamp))
are combined to give the uncorrelated and correlated uncertainties of 𝑢𝑅𝑖,𝑗 , the uncertainty of
𝑅𝑖,𝑗 , the combined “NMI+Pilot” quantity. At this point the uncertainties of the NMI and the
Pilot are not correlated, so the NMI total uncertainty is combined in quadrature with the Pilot
correlated uncertainty to give the final correlated component for the combination of the
measurements for each NMI. The uncorrelated component for this combination is the
combination of the Pilot uncorrelated uncertainty (Pilot uuncorr ) and the lamp uncorrelated
uncertainty (uuncorr(lamp)).
The calculations for the determination of the average value 𝑅𝑖 for each participant are also
presented in Appendix D. The average value is a weighted mean where the weights are
determined from the final ‘Participant + Pilot’ uncertainty 𝑢𝑅𝑖,𝑗 for each lamp given in the
‘combined uncertainty’ column uT (relative standard uncertainty) and subsequent uT(cd/V).
The uncertainty 𝑢𝑅𝑖 for this weighted mean 𝑅𝑖 is determined from the uncertainties of the
individual uncertainties (uncorrelated and correlated) of 𝑢𝑅𝑖,𝑗 using the formulas of equations
(1) to (5).
The calculation of the uncertainties of the comparison measurements made at the pilot
laboratory (transfer uncertainty of the comparison, Section 2.6.2), are also presented in
Appendix D. This was calculated for each participant as the difference uncertainty
2
𝑢𝑁𝑀𝐼_𝑡𝑟𝑎𝑛𝑠𝑓𝑒𝑟 = 𝑢𝑅2 𝑖 − 𝑢𝑁𝑀𝐼
2
as given in equation (22) of the CCPR Guidelines (reference [6], Appendix B).
The results are plotted in Figure Two.
Figure Two
CCPR-K3.2014 Luminous Intensity Unilateral Degrees of Equivalence
Uncertainties are k=1 values.
and detailed in Appendix C. As discussed in Sections 3.9 and 4.2 above, care should be used
in the determination of when (at which stage of data analysis and combination) uncertainty
components are considered correlated or uncorrelated.
Table Six
Summary of Comparison Values
KCRV relative standard (k=1) uncertainty u(KCRV) = 0.068%
Mandel-Paule relative standard (k=1) adjustment uncertainty s = 0.031%
Table Seven
Comparison of CCPR-K3.a 1999 and CCPR-K3.2014 Luminous Intensity
Comparisons
CCPR-K3.a 1999 CCPR-K3.2014
NMI u(DOE) u(DOE)
DOE DOE
(k=1) (k=1)
Figure Three
Comparison of CCPR-K3.a 1999 and CCPR-K3.2014
Uncertainties are relative standard (k=1)
The 1999 KCRV was determined from the results of 15 NMIs. The uncertainty of the KCRV
1999 2014
has changed slightly from 𝑢𝐾𝐶𝑅𝑉 = 0.086% to 𝑢𝐾𝐶𝑅𝑉 = 0.068%. As can be seen from
Figure Three, this change is negligible compared to the DOE and the DOE uncertainties.
The relationships between the NMIs are also very similar for the two comparisons, with
perhaps only one NMI (VNIIOFI) showing a significant shift in its DOE compared to the
other NMIs. The second NMI (NMISA) with a significant shift does not have an independent
LI scale and was not used for the calculation of the KCRV for this 2014 comparison.
The changes in the DOE for each NMI between the two comparisons is within the (k=1)
uncertainties of the two comparisons, except for the two NMIs mentioned above (VNIIOFI
and NMISA).
𝑢2014
There has been a small change in the DOE uncertainty values 𝑢𝑖 .The ratios 𝑢𝑖1999 vary from
𝑖
approximately 0.5 to 1.5, with an average of 0.8.
The 1999 comparison measurements were made in 1997 [1] and these 2014 comparison
measurements were made in 2014. This comparison would suggest that there has not been a
significant change in the luminous intensity scales of many NMIs during these 17 years.
There could be several possible reasons for this observation:
1. There have been no new independent realisations of luminous intensity since 1997.
2. The primary realization of luminous intensity requires the measurement of
luminous/optical power within specified geometrical conditions. Have we reached an
important limitation in our measurement ability for either of these measurement challenges?
3. The primary realization of luminous intensity indicated in (2) above must be
transferred to working standards and transfer standards. The CCPR comparisons used
incandescent lamps as these transfer standards. Are we limited by the characteristics of these
standards?
3.1 The second largest uncertainty at the pilot lab for the
measurement/comparison of all the lamps was the Lamp-to-Photometer distance at
0.03%, predominantly due to the issues of aligning a non-planar thick filament plane.
While this is an order of magnitude less than the typical DOE uncertainty, it is a
significant component in defining the geometrical measurement conditions.
3.2 The reproducibility/repeatability/ageing/portability of the transfer
standards. The average repeatability of the 62 lamps measured at the pilot was 0.09%,
with a standard deviation (of the dataset) of 0.08%. While this repeatability will
contain some of the realignment uncertainty (0.03%), and some of the photometer
responsivity drift, it is still an important uncertainty. (Note that for this comparison it
was difficult to completely separate the lamp repeatability and the photometer drift
uncertainties during the pilot measurements.)
6. Summary
The CCPR Key Comparison of Luminous Intensity using incandescent lamps as transfer
standards (CCPR-K3.2014) was carried out between the years of 2014 to 2020 with NRC
selected as the Pilot laboratory. The 12 participants (NMIs) were selected from/by the 3
RMO groups to participate and to be willing to serve as a link laboratory to other NMIs in
their RMO. The decision to use incandescent lamps rather than photometers as the transfer
standards was determined by the 8-member task group after discussions with the participants.
The comparison was organised as a star comparison with measurement sequence NMI-Pilot-
NMI, and transfer lamps supplied by each NMI. To facilitate the measurement of all lamps
from all participants under as identical conditions as possible, the measurements at the pilot
did not commence until all NMI travelling standard lamps had been received at the pilot
laboratory. All lamps were then measured sequentially using the same measurement set-up,
over a time period of approximately 2 months during 2014-November to 2015-January.
A total of 70 transfer standards (62 of Type Osram and 8 of Type Polaron) were received at
the pilot. Two lamps (Type Polaron, fixed to heavy sockets) were received at pilot broken in
shipment. One lamp (Type Polaron) failed during measurement at the pilot lab. Five lamps
(Type Osram) were removed from the comparison after re-measurements at the originating
NMI indicated changes in the lamps larger than could be explained by the uncertainties of
measurement. Three more lamps (2 Type Osram and one Type Polaron) were removed after
publication of the Draft A v1.0 report. These three lamps were part of the same shipment of
lamps that arrived at the pilot with two broken lamps and their withdrawal from the
comparison meant the withdrawal of all the LNE-CNAM lamps from the comparison.
Consequently a total of 59 lamps (55 of Type Osram and 4 of Type Polaron) were used to
produce the final results of the comparison.
one final value for the lamp, then combination of all lamps from each participant into a final
participant representative value, and finally the combination with the measurements of each
lamp at the pilot to estimate a KCRV, DOEs and their uncertainties. The correlations
between uncertainty components will change for each stage. In addition, the determination of
the weighting methods (relative uncertainties, absolute uncertainties, or other) used for each
of these combinations needs to be considered.
The method for combining measurement values and their uncertainties will also depend upon
the purpose of the key comparison: to obtain a best worldwide determination of the value of
the candela unit, to determine current NMI measurement capability (best or calibration level),
or in some way to justify CMC claims.
The protocol developed for this key comparison did not anticipate the detail and
documentation required to address this amount of detail in uncertainty evaluations, which
was later noted by several participants during the pre-draft A and Draft A stages of the report
preparation. However, the protocol did identify the use of the CCPR G2 guideline that was
followed reasonably closely for this comparison.
Nevertheless, the several versions of the Draft A with several changes to the combination of
the uncertainties showed that the differences in the final KCRV and DOE values and their
uncertainties were very small between different methods of combining the measurement
values and uncertainties. This may indicate that a more important aspect indicated by this key
comparison is the absolute accuracy of the primary luminous intensity scales developed by
each participant and/or the suitability of present travelling transfer artefacts to evaluate the
relationships between the participants at the uncertainty levels presently attainable by the
participants in developing their luminous intensity values.
The pilot laboratory (NRC) would like to thank all the participants for their constructive
support and collaboration during the course of this comparison. The intercontinental,
international, shipment of fragile transfer standards requires considerable effort from all
participants. The subsequent evaluation of transfer standard data acceptability and
measurement analysis requires collaboration from all participants. The pilot is grateful that
all participants have readily participated during the many facets of the comparison.
7. Acronyms
BIPM Bureau International des Poids et Mesures
CCPR Consultative Committee for Photometry and Radiometry
CCT Correlated Colour Temperature
CIPM Comité international des poids et measures
KC Key Comparison
KCDB Key Comparison Data Base
KCRV Key Comparison Reference Value
LIS Luminous Intensity Standard
NMI National Metrology Institute
RMO Regional Metrology Organization
TG Task Group for CCPR-K3.2014
WG-KC CCPR Working Group on Key Comparisons
8. References
[1] K3.a: Georg Sauter, Detlef Lindner, Matthias Lindemann, CCPR Key Comparisons K3a of
Luminous Intensity and K4 of Luminous Flux with Lamps as Transfer Standards, PTB
Bericht, PTB-Opt-62, 1999.
K3.b: R. Köhler, M. Stock, C. Garreau, Final Report on the International Comparison of
Luminous Responsivity CCPR-K3.b, Metrologia 41, 2004, Tech. Suppl., 02001.
Summary results are available at the BIPM Key Comparison Database (KCDB) at
www.bipm.org.
[2] Consultative Committee for Photometry and Radiometry (CCPR), Report of the 20th meeting
(17-18 September 2009) to the International Committee for Weights and Measures, Version
2: amended 13 April 2011, BIPM, Paris, file CCPR20.pdf available from www.bipm.org.
[3] CIPM MRA-D-05, Measurement Comparisons in the CIPM MRA, Version 1.3 October
2012, BIPM, Paris, file CIPM_MRA-D-05.pdf available from www.bipm.org.
[4] CCPR-G4, July 01, 2013 Guidelines for preparing CCPR Key Comparisons, CCPR WG-KC,
BIPM, Paris.
[5] Minutes of CCPR WG-KC meeting 9 July 2010, NPL, Teddington, UK, file
WGKC-10-Minutes.pdf, available from www.bipm.org.
[6] CCPR-G2 Rev.3, July 01, 2013 Guidelines for CCPR Comparison Report Preparation,
CCPR WG-KC, BIPM, Paris.
[7] JCGM 100:2008, Joint Committee for Guides in Metrology (September 2008), Evaluation of
Measurement Data — Guide to the expression of uncertainty in measurement (GUM).
Available from http://www.bipm.org . See also JCGM 104:2009, Evaluation of measurement
Data — An introduction to the “Guide to the expression of uncertainty in measurement” and
related documents.
[8] CIE Publication 198:2011, Determination of Measurement Uncertainties in Photometry,
Commission Internationale de l’Eclairage, Vienna, Austria.
[9] CIE Publication 198-SP1.1:2011, Supplement 1: Modules and Examples for the
Determination of Measurement Uncertainties, Part 1: Modules for the Construction of
Measurement Equations, Commission Internationale de l’Eclairage, Vienna, Austria.
9. Appendices
Luminous Intensity
Final Report
Appendix A
NMI Reports
Luminous Intensity
Final Report
Appendix A
NMISA Report
NMISA:
Model of evaluation:
See attached page for NMISA Uncertainty Budget Matrix (UBM).
𝐾𝐾𝑚𝑚 𝑑𝑑2 𝐹𝐹𝐼𝐼𝑐𝑐
𝐼𝐼 = 𝑆𝑆𝑆𝑆
where
I is the luminous intensity
Km is the luminous efficacy
d is the distance from the lamp filament to the photometer
F is the spectral mismatch factor
Ic is the current, determined for the gain of the amplifier and the voltage as measured
for the LMT photometer
S is the responsivity of the LMT photometer
A is the area of the LMT photometer
• Spectral mismatch: We corrected for spectral mismatch and therefore did not include it in
the model of evaluation.
• Lamp alignment: We allowed for 1° uncertainty in the alignment of the lamps, as you can see
in the model of evaluation.
Metrologist
Type & Serial
Description: CCPR-K3 Luminous Intensity Intercomparison Range:
Number
Mathematical Model:
▼ Unit Under Test / Calibration (Uncorrelated) ▼ NOTE! ONLY CHANGE BLUE CELLS - All OTHER CELLS (WHITE) ARE PROTECTED
UBM V5.03 - CSIR-National Metrology Laboratory - dc Low Frequency (Designed by BvO) Final UB for CCPR-K3.a Uncertainty Budget Matrix
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
NIM Report
Figure One
Schematic of NIM measurement configuration:
1 photometer; 2-6 baffles; 7 luminous intensity lamp; 8 radiation
trap.
The distance between the photometer input aperture and the lamp filament plane
on the NIM photometric bench is 3.2 m. The NIM photometer has an input aperture
diameter of approximately 9 mm. Thus the solid angle for the light emitted from the
lamp that is measured by the photometer is approximately 6 µsr.
In Figure One, 2-5 are baffles with rectangular aperture sizes of 50mm×50mm,
80mm×80mm, 150mm×150mm, and 200mm×200mm, respectively; 6 is a baffle
located 200mm before the lamp with a 50mm diameter circular aperture; a radiation
trap (8) made of a piece of black velvet cloth is mounted 2.0 m behind the luminous
intensity lamp (7) to ensure the reflectance to be less than 0.02%.
All lamps were operated with DC power at the fixed polarity and fixed current.
The electrical operating parameters of the lamps were measured using the standard
four-terminal measurement to permit an accurate measurement of the lamp operating
current and voltage. The voltage was measured at the lamp socket, rather than the
lamp base.
The lamp current was ramped up slowly over approximately one minute to the
specified value. The luminous intensity of the standard lamps was measured together
with the electrical values. After measurements, the lamp voltage was ramped down
slowly over approximately one minute.
After connecting the electrical power to the lamp, ten minutes warm-up
procedure for each lamp was followed.
The measurement is conducted on an 8 m photometric bench using a group of
eight BDQ8 luminous intensity lamps as reference to calibrate a group of six Wi41/G
lamps.
Lamp Wi41/G-152, Wi41/G-180, and Wi41/G-159 were calibrated on
26-01-2014 and 27-01-2014. All the six Wi41/G lamps were calibrated on 06-03-2014
and 11-03-2014. All the lamps and measurement facilities were reinstalled in each
measurement.
The measurement is conducted using an LMT full filter photometer, with the
following features:
- no temperature control is applied.
- the f 1 ’ is 1.2%.
- no diffuser is used.
- the diameter of the detector is about 9 mm.
regulator coincides with the photometric bench axis. The coordinate axis system is
same as that illustrated in Figure One of the Luminous Intensity Technical Protocol.
The origin of the coordinate system is established in NIM using an alignment
telescope. The alignment telescope is positioned such that its optical axis is identical
to the optical axis of the center of the regulator and perpendicular to the indicator
board. The telescope mount and the center regulator mount are adjusted such that the
crosshair of the indicator board coincides with the crosshair of the telescope at any
position along the X direction.
The spatial position of the lamp is defined as “For Center Filament Support #2”
of the protocol. The alignment of the filament was at the room temperature.
A special carriage having five degrees of freedom in its physical adjustments was
used for lamp position adjustment.
The vertical direction of the lamp was adjusted using the shadow of the filament
(Figure Two a). A collimated light beam projects the lamp filament on an indicator
board with a mark of line (plumb line) which is perpendicular to the optical axis. The
lamp was rotated about the Z-axis until the width (in the X direction) of the shadow of
the filament is minimized.
Figure Two
Filament alignment.
Rotation about the Y-axis is adjusted until the shadow of the filament on the
indicator board is parallel to the plumb line.
A regulator (Figure Two b) is used for the filament plane and the photometer
input aperture center alignment. The regulator is an optical imaging system consists of
a lens and an indicator board and its optical axis coincides with the optical axis of the
photometric bench. The lamp was adjusted until the filament center coincides with the
crosshairs in the indicator board (Figure Two c), so that the optical axis of the
photometric bench passes through the center of lamp filament plane.
During the measurement for this comparison the ambient temperature at NIM is
stabilized at 24.5℃ w ith fluctuations less than 1℃.T he hum idity is (35% -39%) RH.
BDQ8 (see Figure Three for an example) luminous intensity lamps are used as
laboratory transfer standards with the following features:
- current is about 6.6000 A.
- voltage is about 97 V.
- luminous intensity is about 1230 cd.
- CCT is about 2835 K.
These lamps are made in China.
Figure Three
A BDQ8 luminous intensity lamp
NMI: NIM
Signature:
Random effects:
Lamp parameters:
- lamp reproducibility(including lamp
A 0.047
alignment)
- lamp output fluctuations A 0.007
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.049
Random effects:
Lamp parameters:
- lamp ageing B 0.018
- lamp reproducibility(including lamp
A 0.077
alignment)
- lamp output fluctuations A 0.002
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.080
Random effects:
Lamp parameters:
- lamp reproducibility(including lamp
A 0.062
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.063
Random effects:
Lamp parameters:
- lamp ageing B 0.024
- lamp reproducibility(including lamp
A 0.077
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.081
Random effects:
Lamp parameters:
- lamp reproducibility(including lamp
A 0.006
alignment)
- lamp output fluctuations A 0.002
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.012
Random effects:
Lamp parameters:
- lamp ageing B 0.036
- lamp reproducibility(including lamp
A 0.056
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.068
Random effects:
Lamp parameters:
- lamp reproducibility(including lamp
A 0.039
alignment)
- lamp output fluctuations A 0.002
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.040
Random effects:
Lamp parameters:
- lamp ageing B 0.018
- lamp reproducibility(including lamp
A 0.057
alignment
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.061
Random effects:
Lamp parameters:
- lamp reproducibility(including lamp
A 0.046
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.047
Random effects:
Lamp parameters:
- lamp ageing B 0.024
- lamp reproducibility(including lamp
A 0.040
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.048
Random effects:
Lamp parameters:
- lamp alignment B 0.065
- lamp reproducibility(including lamp
A 0.001
alignment)
Electrical parameters:
-lamp current supply fluctuations B 0.006
Photometer noise B 0.01
RMS total random effects: 0.066
Random effects:
Lamp parameters:
- lamp ageing B 0.018
- lamp reproducibility(including lamp
A 0.081
alignment)
- lamp output fluctuations A 0.001
Electrical parameters:
NMI: NIM
Signature:
Specification:
The warm up time for each lamp is ten minutes.
The centre of the lamp is positive, the screw is negative.
The lamp socket has been changed since March 11th. We found that the electrical
connection characteristics of the old socket is not good, the characteristics of the
new one has been improved. In order to avoid increasing the lamp burning time, we don’t
measure too much. The measurement results of voltage in March 11th should be the lamp
voltage.
*Measurement with old lamp socket.
NMI: NIM
Signature:
Specification:
The warm up time for each lamp is ten minutes.
The centre of the lamp is positive, the screw is negative.
NMI: NIM
Signature:
Luminous Intensity
Final Report
Appendix A
NMIA Report
Luminous Intensity
Final Report
Appendix A
NMIJ Report
FAX: 613-952-1394
Email: Arnold.Gaertner@nrc-cnrc.gc.ca
In the calibration process at NMIJ, luminous intensity of a lamp is calibrated by a comparison with
a group of luminous intensity standard lamps. During measurement, a lamp and a photometer are
located on a photometric bench. A wall lying midway on the photometric bench separates the lamp
area and the measurement area. A hole on the wall allows the light from the lamp to go to the
measurement area where the photometer is placed. Baffles and a shutter are also placed between the
lamp and the photometer. The number of baffles is four. In the measurement of an Osram Wi41/G
lamp, a limiting aperture is used additionally to reduce stray light and to measure the light through
the rectangular mask of the lamp only. The measurement geometry is shown in Fig. 1.
The lamp alignment system consists of a lamp alignment stage and alignment apparatus such as a
laser and cameras. The lamp alignment stage is composed of six stages to adjust the lamp positions
along the X, Y, and Z axes and the rotation angles of θX, θY and θZ. The lamp alignment procedure is
described later.
The position of the photometer is determined so that the laser beam that coincides with the optical
axis comes to the center of the shielding tube of the photometer.
The sizes and the distances of the limiting aperture, the hole on the wall, the baffles, and the shutter
are shown in Fig. 1.
The distance between the center of the lamp filament and the entrance diffuser of the photometer
(reference plane) is 2.7 m. The diameter of the entrance diffuser of the photometer is 40 mm.
Therefore, the solid angle formed by the entrance diffuser of the photometer is about 1.7×10-4 rad.
Wall
with Baffles
Limiting Entrance
aperture hole Shutter diffuser
Shielding
Lamp tube
Photometer
Stages
0.0 0.50 1.08 1.24 1.50 1.77 2.01 2.39 2.60 2.70 Distance [m]
75 250 200 150 125 100 75 50 40 Diameter [mm]
The luminous intensity of a traveling lamp Ii (i=1 to 6) is determined as the average of two values
derived separately from the calculations based on the individual luminous intensity of the standard
lamps.
The value to be calculated from one standard lamp for a traveling lamp is the product of three
quantities, i.e., the luminous intensity of the standard lamp, the ratio of the output signal for the
traveling lamp to that for the standard lamp, and the color correction factor. Therefore, the luminous
intensity of the traveling lamp is obtained in the following equation.
1 V V
I i = k ai i I a + k bi i I b , (1)
2 Va Vb
where Ii is the luminous intensity of the i-th traveling lamp, Ia and Ib are the luminous intensity of the
standard lamps SA and SB, Vi, Va, and Vb are the output signals of the i-th traveling lamp, the standard
lamps SA and SB, and kai and kbi are the color correction factors between the i-th traveling lamp and
SA and SB, respectively. As mentioned above, the output signals used here are the averages of the
measurement sequences "Go" and "Return" for the respective lamps.
*) In this final measurement report, the data of the sixth lamp (No. 69) is excluded because of the
large discrepancy in luminous intensity between the first round measurement and the second
round measurement measured at NMIJ, which implies unexpected instability of this lamp.
The photometer used for the calibration is manufactured by Kouno Kouki Sangyou KK, which has
closed down its business already. The photometer consists of an 100 mm diameter integrating sphere
with an entrance diffuser (matte opal glass) and three filtered Si photodiodes; B(blue), Y(yellow),
and R(red) detectors. Spectral responsivity of the Y detector is approximated to V(λ) whose V(λ)
mismatch index f1' is 2.11. The output signal from the Y detector is used for the calibration of
luminous intensity. Spectral responsivity of the B and R detector has the peak around 460 nm and
660 nm, respectively. They are used to check the distribution temperature of lamps to be measured.
The lamp filament is observed with two cameras. One camera is located along with the X axis (
optical axis) to see the rear view of the filament, and the other is located along with the Y axis to see
the side view. The coordinate system is taken to agree with the description in the protocol. Each
view has cross hairs pre-aligned to coincide with the origin of the coordinate system and the
coordinate axes. The lamp alignment is made in such a way that the filament position comes in
accordance with these cross hairs.
The rotation about the Z axis (θZ) is adjusted so that the shape of the filament in the side view
becomes narrowest. Care must be taken when aligning the traveling lamps #58 and #69, because the
wires that compose the filament of both lamps are slightly uneven and not formed the single plane.
For these lamps, the angle about the Z axis (θZ) is adjusted so that the widths of the upper half and
lower half of the filament is balanced (see Fig. 2).
The rotation about the X (θX) axis is adjusted so that the rectangular shape of the filament in the
rear view stands upright.
Then the rotation about the Y axis (θY) is adjusted so that the filament in the side view coincides
with the vertical axis. Care must also be taken when aligning the traveling lamps #58 and #69
because the shapes of the filaments seem slightly curved. For these lamps, the angle about the Y axis
is adjusted so that the line fitted to the shape of the filament coincides to the Y axis.
The height of the filament is adjusted so that the half height of the filament coincides with the
horizontal line of the cross hair that is in accordance with the origin of the coordinate system.
The position of the filament along with the Y-axis is adjusted so that the origin of the coordinate
system comes to the center between the 6th and 7th wires of the filament. The wires are numbered as
shown in Fig. 3.
1 2 3 4 5 6 7
The position of the filament along with the X-axis is adjusted so that the center of the filament
coincides with the origin of the coordinate system.
The size and position of the limiting aperture are described in Fig. 1.
For all the traveling lamps, electrical polarity is defined so that the negative pole is connected to the
center of the socket when electrical current is supplied. The measured lamp current and voltage for
each traveling lamp are as follows. The lamps are operated with the constant current mode.
#37 5.7563 A 29.0689 V
#40 5.7943 A 29.5493 V
#51 5.7362 A 29.2641 V
#52 5.7646 A 29.1673 V
#58 5.6101 A 29.9704 V
#69 5.6198 A 29.9225 V
When turning on a traveling lamp, electrical current is increased gradually from zero to the fixed
value in two minutes. After the lamp current reaches the fixed value, the warm-up time of 13
minutes is applied before measurement.
A schematic diagram of the luminous intensity calibration facilities is shown in Fig. 4. A lamp
voltage is measured by a 8.5-digit digital multimeter. A lamp current is determined by measuring the
voltage between the terminal of the standard shunt resistor that has a calibrated resistance. An output
of a DC power supply for the lamp is regulated by a voltage/current source, whose output is
controlled by the software with feedback-control using signals from the multimeters to stabilize the
lamp voltage (constant voltage mode, for NMIJ standard lamps) or current (constant current mode,
for travelling lamps: Osram Wi41/G). An amplified output of the photometer is measured by another
8.5-digit digital multimeter.
The luminous intensity scale of NMIJ is maintained with four standard lamps. The type of the
standard lamps is Toshiba 55 V-330 W coil-M-type luminous intensity standard lamp. The
traceability diagram is shown in Fig. 5. Two lamps are used as the luminous intensity standard lamps
and another two are used as the luminous intensity working standard lamps according to Fig. 5. In
this comparison, the traveling lamps are directly calibrated with the luminous intensity standard
lamps.
The date of last realization of the NMIJ primary scale is Jan. 30th, 1998. The uncertainty budget is
shown in Table 1.
Signature:
Lamp number: 37
Signature:
Lamp number: 40
Signature:
Lamp number: 51
Signature:
Lamp number: 52
Signature:
Lamp number: 58
Signature:
Lamp number: 69
Signature:
Random effects:
Lamp parameters:
- lamp ageing B 0.11
- lamp alignment (*) B 0.06
- lamp reproducibility included in (*)
- lamp output fluctuations B 0.02
Electrical parameters:
- power supply fluctuations included in (*)
Photometer noise included in (*)
(Measurement Set standard deviation of mean)
Signature:
The following equations are the physical model of uncertainty of luminous intensity at NMIJ.
λ2
K (d + ∆d1 ) V0
I1 = m 1
2
∫λ Φ
1
e ,λ (λ ) V (λ ) dλ
(1 + ct ) (1 + c1 ) (1)
λ
A G
∫λ Φ (λ ) s e (λ ) dλ
2
e ,λ
1
V2 (d 2 + ∆d 2 )
2
I 2 = I1 k c 2
(1 + ci ) (1 + ca ) (1 + c2 ) (1 + c3 ) (2)
V1 d2
Equation (1) is the model to determine the luminous intensity of the standard lamp. Equation (2) is the
model to transfer luminous intensity from the standard lamp to the transfer lamp. The meanings of each
variable are listed below.
d1 : Distance between the standard lamp and the standard photometer. Constant. No uncertainty.
negligible.
V (λ ) : Luminous efficiency function. No uncertainty.
se (λ ) : Spectral responsivity of the standard photometer. Uncertainty of this factor consists of two parts
in the budget. One is "Spectral responsivity of the silicon photodiode measured with the cryogenic
radiometer", and another is " Illuminance responsivity of the standard photometer with respect to the
spectral responsivity of the silicon photodiode".
ct : Deviation of the standard photometer responsivity by the room temperature.
c1 : Deviation of the luminous intensity measurement for the standard lamp set on and removed from the
lamp mount in many times. Accumulated data.
kc : Colour correction factor between the standard lamp and the transfer lamp. Uncertainty negligible.
ca : Deviation of luminous intensity through the period of recalibraion-limit burning time. We take this
effect into the uncertainty without correction. So it is listed in "Random effects" because we cannot predict
what value a lamp will take at each burning.
c 2 : Deviation of the luminous intensity measurement for the transfer lamp set on and removed from the
lamp mount in many times. Accumulated data.
c3 : Fluctuation of lamp signal.
Random effects:
Lamp parameters:
- lamp ageing B 0.11
ca
- lamp alignment (*) B 0.06
c2
- lamp reproducibility included in (*)
- lamp output fluctuations B 0.02
c3
Electrical parameters:
- power supply fluctuations included in (*)
Photometer noise included in (*)
(Measurement Set standard deviation of mean)
The effect of baffles is regarded as negligibly small. We expect that that effect can be as small as 0.007 %,
which is negligible in the NMIJ's uncertainty budget.
Lamp Number: 37
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.7563 29.069 242.14 cd 0.127 0.256
2014/11: A 2V
02am
2 2 Feb. 8th, 5.7563 29.068 242.15 cd 0.127 0.256
2014/9:5 A 9V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.7563 29.064 242.20 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.7563 29.064 242.11 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Lamp Number: 40
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.7943 29.550 250.55 cd 0.127 0.256
2014/11: A 5V
02am
2 2 Feb. 8th, 5.7943 29.549 250.46 cd 0.127 0.256
2014/9:5 A 3V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.7943 29.543 250.34 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.7943 29.544 250.23 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Lamp Number: 51
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.7362 29.264 240.91 cd 0.127 0.256
2014/11: A 6V
02am
2 2 Feb. 8th, 5.7362 29.264 240.79 cd 0.127 0.256
2014/9:5 A 1V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.7362 29.261 240.54 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.7362 29.262 240.59 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Lamp Number: 52
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.7646 29.167 241.51 cd 0.127 0.256
2014/11: A 6V
02am
2 2 Feb. 8th, 5.7646 29.167 241.49 cd 0.127 0.256
2014/9:5 A 3V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.7646 29.160 241.48 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.7646 29.159 241.50 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Lamp Number: 58
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.6101 29.968 244.27 cd 0.127 0.256
2014/11: A 9V
02am
2 2 Feb. 8th, 5.6101 29.970 244.29 cd 0.127 0.256
2014/9:5 A 4V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.6101 29.965 244.50 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.6101 29.964 244.51 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Lamp Number: 69
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Feb. 2nd, 5.6198 29.920 243.26 cd 0.127 0.256
2014/11: A 4V
02am
2 2 Feb. 8th, 5.6198 29.922 243.09 cd 0.127 0.256
2014/9:5 A 5V
7am
Standard Uncertainty
Number of in Luminous Intensity
Measurement Lamp Lamp Luminous
measurements Date/time (%)
Set Number current voltage Intensity
per set
Random Systematic
1 2 Mar. 5.6198 29.915 244.14 cd 0.127 0.256
21st, A V
2015/11:
31am
2 2 Mar. 5.6198 29.915 244.74 cd 0.127 0.256
27nd, A V
2015/9:4
2am
Signature:
Luminous Intensity
Final Report
Appendix A
IO-CSIC Report
1.- Introduction
This report describes the procedure followed at IO-CSIC (covering first round of
measurements) to determine the luminous intensity of six lamps: four OSRAM Wi 41/G
identified as: Wi95A,Wi95B,Wi95C and Wi95D; and two NPL/GEC lamps (now called
NPL/Polaron Heavy Current LIS incandescent lamps) identified as: A-454 and A-456.
The lamps have been calibrated in an optical bench, measuring the illuminance
with a V(λ) corrected detector. Two standard photometers have been used as
reference, which were calibrated for absolute responsivity against our cryogenic
radiometer; and for relative responsivity against the IO-CSIC spectral responsivity
scale.
Full description of the method used for the realization of the candela is reported
in “Realization of the candela from a partial filtering V(λ) detector traceable to a
cryogenic radiometer” by J. Campos, A. Corróns, A. Pons and P. Corredera.
Metrologia. 1995, 32 and in “Luminous intensity standard based on a cryogenic
radiometer” CIE 119-1995-23rd Session. New Delhi. Volume 1, 102-105
A laser beam has been used to visualize the optical axis in order to simplify the
positioning and orientation of lamps and detector. A second laser beam, intercepting
the first one at an angle of 45º approximately, has been used to define the reference
plane for the measurement of illuminance, which is normal to the optical axis at the
point where both laser beams crossed. Photometers were placed at the reference
plane at normal incidence by using high precision positioning equipment. Lamp
reference plane, as fixed in the measurement specifications, has been defined with a
third laser beam. Figure 1
The lamps have been aligned with a cold (room temperature) filament. Next table
shows individual values for electrical supply of each lamp and the measured Colour
Temperature.
6.- UNCERTAINTY
m.mT − m I
c jV j
m
TR
.(d + ∆d p + ∆d L ) .(1 − cstray − ε + hϕ + fυ ) )
cV
I= v 2
(1)
J R
Sv TA R j
Quantities:
Enclosed please find a the Anex 5 with an extra column with the classification of uncertainty
components,
Regards,
Alicia Pons
Instituto de Optica-CSIC
Serrano 144, 28006 Madrid
Tf. 91 5618806; Fax. 91 5642122
e-mail: alicia.pons@csic.es
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 100 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
1.- Introduction
This report describes the values obtained in the calibration, in second round, of
five lamps: four OSRAM Wi 41/G identified as: Wi95A,Wi95B,Wi95C and Wi95D; and
one NPL/GEC lamps (now called NPL/Polaron Heavy Current LIS incandescent lamps)
identified as: A-454. Polaron-type lamp identified as A-456 failed during the
measurement at pilot laboratory.
Shown values of current and voltage are effective values of electrical supply used in
our laboratory, during the second round. They are mean values of the three sets of
measurements made. The lamp current is measured as voltage across a shunt
resistance.
Laboratory conditions
For the second round the measurements were performed during July 2015. Laboratory
temperature was always between 24 ºC to 25 ºC. Humidity was in the range 30%-40%.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 101 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 102 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 103 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 104 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
2016-September-09
Dear Arnold,
I have been analyzing the relative data of IO-CSIC and I have noticed some problems with data
of Round #2 of our lamp identified as A-454.
After a detailed revision of the measurements and the calculus, I have noticed that there was
an error in the data sent to you corresponding to round #2 of A-454 lamp. The problem was
that in the measurements of this lamp we used a different standard resistor with a different
value. At the time we calculate the luminous intensity I made a mistake.
Enclosed please find the corrected values (marked in red) for this lamp.
Random Systematic
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 105 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
LNE-CNAM Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 106 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
LNE-CNAM — Appendix A3
Baffles
Standard Light trap
Photometer lamp
i z
y
I γ β
x
α
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 107 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Were Iph (T) , luminous intensity when the lamp is operated perfectly
T , colour temperature of the lamp
d , distance between the lamp filament and the photometer limiting aperture.
Sph(T) , absolute sensitivity of the photometer at the colour temperature T.
dL , misplacement of the lamp.
dph , misplacement of the photometer.
yph , photocurrent corrected for straylight and offset y0
cor , correction factor with about unity value.
εph , misorientation of the photometer
εL , misorientation of the lamp
∆J , relative difference in the lamp current setting.
γ , ageing coefficient of the lamp.
∆t , burning time
Measurement uncertainty
The lamps run at a colour temperature of (2800 ± 15) K. The sensitivity of the photometers is
calculated at 2800K. The slope of the spectral mismatch factor according to the CCT is low. The
uncertainty component associated to the colour temperature is neglected
The other contributions to the combined uncertainty are summarised in Appendix A5.
Uncertainties on the photometer sensitivity and repeatability of the lamps are the main parts of
the combined uncertainty.
Description of the primary photometers
The photometers are made with four mains parts, the trap detector, the filter holder, the filter and
the aperture. The trap detector is made with three identical silicon photodiodes of 18x18mm,
provided by Hamamatsu. They are oriented in order to trap the light after 5 internal reflexions
and to minimize the polarization effects.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 108 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
The V(λ) filters are elaborated in the institute using different Schott glasses (GG10, BG39, FG13,
VG4). We have 2 types of filters, made with 3 or 4 layers. As the transmittance of a filter is
strongly dependent on its temperature, we designed a dedicated holder. The filter is fixed inside a
copper oven and the temperature of the oven is regulated at 23°C with a Peltier element and an
external controller. The temperature stability in the oven is better than 0.1 °C.
A precision aperture defines the illumination area. We have three diameters of aperture (10mm,
8mm and 6mm).
The table below details the different elements associated in the photometers.
Photometer ref Filter Aperture Solid Angle
PH04A 4 layers ∅ 10 mm 10.5 µsr
PH04B 4 layers ∅ 8 mm 6.7 µsr
PH04C 3 layers ∅ 10 mm 10.5 µsr
The absolute sensibility of the trap radiometer is calibrated according to our cryogenic radiometer
at 3 laser wavelengths (514.53 nm, 543.36 nm, 611.80 nm). The calibration is extended between
380 and 780 nm on our relative spectral sensitivity measurement facility.
The V(λ) is calibrated on the visible domain using our primary transmittance measurement
facility. The area of the diaphragms have been calibrated on our dedicated facility.
The presentation of theses facilities and the intermediate uncertainty budget for all these
calibration steps can be found in the 2 following publications
OBEIN, G., GONZALEZ-GALVAN, L., BASTIE, J., 2007, Nouvelle réalisation de la candela au
LNE-INM/CNAM, revue française de métrologie, 12, p19-28.
OBEIN, G., GONZALEZ-GALVAN, L., BASTIE, J., 2007, A new realization of the candela at the
Lne-Inm/Cnam, Proceedings of the 26th session of the CIE, Vol. 1, part. 1,
pp192-195.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 109 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
The 6 lamps have been sent by private transporter at NRC after the first round. The box arrived at
NRC in good shape, but unfortunately, 2 polaron lamps (ref A431, A434) were broken inside.
The lamps were in a wood box specially designed to protect them. They were in the institute
since more than 20 years, and had already travelled many times for CCPR or EURAMET key
comparisons. The shock during the transport must has been of high violence to brake the boxes
and the lamps.
After discussion with the pilot lab, we took the decision to maintain the comparison on a
restricted set of the 4 remaining lamps (A430, #926, #927, #936).
Unfortunately, after round2 measurements, it appears that lamp #927 shows a drift of 0.88%
between Round1 and Round2. This comportment is abnormal for such a lamp. We believe that
this evolution might be due to the shock during the transport. We proposed to the pilot lab to
remove that lamp also.
Operating conditions of the lamps
Lamp Current Voltage CCT
[A] [V] [K]
926 5,690 29,01 2796
927 5,690 28,97 2795
936 5,690 29,15 2799
A430 25,000 11,95 2815
All the lamps are aligned at room temperature.
— Wi41/G lamps are aligned as in case #2 (w/2, h/2) described in the protocol section 4.4.8
— LIS Polaron lamps are aligned with retro-reflection of a laser on the front of the bulb.
The negative pole of the power supply is connected to the central electrical foot contact of the cap
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 110 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Random effects:
Lamp parameters:
- lamp ageing A 0.04
- lamp alignment A 0.21
- lamp reproducibility A 0.10 (typical)
Electrical parameters:
- power supply fluctuations A 0.01
Photometer noise A <0.001
The RMS total refers to the usual square root of the sum of the squares of all the individual
uncertainty terms.
Participant: LNE-CNAM
NMI: France
Date: 27/11/2015
Signature: ………………………………
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 111 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Lamp Number:
The random/systematic labels in this table are those related to the measurements within the
particular round of the measurements. If the systematic factors change between the measurement
rounds, this information should be indicated separately.
Signature: ....................................
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 112 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
METAS Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 113 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
METAS data:
The items listed on this form should be used as a guide. It is anticipated that many of the
questions will require more information than the space allocated on this page. Please expand
your reply document as necessary.
Description of measurement geometry (please include a diagram):
- positions of lamp, detector, bench, shielding, baffles (number, distances and sizes)
fig. 1
- alignment devices
Laser beam, telescopes
- limiting aperture?
The photometers have a sharp limiting aperture defining the reference plane
METAS built reflective trap detector based on 3 large sized hamamatsu Si photodiode (S6337).
Each photometer has a precis aperture and a V(λ)-matching filter . The f'1 of the photometer
being approximately 1.8. The filters are temperature stabilized
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 114 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
- geometrical alignment
The optical axis is horizontal and passes through the center of the filament.
- for Osram lamps with center filament supports, which center filament support type is used for
the alignment (see Figure Two and Section 4.4.8.)
CenterFilament Support #1
- alignment procedure
The spatial position of the lamp is adjusted in the Y direction until the vertical crosshair of the
telescope is equidistant from the two filament wires at the center of the filament.
The spatial position of the lamp is adjusted in the Z direction until the horizontal crosshair of
the telescope passes through the defined point of the filament plane.
Rotation about the Z-axis is adjusted until the width (in the Y direction) of the image of the
filament in the telescope is minimized. In the case of the filament with the center support, only
the top half of the filament will be visible for this alignment.
The distance along the X-axis is measured to the center along the X-axis of the image of the
lamp filament in the telescope.
- length of warm-up time for each lamp before measurements are taken > 15 min.
- measured CCT (or Distribution Temperature or Colour Temperature, see Section 3.5).
See table
- stray-light reduction
three apertures are placed in the path between the lamp and the photometers (see also
figure above): at around 100 mm from the photometer a round aperture of 42mm diameter, at
960 mm a round aperture of around 120mm and at 2000mm a round aperture of 100 mm. In
addition a light trap is placed at around 725 mm after the lamp.
Establishment or traceability route of primary scale including date of last realisation and
uncertainty budget.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 116 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Vmm0
VICCS1
2* mean val photo. signal / (V) 0.73384 0.00013 A 9 376.7446 0.0479 0.017%
yPS10
3* mean val photo. dark signal / (V) -0.000438 0.000018 A 9 -376.7446 -0.0068 0.002%
4
luminous respons photometer / sCP1 16.580 0.051 B ∞ 16.6746 0.8579 0.310%
(nA/ lx)
6
gain setting resistance, GPS1 1001.36 0.02 B ∞ -0.2761 -0.0044 0.002%
photometer picoammeter / Kohm
distribution temperature of lamp / TdC1
7* 2855.7 11.5 B ∞ -0.0007 -0.0086 0.003%
K
nominal distribution temperature, TA
8 2856 0 B ∞ 0 0 0.000%
"illuminant A" / K
spect. mismatch factor for mP1
9 0.0077 0.0003 B ∞ -1.09E-02 0.0000 0.000%
photometer mPR
J C1
12 nominal current for the lamp / A 5.76 0 B ∞ 0 0 0.000%
18
deviation to nominal ambient ∆TP 1 0.28 B ∞ 0.0553 0.0155 0.006%
temperature DUT/ K
19*
angular misalignment of DUT g P (ε P ) 0 0.00007 B ∞ 276.4719 0.0194 0.007%
photometer head
20*
distance alignment of lamp ∆dS1 d PS 0 0.00023094 B ∞ -552.9438 -0.1277 0.046%
filament
relative temperature coefficient of α S 1
21 0.0002 0.0001 B ∞ 0.0000 0.0000 0.000%
lamp / K-1
22
deviation to nominal ambient ∆TS1 0 0.28 B ∞ -0.0553 -0.0155 0.006%
temperature ref/ K
23*
angular misalignment of ref k S 1 (ϕ S 1 ) 0 0.000054 B ∞ -276.4719 -0.0149 0.005%
photometer head
24*
angular misalignment of ref hS 1 (ϑS 1 ) 0 0.0002 B ∞ -276.4719 -0.0553 0.020%
photometer head
sCP1
luminous intensity lamp / (cd) 276.47 DOF 1.0E+06 0.8761 0.317%
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 117 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 118 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 119 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Blattner Peter
Digitally signed by Blattner Peter DN: cn=Blattner Peter, o=METAS, ou,
email=peter.blattner@metas.ch, c=CH
Date: 2015.11.07 22:23:01 +01'00'
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 120 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
NPL Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 121 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
APRIL 2016
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 122 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 123 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 124 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Extracts from this report may be reproduced provided the source is acknowledged
and the extract is not taken out of context.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 125 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CONTENTS
1 INTRODUCTION ..........................................................................................................................1
2 SELECTION OF COMPARISON ARTEFACTS ......................................................................1
3 MEASUREMENT FACILITY ......................................................................................................1
3.1 LAMP ALIGNMENT ....................................................................................................................3
4 RESULTS ........................................................................................................................................5
5 UNCERTAINTY BUDGET...........................................................................................................6
5.1 CALIBRATION OF REFERENCE LAMP INTENSITY .............................................................7
5.2 AGEING OF REFERENCE LAMPS ............................................................................................7
5.3 DISTANCE SETTING AND BENCH SCALE.............................................................................7
5.4 ACCURACY OF LAMP CURRENT SETTING ..........................................................................8
5.5 PHOTOMETER PHOTOCURRENT ............................................................................................8
5.6 PHOTOMETER SPECTRAL MISMATCH .................................................................................8
5.7 STRAY LIGHT..............................................................................................................................8
5.8 STABILISER CURRENT CONTROL..........................................................................................8
5.9 PHOTOMETER CALIBRATION FACTOR ................................................................................8
5.10 TEST LAMP REPEATABILITY ..................................................................................................8
6 REFERENCES ...............................................................................................................................9
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 126 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 127 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
1 INTRODUCTION
NPL is one of twelve laboratories that has participated in the CCPR Key Comparison CCPR
K3:2014 for Luminous Intensity which commenced in 2014. This report summarises the
results of the measurements performed at NPL of the selected Luminous Intensity Standard
lamps.
The comparison protocol called for luminous intensity standard lamps as the comparison
artefacts. The nominated lamp types were Osram Wi41/G lamps and NPL/Polaron LIS lamps.
Participants were requested to submit between four and six lamps which could be of either
one or both types. NPL chose to submit five lamps as travelling artefacts, three Polaron type
and two Osram type, as indicated in Table 1.
3 MEASUREMENT FACILITY
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 128 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
An NPL-designed photometer was used to transfer the calibration from the NPL primary
standard luminous intensity lamps to the travelling standard lamps used for the comparison.
The photometer consists of a single element silicon photodiode with a four-element glass
filter, which has been individually-designed to give a close match to the V(λ) function, and a
precision aperture (10 mm in diameter) which defines the photometer reference plane for
distance measurement. The photometer was aligned so that the aperture was perpendicular to,
and centred on, the optical axis. The complete unit was temperature-controlled at 30.0 ºC; the
spectral responsivity of the photometer has also been calibrated at this temperature.
Stray light was minimised using a series of baffles between the lamp being measured and the
photometer (three baffles were used, placed approximately 200 mm, 1900 mm and 2100 mm
from the limiting aperture of the photometer – exact placement of the baffles depended on a
visual assessment of stray light reaching the photometer). An additional baffle was placed
immediately in front of the lamp so that only light passing through the lamp mask could reach
the photometer. A black cloth screen was placed between the rear of the lamp and the
alignment laser to eliminate reflections from the laser aperture. The walls, floor and ceiling of
the laboratory are painted black. Residual stray light was allowed for by making a ‘dark’
measurement with the stray light screen closest to the lamp obstructed.
Measurements were made with the lamps aligned as described in Section 3.1 below with the
photometer set so that its limiting aperture was at a distance of 2.4 m from the mean plane of
the lamp filament; this gave a measurement solid angle of approximately 0.12 sr. The
reference standards used were of the same type as the travelling standards and were aligned in
the same way and measured at the same distance. The reference standards were directly
traceable to NPL’s cryogenic radiometer and were established as described in [1]. The NPL
luminous intensity scale has been re-established directly against the cryogenic radiometer on a
regular basis since the time of the first realisation described in [1] and this has confirmed the
stability of the disseminated scale over this period. The calibration of the reference lamps has
also been checked at regular intervals and found to be within the limits allowed for ‘lamp
ageing’ in the uncertainty budget.
The lamps were operated from a stabilised dc power supply with a current stability of better
than 0.005 %. Current was determined by measuring the voltage drop across a calibrated
precision resistor (0.1 Ω in the case of the Osram lamps, 0.01 Ω for the Polaron lamps); the
resistors were used in an oil bath to minimise any temperature fluctuations during use and
were calibrated at the same temperature at which they were used. Each lamp had previously
been calibrated against the NPL relative spectral irradiance scale to determine the current
required for a correlated colour temperature of approximately 2856 K and this was the current
set for this comparison; the individual lamp currents set and the corresponding correlated
colour temperatures are given in Table 2. The reference lamps were also operated at a
correlated colour temperature of approximately 2856 K and the close match of the photometer
spectral responsivity to the V(λ) function meant that the spectral mismatch correction was
negligibly small.
At least two independent measurements were made on each lamp, with the lamp being
completely realigned for each measurement. Measurements were made over a period of
several days. On each occasion of measurement the lamp was run up gradually to the required
current and allowed to stabilise for at least 15 minutes before measurements commenced. The
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 129 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
lamps were operated in a 4-pin lamp holder and the lamp voltage and current at the time of
measurement were recorded, together with the photocurrent from the photometer.
Table 2 Currents and correlated colour temperatures for NPL travelling artefacts
Correlated colour
Current
Lamp identifier temperature
(A)
(K)
A644 25.360 2850
A647 25.310 2850
PA758 25.220 2850
877 5.818 2853
890 5.804 2853
During the course of the measurements the laboratory was maintained at a temperature of
21.0 ºC ± 2.0 ºC and a humidity of 50 % RH ± 25 % RH.
The lamps were mounted base down and aligned with a cold filament (i.e. no current
flowing).
The alignment procedure used for the Osram lamps was as follows:
1. The lamp was adjusted so that the filament was vertical when viewed along the optical
axis.
2. The lamp was rotated so that the width of the image of the filament viewed through
the cathetometer set perpendicular to the optical axis was minimised.
3. The tilt of the lamp was adjusted so that the image of the filament was vertical (i.e.
parallel to the cathetometer vertical cross hair)
4. The lamp was adjusted in the horizontal and vertical direction so that the laser
defining the optical axis passed through the centre of the filament.
5. A screen was placed immediately in front of the lamp so that only light passing
through the aperture in the painted lamp mask could reach the photometer.
6. Distance was measured from the mean plane of the lamp filament as viewed using the
cathetometer mounted perpendicular to the optical axis of the optical bench. The
measurement distance was 2.40 m.
The alignment procedure used for the Polaron lamps was as follows:
1. The lamp was adjusted so that the envelope was vertical when viewed along the
optical axis.
2. The lamp was rotated and tilted so that the flat front window was set perpendicular to
the optical axis (i.e. so that the laser defining the optical axis was reflected from the
rear of the front window directly back to the laser).
3. The lamp was adjusted in the horizontal and vertical direction so that the laser
defining the optical axis passed through the centre of the filament.
4. A screen was placed immediately in front of the lamp so that only light passing
through the aperture in the lamp mask fixed to the front window could reach the
photometer.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 130 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
5. Distance was measured from the mean plane of the lamp filament as viewed using the
cathetometer mounted perpendicular to the optical axis of the optical bench. The
measurement distance was 2.40 m.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 131 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
4 RESULTS
Standard Uncertainty in
Luminous Intensity (k = 1)
Number of
Source Lamp Lamp Luminous independent
Random Systematic
Identifier Current Voltage Intensity measure-
ments
(A) (V) (cd) (%) (%)
A644 25.360 12.505 451.78 3 0.082 % 0.158 %
A647 25.310 12.510 459.43 2 0.082 % 0.158 %
PA758 25.220 12.743 460.33 3 0.082 % 0.158 %
877 5.818 30.013 276.34 2 0.082 % 0.158 %
890 5.804 29.871 273.93 3 0.082 % 0.158 %
Standard Uncertainty in
Luminous Intensity (k = 1)
Number of
Source Lamp Lamp Luminous independent
Random Systematic
Identifier Current Voltage Intensity measure-
ments
(A) (V) (cd) (%) (%)
A644 25.360 12.500 451.97 2 0.082 % 0.158 %
A647 25.310 12.533 459.63 2 0.082 % 0.158 %
PA758 25.220 12.751 460.70 2 0.082 % 0.158 %
877 5.818 30.013 275.91 3 0.082 % 0.158 %
890 5.804 29.878 273.24 2 0.082 % 0.158 %
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 132 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
5 UNCERTAINTY BUDGET
Table 5 Uncertainty budget (identical for both rounds of measurements and both types of lamp)
Type A
Source of uncertainty or value Divisor ui
Type B
Systematic effects:
Calibration of reference lamp intensity B 0.200 % 2 0.100 %
Ageing of reference lamps B 0.125 % 1.732 0.072 %
Distance setting B 0.050 % 1.732 0.029 %
Accuracy of lamp current setting B 0.160 % 1.732 0.092 %
Photocurrent measurement B 0.010 % 1.732 0.006 %
Spectral mismatch B 0.010 % 1.732 0.006 %
Stray light B 0.020 % 1.732 0.012 %
RMS Total Systematic Effects 0.158 %
Random effects:
Stabiliser current control A 0.016 % 2 0.008 %
Photometer calibration factor repeatability A 0.064 % 1 0.064 %
Test lamp repeatability A 0.050 % 1 0.050 %
RMS Total Random Effects: 0.082 %
The basis of these uncertainties is described in 5.1 to 5.10 below. The associated measurement
equation is:
𝐼𝐼v,𝑡𝑡 = 𝐶𝐶cal 𝑉𝑉𝑡𝑡 (1 + 𝐶𝐶d,𝑡𝑡 )(1 + 𝐶𝐶J,𝑡𝑡 )(1 + 𝐶𝐶p,𝑡𝑡 )𝐹𝐹SM,𝑡𝑡 (1 − 𝐶𝐶stray,𝑡𝑡 )(1 + 𝐶𝐶align,𝑡𝑡 ) (1)
where
(𝐼𝐼v,𝑟𝑟 +𝐶𝐶age,𝑟𝑟 )
𝐶𝐶cal = (2)
𝑉𝑉𝑟𝑟
and
𝐼𝐼v,𝑡𝑡 is the luminous intensity of test (comparison) lamp t
𝐶𝐶cal is the mean photometer calibration factor, calculated using Equation 2 and averaged
across all the reference lamps used
𝐼𝐼v,𝑟𝑟 is the luminous intensity of reference lamp r
𝐶𝐶age,𝑟𝑟 is the change in luminous intensity of reference lamp r since its original
calibration due to ageing
𝑉𝑉𝑟𝑟 is the mean reading from the photometer for reference lamp r
𝑉𝑉𝑡𝑡 is the mean reading from the photometer for test lamp t
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 133 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
𝐶𝐶d,𝑡𝑡 is the error in luminous intensity for test lamp t due to error in setting the filaments
of the reference and test lamps in the same vertical plane
𝐶𝐶𝐽𝐽,𝑡𝑡 is the error in luminous intensity for test lamp t due to error in setting the current for
the test lamp to the specified value (the uncertainty due to error in setting the current for
the reference lamp to the specified value is included in the uncertainty budget for the
luminous intensity of the reference lamp)
𝐶𝐶p,𝑡𝑡 is the error in luminous intensity for test lamp t due to differences in amplifier gain
and DVM sensitivity between measurement of the photocurrent produced by the
reference lamp and that produced by the test lamp
𝐹𝐹SM,𝑡𝑡 is the spectral mismatch correction factor for test lamp t
𝐶𝐶stray,𝑡𝑡 is the error in luminous intensity for test lamp t due to differences in stray light
between the reference and test lamps
𝐶𝐶align,𝑡𝑡 is the error in luminous intensity for test lamp t due to misalignment of the lamp
(the uncertainty due to misalignment of the reference lamp is included in the uncertainty
budget for the luminous intensity of the reference lamp)
Note all of the C terms listed above have an expected value of zero and an associated
uncertainty that has been estimated as described in sections 5.1 to 5.10 below.
The calibration of the reference lamps, and the associated uncertainties, is detailed in [1].
The reference lamps are used only for checks to confirm the stability of the NPL luminous
intensity scale and as standards for calibration of working standards. They have been used for
a maximum of 25 hours since their initial calibration (most have been used for less than this)
and their polarity is reversed on each occasion of use to minimise ageing effects.
Measurements on other lamps of the same type operated at the same correlated colour
temperature and under the same conditions have shown ageing rates of approximately 0.5 %
per 100 hours; a conservative allowance for ageing of 0.125 % has therefore been included in
the uncertainty budget to allow for ageing. Regular checks using the NPL photometer as a
reference (freshly calibrated against the cryogenic radiometer) have confirmed the reference
lamps have been stable to within this limit.
The reference lamps and the travelling standards are both used at the same distance, which is
also the distance at which the reference lamps were originally calibrated. Therefore the only
contributions that need to be considered when assessing the uncertainty due to distance setting
are the precision with which the bench scale can be read (since this limits the ability to set a
reproducible distance value) and the uncertainty in defining the mean plane of the filament.
The combined effect of these two contributions is estimated as 0.5 mm, which corresponds to
an uncertainty in luminous intensity of 0.05 % at the measurement distance of 2.40 m.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 134 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
The accuracy of the lamp current setting is determined by the uncertainty associated with the
calibration of the standard resistor (including an allowance for possible drift in the resistance
since the time of calibration) and the uncertainty associated with the calibration of the
voltmeter (again including an allowance for possible drift since the time of calibration). These
were estimated to give a combined uncertainty of 0.02 % in current, which corresponds to an
uncertainty of 0.160 % in luminous intensity (using an 8:1 relationship between intensity and
current).
Since the measurement procedure used at NPL involves a direct comparison between lamps of
similar types, the majority of the factors that influence the accuracy of the measurement of the
photometer photocurrent (such as amplifier gain and digital voltmeter accuracy) have
negligible impact on the final luminous intensity value. A small contribution (0.01 %) is
included in the uncertainty budget to allow for any residual uncertainty e.g. due to the effect
of ambient temperature fluctuations.
As indicated in Section 3, the reference and test lamps have similar correlated colour
temperatures and the photometer has a good match to the V(λ) function (f1’ better than 3.5 %).
No spectral mismatch correction was therefore necessary, but a small contribution (0.01 %)
was allowed for spectral mismatch in the uncertainty budget.
Stray light was minimised through the use of stray light screens between the lamp and the
photometer. A small contribution of 0.02 % was included in the uncertainty budget to allow
for any residual stray light.
The lamps were operated from a stabilised power supply, able to control current to 0.002 %.
The corresponding uncertainty in lamp luminous intensity was estimated as 0.016 % (using an
8:1 relationship between intensity and current).
The repeatability of the photometer calibration factor was determined by statistical analysis of
the results using a number of the NPL reference standard lamps. The standard uncertainty was
included as a Type A contribution in the uncertainty budget.
The repeatability of the measurements on the test lamps was estimated based on statistical
analysis of the results of previous measurements on similar lamps, in which the lamp was
realigned a number of times at various extremes of what would be regarded as an ‘acceptable’
alignment. In practice, the measurement repeatability achieved was better than this, but the
worst case estimate (0.05 %) was used in the uncertainty budget.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 135 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
6 REFERENCES
[1] Goodman TM and Key PJ. The NPL radiometric realisation of the candela. Metrologia
1988; 25: 20-40.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 136 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
1. The lamps used for the comparison were calibrated directly against NPL’s primary reference
standard luminous intensity lamps, which are of exactly the same type as the comparison
lamps. Any reflections from the inside edges of baffles or shutters are therefore common to
both the reference and comparison lamps and the effects cancel; no correction is necessary.
Extensive investigations into stray light effects (including light scattered, reflected or
diffracted by apertures and baffles) were carried out during the realisation of the luminous
intensity scale and assessed to be less than 0.01 % - this is included in the uncertainty
budget for NPL’s realisation of the candela.
2. The alignment of the NPL photometer was not changed between the calibration using the
reference lamps and the measurements of the comparison lamps; therefore it is not
necessary to include an uncertainty component for misalignment of the photometer
aperture.
3. NPL did not follow the model given in CIE 198:2011 since this is not how we usually structure
our uncertainty budget. We did, however, provide a detailed description of each of the
uncertainty contributions included in our uncertainty budget, which we believe gives the
information necessary to judge the legitimacy of each of these. For completeness, our
measurement equation is given below (this has also been added to our measurement
report):
𝐼𝐼v,𝑡𝑡 = 𝐶𝐶cal 𝑉𝑉𝑡𝑡 (1 + 𝐶𝐶d,𝑡𝑡 )(1 + 𝐶𝐶J,𝑡𝑡 )(1 + 𝐶𝐶p,𝑡𝑡 )𝐹𝐹SM,𝑡𝑡 (1 − 𝐶𝐶stray,𝑡𝑡 )(1 + 𝐶𝐶align,𝑡𝑡 ) (1)
where
(𝐼𝐼v,𝑟𝑟 +𝐶𝐶age,𝑟𝑟 )
𝐶𝐶cal = 𝑉𝑉𝑟𝑟
(2)
and
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 137 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
current for the reference lamp to the specified value is included in the uncertainty
budget for the luminous intensity of the reference lamp)
𝐶𝐶p,𝑡𝑡 is the error in luminous intensity for test lamp t due to differences in amplifier gain
and DVM sensitivity between measurement of the photocurrent produced by the
reference lamp and that produced by the test lamp
𝐹𝐹SM,𝑡𝑡 is the spectral mismatch correction factor for test lamp t
𝐶𝐶stray,𝑡𝑡 is the error in luminous intensity for test lamp t due to differences in stray light
between the reference and test lamps
𝐶𝐶align,𝑡𝑡 is the error in luminous intensity for test lamp t due to misalignment of the lamp
(the uncertainty due to misalignment of the reference lamp is included in the
uncertainty budget for the luminous intensity of the reference lamp)
Note all of the C terms listed above have an expected value of zero and an associated uncertainty
that has been estimated as described in our measurement report.
1. Yes, it is impossible to isolate the effect of ‘stabiliser current control’ from ‘test lamp
repeatability’ so there is potentially a small element of double counting in the random
effects. However since the test lamp repeatability component is intended primarily to allow
for lamp alignment variations and is treated as a worst case estimate, we have chosen to
ignore this small element of double counting. The effect on the final uncertainty is
insignificant.
2. We do not know the actual change in luminous intensity due to ageing for each individual
reference lamp used. Each reference lamp has been used for a different length of time since
the original calibration and will also age at a slightly different (unknown) rate. We therefore
do not correct for ageing effects. The uncertainty estimate is a conservative allowance,
which is based on measurements on other lamps of the same type operated at the same
correlated colour temperature and under the same conditions coupled with knowledge of
the maximum length of time for which the reference lamps have been used since the
original calibration.
3. We apologise for these typing mistakes, which were due to importing the table from an
Excel file. We have provided a corrected version of the report to the pilot laboratory.
4. The uncertainty due to lamp alignment is included under ‘Test lamp repeatability’ as
described in section 5.10 of our report.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 138 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Dear Arnold,
Many thanks for sending the NPL relative data for CCPR-K3.2014 for review. I have the following
comments:
1. We had noted from our measurements that the luminous intensity values for our two
Wotan lamps, 877 and 890, were significantly different for our round 1 and round 2
measurements; the differences were significantly larger than would be expected based on
the random uncertainty associated with measurements of these lamps. The Polaron lamps
showed much better stability during the course of the comparison, with values from the two
rounds agreeing at the levels we would expect (i.e. within the random uncertainty).
2. This suggested that the luminous intensity of both Wotan lamps had changed as a result of
transportation. Other possible causes of a change in output, such as ageing, appeared
unlikely because of the very short burn time during the course of the comparison
measurements. Furthermore, we considered it likely that the change in output would have
occurred either during transportation to NRC or during return to NPL; changes during both
transportations could occur, but are less probable.
3. We obviously could not tell from our measurements whether the values had changed during
the first transportation, from NPL to NRC, or during the second, from NRC to NPL. Based on
the evidence of the relative data, it appears that for both lamps the change is most likely to
have occurred after the NRC measurements, i.e. during transport back to NPL. This is
demonstrated by the small standard deviation in the candela/volt ratios using the Round 1
luminous intensity values (0.06 %, which is consistent with the random uncertainties of
0.08 %) compared with the much larger standard deviation using the round 2 values (0.14 %,
which is significantly higher than the random uncertainties).
4. Based on this review, I would request that the analysis for the NPL Wotan lamps should use
only the round 1 luminous intensity values. Both rounds of measurements should be used
for the Polaron lamps.
Please let me know whether this is acceptable. I am happy to give further details if necessary.
Best wishes
Teresa
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 139 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
PTB Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 140 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
Appendix A.3
Description of the measurement facility at
PTB
The measurements are carried out at the photometer bench system. The photometer bench system is
composed of three different photometer benches aligned in a row. They can be used singly or
together, so that measuring distances up to 40 meter become possible. The distance readings are
from absolute electronic linear encoders with resolution below 0.01 mm and linear well within that
range. The calibration of all geometric relations is performed by a laser-tracker. The latter is traced
back to the national PTB length standard with an expanded uncertainty of 0.1 mm for distances up
to 8.5 m. The following Fig. 1, 2, 3, 4 and 5 show the main components of the bench and illustrate
their interactions.
Fig. 1: Main components of the photometer bench (schematic) including light trap, baffles and the
aperture plus shutter for field-of-view limitation and dark measurements, respectively.
6 photometer
spectrometer
600 lx
P1…P6 = photometer
photometer carousel
P6 SPM P1 SPM = spectrometer
V1 V2
L = alignment laser
P5 P2
L
RP = reference point
P4 P3
M
M = micrometer
RP
Indication of distance V1, V2 = shutter Carriage motor
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 141 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
video camera
ph
ot
om
et
Telescopes + h er
nc be
e nc
video cameras sb h
ros
c
Fig. 3: Tools for the camera aided alignment are two video cameras (left) to the left and right side of
the lamp and, behind the rolling gate, the third video camera (right) mounted temporarily within the
photometer bench for front view.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 142 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
At the PTB the unit of luminous intensity is the realized and maintained by a network of lamps and
photometers [1]. All objects are characterized over long periods of time and well known for their
relevant properties such as coefficients for ageing, geometrical misalignment, electric
misadjustment, and ambient influences (temperature, humidity, air stream).
The 6 lamp transfer standards participating in this key comparison are organized as a fixed group
since the last CCPR key comparison [2] which acts as a PTB- internal duplication for the lamp
transfer standard group taking part in the last key comparison. Since then, they were operated for
only 4 hours and were calibrated according to the value represented by PTBs network of lamps (see
“Traceability chain and date of last realization”), before transport to the pilot laboratory. Hence,
their values represent the valid national luminous intensity unit of PTB.
Two photometer heads LMT with thermostatic stabilization at 35°C are used and permanently
heated.
a) Type P30, aperture with diffuser, the reference plane is outside of the opal glass of the entrance
window (diameter 30 mm)
b) Type P10, aperture without diffuser, the reference plane is outside of the glass of the entrance
window (diameter 10 mm)
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 143 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
Geometrical conditions:
The lamps OSRAM WI41/G are aligned (see Fig. 4) without glowing:
- lamp’s optical axis is central and rectangular to the filament plane
- lamp’s optical axis is parallel to the bench’s horizontal axis
- plane, containing lamp’s optical axis and lamp axis (cap down) is vertical
- distance is measured from the centre of the filament
- only the light passing through the opening (see Fig. 4) in the mask is measured
For a measurement of the luminous intensity values the assigned distances vary depending on the
effective location of the beginning of the light path within the filament of the lamps. Therefore, at
PTB all luminous intensity measurements were carried out in a (large) distance of 5.5 m between
the plane associated with the filament and the limiting aperture of the photometer. In most cases
additional readings at reduced distances were taken to find out the sensitivity coefficient for a
translation in the direction of the bench’s optical axis. These coefficients are used for the evaluation
of uncertainty as well as for a correction between the different measurements conditions, if needed.
However, it turns out - as expected - that the influence is negligible under the conditions realized at
PTB (distances 3 m to 7 m and apertures 10 mm to 30 mm in diameter; which corresponds to solid
angels between 1.6·10-6 sr and 79·10-6 sr).
The lamps are operated with constant DC-currents and the values are selected for a distribution
temperature of about 2800 K. Every individual lamp is operated for a period of 15 minutes at
nominal current before the measurement starts to warm up and to allow for the stabilization of its
luminous output.
Stray-light reduction
The room for the measurements is divided by the rolling gate in two parts, one room for the lamp
and a second room for measurement with the photometer bench (see Fig.3) ensuring large distances
to the walls, ceiling and the floor. All sources of light except the lamp standard are switched off
during measurement.
A light trap more than 1 m behind the lamp reduces the back reflected stray light. Baffles with
various openings are placed on the photometer bench such that light illuminates neither the rails nor
the room for measurements. The Fig. 6 gives an impression how baffles look like and are placed.
The land of a baffle if illuminated originates a relative stray-light of 10-5 which is corrected
numerically. The box with the photometer carousel screens the photometer heads from any side-
direction and a baffle with adjustable opening limits the field of view for the photometers. The
illuminated entrance window of the photometer head reflects back and would illuminate the lamp.
This is avoided by a minimal tilting just to direct the spot onto the baffles in between.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 144 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
The luminous intensity unit at PTB is realized annually. The last realization was carried out in
December 2013. The traceability chain at PTB (see Fig. 7) starts with the cryogenic radiometer to
establish the unit of spectral radiant power, which is used to determine the spectral power
responsivity of trap detectors. Using a uniform source based on tunable lasers and trap detectors
with precision apertures, the responsivity with respect to optical power is transferred into a spectral
irradiance resonsivity, and, in a second step, using V(λ)-corrected photometers, into the photometric
responsivity. Parallel to this step, filtered detectors are calibrated to determine the temperature of a
high temperature Black-Body radiator, used to provide the relative spectral distribution of transfer
standard lamps at Illuminant A. Using the photometric bench system at PTB, the photometric
responsivity of the calibrated photometers and the relative spectral distribution of the transfer
standard lamps are combined to verify and establish the SI base unit Candela at PTB.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 145 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
Parallel to this realization of the Candela, the unit Candela is also maintained at PTB since
introduction of the new definition of the Candela in 1979 using a set of 17 Toshiba lamps operating
at a distribution temperature of 2042 K, a set of 5 Toshiba lamps at 2353 K and 6 OSRAM WI41/G
at 2600 K, and additional 12 OSRAM WI41/G lamps separated in two groups working at a
distribution temperature of 2800 K, close to CIE-Illuminant A (see Fig. 8).
Fig. 8: Maintenance of the unit Candela by the luminous intensity lamps operated at various
distribution temperatures.
Due to the different but very low aging rates of the various groups of lamps, which are only
operated for traceability check once a year, the uncertainty of the preserved magnitude of the unit
defined by this network of lamp groups is well below 1·10-3 (k = 1). The preserved quantity,
together with the realized quantity via the detector based traceability chain finally establish official
magnitude of the unit Candela of PTB with an uncertainty of 1.02·10-3 (k = 1). The magnitude of
the unit is then disseminated by PTB by transfer standard lamps and transfer standard detectors with
an uncertainty of typically 3.6·10-3 (k = 2).
References:
[1] Erb, W., Sauter, G., PTB network for realization and maintenance of the candela, Metrologia,
1997, 34, 115-124
[2] Georg Sauter, Detlef Lindner, Matthias Lindemann, CCPR Key Comparisons K3a of Luminous
Intensity and K4 of Luminous Flux with Lamps as Transfer Standards, PTB Bericht,
PTB-Opt-62, 1999.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 146 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 2014-March-15
Appendix A.5
Uncertainty Budget
(Example)
The following example of the complete measurement budget is based on the document
“CIE 198-SP1:2011”. The explanations for all entries are given in that document and the
values are taken out of the quality management system of the photometry laboratory except
those which are found from the measurement of the individual lamps. The measurement
uncertainty in line 35 is stated as standard measurement uncertainty. It has been
determined in accordance with the “Evaluation of measurement data – Guide to the
Expression of Uncertainty in Measurement; JCGM 100:2008”.
CIE E 13 Calibration of a luminous intensity standard (source based) manual entries are blue colored
mP m P ⋅m TS 1 − m IS1 − m P ⋅m TSR + m IS R
y PS1 TdC 1 c J ⋅ U JS1 cJ ⋅ U JSR corSR
I CS1 = I C SR
y P SR Td CR J C1 ⋅ RJ J CR ⋅ RJ corS1
Note:
The type A/B evaluation of uncertainties valid for the different quantities is stated above.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 147 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
PTB-Answers
General Comment 1:
Straylight created by baffles in the light path depends strongly on their shapes and the
construction of the edges and it is not corrected by background subtraction. This yields
similarly for straylight back reflected from the light trap behind the lamp. The effect of
this type of straylight is mostly compensated if luminous intensity lamps are used as
reference standards for the transfer standards within the CCPR comparison as
performed by the PTB. Provided this type of straylight contributes significantly to the
combined uncertainty then it has to be mentioned in the model of evaluation and in the
uncertainty budget. It should be mentioned that the baffles used at the PTB create a
relative straylight < 5 ⋅ 10 −5 . In case the photometer is reference for the calibration of
the luminous intensity standard lamps the uncertainty of the aperture has to be taken
into account and only then the given reference [Metrologia 37, 621 (2000)] is helpful.
General Comment 2:
Usually the photometer´s aperture plane is aligned by help of a mirror and a back
reflected laser beam and any deviation from the perpendicular direction has to be
weighted by the cosine. The effect of this misalignment is mostly compensated if the
mounting of the photometer was unchanged between its calibration as reference and
the transfer to the transfer standards within the CCPR comparison. Provided this
misalignment contributes significantly to the combined uncertainty then it has to be
mentioned in the model of evaluation and in the uncertainty budget.
General Comment 3:
In the Technical Protocol for this CCPR comparison chapter 6.1.1 the GUM is
explicitly claimed as reference for any statement of measurement uncertainty.
Additionally, the chapter 6.1.2 refers to the document CIE 198 as example for
modeling combination and presentation. The protocol itself gives in Appendix A.5 an
example for an abbreviated presentation. Thus, the model of evaluation is an essential
part in the documentation and has to be stated individually by each participant as well
as the complete uncertainty budget from CIE 198 as an intermediate step for the
summarized presentation recommended in Appendix A.5 to simplify the comparison
of individual contributions.
1
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 148 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Questions to PTB:
a) According to the GUM all entries in Appendix A.5 are labeled in column 6 with
“A” for “statistical” or “B” for any “other determination”. These types of entries
are combined and listed separately for each lamp. The list was send to the pilot for
an additional explanation and mean values u(A) = 0.12% and u(B) = 0.13% are
indicated. Thus, the combined standard uncertainty for the transfer by only one
lamp is u ( I ) = u (A) 2 + u (B) 2 = 0.18% .
b) At the bottom of the table Appendix A.5 two values labeled udev = 0.12% for
random (“dev” for devise) and uinst = 0.14% for systematic contributions (“inst” for
instrumentation) are included. These numbers, their meaning and the evaluation
are explained in all details in the publication CIE 198-SP1.2:2011 (see
chapter/example 2.13). The combination of these numbers to determine the
uncertainty of the whole batch for the transferred value of intensity is explained in
great detail in CIE 198-SP1.1:2011 example 1.11.
It turns out that the instrumentation for the two rounds at PTB was stable and the
properties of the PTB-transfer-standards (WI41/G) are uniform. A separation in types
A and B or “random” and “systematic” gives no real difference. So, the uncertainty
u(PTB) associated with the luminous intensity value transferred by the batch with a
number of 6 PTB-transfer standards will be determined by the pilot laboratory from
2
u dev
u ( I ) PTB = u inst
2
+ = 0.15%
6
Braunschweig, 2016-April-28
2
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 149 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 759
- Final result -
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.1225 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 236.21 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.1225 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 236.22 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 150 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 791
- Final result -
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.5643 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 247.55 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.5649 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 247.53 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 151 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 793
- Final result -
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.3866 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 245.97 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.65000 A
Lamp voltage
value UL 29.3867 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 246.00 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 152 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 848
- Final result -
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.5727 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2810 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 228.53 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.5712 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2810 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 228.54 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 153 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 851
- Final result -
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.9316 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2815 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 233.49 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.9313 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2815 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 233.54 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 154 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014 March 2014 / June 2015
Appendix A.6
Initial & Return measurement at PTB
Lamp-No.: 858
- Final result -
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.5610 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 225.12 cd
relative standard uncertainty urel(I) 0.18 %
Lamp current
nominal value JL 5.70000 A
Lamp voltage
value UL 28.5618 V
relative standard uncertainty urel(UL) 0.011 %
Distribution temperature
value Td 2800 K
absolute standard uncertainty u(Td) 20 K
Luminous intensity
value I 225.01 cd
relative standard uncertainty urel(I) 0.18 %
Note: The stated results of the return measurements at PTB include the aging corrections for the
total operating time at the Pilot-Laboratory and at the PTB using the following averaged relative
correction coefficients with associated standard uncertainties:
Lamp voltage +6.0E-5/h with 3.0E-5/h
Luminous intensity +2.5E-4/h with 1.0E-4/h
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 155 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
VNIIOFI Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 156 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
The items listed on this form should be used as a guide. It is anticipated that many of the questions
will require more information than the space allocated on this page. Please expand your reply
document as necessary.
Fig.1. Diagram of VNIIOFI facility used for measuring luminous intensity of comparison lamps
within the CCPR-K3.2014 comparison. A – side view; B – top view. 1 – lamp to be measured; 2 –
photometer; 3 – telescope; 4 – alignment laser; 5 – mirror; 6 – lamp socket; lamp alignment mount;
photometer alignment mount; translation stage; 10 - limiting baffle (aperture is 40x45 mm); 11 –
baffle with shutter (aperture is 55x80 mm); 12 – baffle with aperture diameter of 100 mm; 13 –
baffle with aperture diameter of 80 mm; 14 – baffle with aperture diameter of 50 mm; 16 – ruler.
- positions of lamp, detector, bench, shielding, baffles (number, distances and sizes)
A lamp, photometers and baffles stand on a rail inside a light-tight box. On a parallel rail
there is a telescope for aligning a lamp and measuring distance. All side walls of the box are covered
with black velvet cloth. The ceiling and baffles are painted by diffuse black paint.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 157 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Three photometers were used for the comparison. All three were located on a translation
stage perpendicular to the rail. The photometers were pre-aligned before the measurements and then
replaced each other without additional alignment during the measurement.
A laser beam, reflected by a mirror behind the lamp, is used for aligning the lamp and
photometers. The distance from the lamp to the mirror is 550 mm. During measurements the laser
and mirror are shielded by black velvet cloth.
The lamp area is separated from the other box volume by a black velvet curtain. In the plain
of the curtain in front of the lamp filament there is a limiting baffle. The distance between the baffle
and the lamp bulb is 50 mm. The aperture of the baffles is rectangular 40x45 mm (width x height).
There are four additional baffles between the lamp curtain and the photometers. The widths
and heights of the baffles are 400 mm and 510 mm, respectively. Aperture of the baffle nearest to the
curtain is rectangular of 55x80 mm. This baffle is equipped with a shutter. Apertures of other baffles
are round with diameters of (if counted from the lamp to the photometer): 100 mm, 80 mm and 50
mm. The distances: from the curtain to the first baffle is 320 mm; from the first to the second is also
320 mm; from the second to the third is 500 mm; from the photometer to the fourth is 50 mm.
Make and type of the photometer (or equivalent). LMP Photometerhead of the P150T type with
LMT photocurrent meter of the I1000 type
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 158 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
At room temperature
- alignment jig? If so, how is it used? No jig was used.
- size and position of limiting aperture
40x45 mm (width x height) at the distance of approximately 2050 mm from the
photometer (about 50 mm from the lamp bulb). This aperture is smaller that the lamp window.
“+” mark
THREAD
of the base
- length of warm-up time for each lamp before measurements are taken
15 minutes
- measured CCT (or Distribution Temperature or Colour Temperature, see Section 3.5).
See the table above
- stray-light reduction
No correction for stray-light was done
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 159 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Establishment or traceability route of primary scale including date of last realisation and uncertainty
budget.
Primary scale was realized using a high-temperature blackbody. Luminous intensity of the
blackbody calculated as
I BB = ε ⋅ A ⋅ K cd ⋅ ∫ Lλ,BB (λ , TBB ) ⋅ V (λ )dλ (1)
where
K cd is the luminous efficacy, equals to 683 cd·sr/W
V (λ ) is the photopic luminous efficiency function,
Lλ,BB (λ , TBB ) is the ideal blackbody spectral radiance,
TBB is the blackbody temperature,
ε is the blackbody emissivity and
A is an area of the blackbody aperture.
The temperature of the blackbody was approximately 2856 K. The exact temperature was measured
by a radiation thermometer, which was calibrated against three high-temperature fixed points: Co-C
(1597 K), Re-C (2748 K) and WC-C (3021 K). The fixed points were earlier measured by means of
comparison with the copper fixed point (1357.77 K) in according with the ITS-90. The standard
uncertainty of blackbody temperature measurement was 0.5 K.
The emissivity of the blackbody was estimated using the Monte-Carlo based software STEEP3. as
0.9995 with standard uncertainty of 0.0002.
A water-cooled bronze aperture was used with approximate diameter of 8 mm. The exact value of an
average diameter measure with standard uncertainty of 1.5 µm.
Responsivities of the photometers to the Type A source were measured against the blackbody and
equals:
iphot
s v,phot = ⋅M
g ⋅ (I BB l 2 )
(2)
where
M =
∫ Lλ,BB (λ , TA ) ⋅ sreλ,phot (λ )dλ ⋅ ∫ Lλ,BB (λ , TBB ) ⋅ V (λ )dλ (3)
∫ Lλ,BB (λ , TA ) ⋅ V (λ )dλ ∫ Lλ,BB (λ , TBB ) ⋅ sreλ,phot (λ )dλ
M – Spectral mismatch correction factor;
s rel,phot (l ) – Relative spectral responsivity of the photometer;
TA = 2856 K;
l – Distance between the photometer and the blackbody aperture was about 720 mm;
g – Geometry correction depends on sizes of apertures and the distance.
Because the temperature of the blackbody agreed with TA = 2856 K within 2K only, the difference of
M from the unit and its uncertainty were negligible (less than 0.005%).
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 160 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Source of uncertainty
standard uncertainty, %
Blackbody temperature (0.5 K) 0.16
Blackbody uniformity 0.03
Blackbody stability (0.03 K) 0.01
Emissivity 0.02
Aperture size (1.5 µm) 0.04
Distance (0.1 mm) 0.03
Stray light 0.04
Repeatability of measurement
0.08
(with independent alignment)
Combined Standard
0.19
Uncertainty
The last realization and calibration of the photometers was done in December 2013 (one month
before the first round measurements of the K3 lamps).
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 161 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Lamp number: N 01
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 162 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Lamp number: N 02
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 163 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Lamp number: N 03
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 164 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Lamp number: N 04
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 165 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 166 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 167 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Random effects**:
Electrical parameters:
- power supply fluctuations, upsf B 0.02
Photometer noise (25 readings), unoise A 0.001
Discrepancy between three photometers (random), upd A 0.05
Independent measurement reproducibility *** , urep A 0.05
**** Uncertainty associated with reproducibility (urep) partly includes uncertainties associated with
other random effects. Therefore, the Total random uncertainty is calculated as
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 168 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Measurement parameters given in this table are suggested. Please modify and itemize according to
your particular situation. See Section 6.2 for explanation of the various items.
Note that if lamps are used as the laboratory working standards, a group of uncertainties would need
to be included in the above table to account for their behaviour.
The RMS total refers to the usual square root of the sum of the squares of all the individual
uncertainty terms.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 169 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
NMI: VNIIOFI
Lamp# Measurement# Date Lamp Electrical Lamp CCT Luminous Intensity (cd)
Round# Set# Meas#PerSet Current(A) Voltage(V) K I(cd) Standard Uncertainty (%)
Random Systematic final
Luminous Intensity
Final Report
Appendix A
NIST Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 171 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report UNITEO STATES OEPARTMENT OF COMMERCE
National Institute of Standards and Technology
Gaithersbur g, Maryland 20888-
REPORT OF CALIBRATION
Luminous Intensity and Co lor Temperature Standard Lamps
Submitted to :
1. Calibration Item
Six incandescent lamps model Wi41 /G manufactured by Osram Inc. were calibrated for correlated
color temperature and for luminous intensity. The lamp designations NIST20 100, NIST20 101 ,
NIST20 102, NIST20 103, NIST20 l 04, and NIST20 105 are marked on the lamp base.
z
Vert1cal ax1s
Coordinate
System
X
Opt1ca1 ax1s
Osram Wi 41 /G
filament
Figure 1 - Lamp filament and coordinate system, shown with the Osram Wi41 /G filament
'L y
w
u
Center Fi lament Support #1
Figure 2- Osram Wi41 /G fi lament defined point: Center
(2) Focus the side telescope on the right filament post (viewing the lamp from the photometer).
Rotate the lamp about Y axis and adjust the lamp position along X axis so that the right filament
post is aligned exactly on the vertical fiducial line (the red line in Figure 3) in the side telescope.
(3) Focus the side telescope on the left filament post. Note the image of the left filan1ent is not so
clear but you can still tell when the telescope is focused on the left post by adj usting the focus back
and forward near the left filament post. Rotate the lan1p about Z axis and adjust the lamp position
along X axis so that the left post is also on the vertical fiducial line in the side telescope.
(4) Repeat steps 2 and 3 unti l both right and left fi lan1ent posts are aligned onto the vertical fiducial
line in the side telescope.
(5) Double check if all aligmnents are good by repeating steps 1, 2, 3 and 4.
(6) Focus the side telescope on the right side lamp filament (viewing the lamp from the
photometer). Adjust the lamp position along X axis so that the distance reference point (the
intersection point between the right side filament and the right lamp post) is aligned onto the
vertical fiducial line (the distance origin) as shown in Figure 4.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 173 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
REPORT OF CALIBRATION Manufacturer: Osran1
Luminous Intensity and Color Temperature Calibration Model: Wi41/G
National Research Council of Canada Designation: NIST20 100, NIST20 101,
NIST20102, NIST20103, NIST20104, and NIST201 05
Figure 3 - Alignment of the right filament post onto the vertical fiducial line
Distance
reference point
Figure 4 - Alignment of the distance reference point of the lamp onto the vertical fiducial line
The room temperature was 23 °C and relative humidity was approximately 47% at the times of
calibration. The equipment and the details of the calibration procedures of the luminous intensity
and color temperature measurements are described in Section 3 and Section 7 of reference [ 1].
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 174 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
REPORT OF CALIBRATION Manufacturer: Osram
Luminous Intensity and Color Temperature Calibration Model: Wi41/G
National Research Council of Canada Designation: NIST20100, NIST20101,
NIST20102, NIST20103, NIST20104, and NIST20105
expanded uncertainty (k=2) of the color temperature value is 8 K as shown in Table 18 (page 60)
of reference [1]. The NIST policy on uncertainty statements is described in reference [2].
4. General Information
The lamp should be carefully aligned in accordance with the procedures described above. The
lamp should be operated on DC power at the reported current and at the prescribed polarity.
Photometric measurements should be made at least 10 minutes after turning on. The uncertainty
value is valid only for distances larger than 2 m.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 175 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
REPORT OF CALIBRATION Manufacturer: Osram
Luminous Intensity and Color Temperature Cal ibration Model: Wi41 /G
National Research Council of Canada Designation: NIST20 100, N IST20 10 I ,
N IST20 102, N IST20103, NIST20104, and N IST20 105
The customer should take the uncertainty associated with the aging of the lamp and the calibration
cycle into account.
The Calibration Report shall not be reproduced except in full , w ithout the written approval of
NIST.
0(1>f/~
t~~nf ~
Sensor Science Divis ion
Maria Nadal
Sensor Science Division
Physical Measurement Laboratory Physical Measurement Laboratory
(301) 975-2332 (30 1) 975-4632
Approved by:
t ~~
C. Cameron Miller
For the Director,
National Institute of Standards and Technology
(301) 975-4713
References:
[1] Y. Ohno, N IST Special Publication 250-37 "Photometric Calibration" (1997)
[2] B. N. Taylor and C. E. Kuyatt, "Guidelines for Evaluating and Expressing the
Uncertainty ofNIST Measurement Results," N IST Technical Note 1297 ( 1994).
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 176 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
REPORT OF CALIBRATION Manufacturer: Osram
Luminous Intensity and Color Temperature Calibration Model: Wi41/G
National Research Council of Canada Designation: NIST20100, NIST20101,
NIST20102, NIST20103, NIST20104, and NIST20105
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 177 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Final Report
Appendix A
NRC Report
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 178 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
NRC Report
Six Osram Wi41/G lamps were used at the pilot (NRC) laboratory to represent the NRC luminous
intensity scale for the comparison. These lamps are the six primary Osram Wi41/G lamps that were
calibrated from room temperature absolute radiometers as described in our paper:
L.P. Boivin, A.A. Gaertner, and D.S. Gignac Realization of the New Candela (1979) at NRC,
Metrologia 24, 139-152 (1987).
As described in this paper, the NRC Candela was most recently realised in 1986, using room
temperature electrical-substitution absolute radiometers to calibrate secondary radiometers. The
absolute spectral responsivities of the secondary radiometers were measured at laser wavelengths
476.2 nm, 530.9 nm, 568.2 nm, 647.1 nm, and 676.4 nm. Auxiliary measurements using a
monochromator apparatus were used to obtain calibration points below 476.2 nm, to 380 nm, and
above 676.4 nm, to 800 nm. Interpolation techniques were used to obtain complete calibration data
from 380 nm to 800 nm at 5 nm intervals. These secondary radiometers, which incorporated
diffusers, were then used with V(λ)-correcting filters to calibrate the lamps. These photometers were
not thermostated.
The equation for the luminous intensity of a lamp (Iv), as measured by the output voltage (Vout) of
the photometer, is given by (see equation 1 of our above-mentioned paper):
2
683 ∙ V𝑜𝑜𝑜𝑜𝑜𝑜 ∙ �𝑑𝑑 − 𝑡𝑡�3� ∫ 𝑉𝑉(𝜆𝜆) ∙ 𝐼𝐼𝑒𝑒 (𝜆𝜆) ∙ 𝑑𝑑𝑑𝑑
𝐼𝐼𝑉𝑉 = ∙
𝑅𝑅𝑓𝑓 ∙ 𝐴𝐴 ∫ 𝑆𝑆(𝜆𝜆) ∙ 𝑇𝑇(𝜆𝜆) ∙ 𝐼𝐼𝑒𝑒 (𝜆𝜆) ∙ 𝑑𝑑𝑑𝑑
where
Rf = feedback resistance of the detector amplifier,
A = area of the radiometer aperture,
D = distance between the lamp filament and the radiometer aperture,
t = thickness of the V(λ)-correcting filter,
Ie(λ) = relative spectral distribution of the lamp,
S(λ) = absolute spectral responsivity of the secondary radiometer, and
T(λ) = spectral transmittance of the filter.
The lamps are OSRAM type Wi41/G, operating at a colour temperature of 2800 K. The colour
temperature of these luminous intensity lamps was set to 2800K when they were first calibrated in
1987.
These lamps have been used very little since that time, so the luminous intensity values assigned to
these lamps is the same as at their calibration in 1986. These six lamps are also used as the
primary standards to maintain the candela at NRC. An estimate of the aging of the lamps due to
use since they were calibrated is included in the uncertainty budget as indicated below.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 179 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
The statement of the uncertainty in the calibration of these lamps is given in Table 2 of our above-
mentioned paper, and reproduced below.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 180 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
*L.P.Boivin, A.A.Gaertner, and D.S.Gignac, Realization of the New Candela (1979) at NRC
-Metrologia 24, 139-152 (1987)
1. The term ‘Lamp Maintenance/aging’ is added to the original uncertainties as an estimate of the
uncertainty in the luminous intensity values of the lamps used for the comparison since the time
that they were calibrated. This is predominantly an estimate of the aging of the lamps due to use
since they were calibrated.
2. The lamp alignment component is included in the lamp reproducibility term.
CCPR-K3.2014: Luminous Intensity Final Report, Appendix A Appendix A Page 181 of 181
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report
Luminous Intensity
Draft B Report
Appendix B
- attachments:
- VNIIOFI, NPL, NMIJ, NMISA
VNIIOFI: Following the first General Comment of the pre-Draft A Process 2, we recently
tried to estimate an uncertainty associated with scattering in inner edges of baffles, but
suddenly realized that our measurements (both rounds) were mistakenly done without a
thin edge aperture on a shutter. So, the thickness of the actual edge was too thick and
gave quite strong reflectance. We have measured this effect and found that the
luminous intensity values have to be reduced. Please find the corrected files attached.
(See attached pages VNIIOFI response for the revised Appendix A.5 uncertainty
budget.) Corrected values are marked red. All luminous intensity values are reduced.
The uncertainty component associated with stray light is increased from 0.02% to
0.1%; the total uncertainty is increased up to 0.25% and 0.26% for the 1st and 2nd
rounds, respectively.
NIST: There are two irises (used as baffles) and one electric shutter between the lamp and the
photometer. The first and second irises are located at approximately 0.6 m and 1.85 m,
respectively from the lamp. The thickness of the iris blades is 0.2 mm. The photometer
is at 3.5 m away from the lamp. The measurement uncertainty resulting from the
reflected light is analyzed using the optical ray tracing technique. The shutter does not
cause any reflected light because its opening is larger enough so that it is completely
hidden behind the second iris (i.e., it is in the dark). The first iris does not contribute to
the measurement error because the angle of its reflected light is large enough so that
the reflected light cannot reach the photometer. The uncertainty resulting from the
reflected light from the second iris is estimated to be less than 0.01 % and therefore no
correction is applied. Instead it is rolled into our 0.05% stray light uncertainty
component, which also includes the scatter light from the edge of iris, the inter-
reflection between the photometer, photometer mount, irises, wall of the photometry
bench, and light trap, etc.
2. Comment for all the laboratories that are using an aperture:
If the aperture plane is not perpendicular to the optical axis, the effective aperture area will be
smaller than that at the normal position. This uncertainty component is not in the uncertainty list.
(NMIA, LNE-CNAM, VNIIOFI, NRC).
Answers:
PTB: Usually the photometer´s aperture plane is aligned by help of a mirror and a back
reflected laser beam and any deviation from the perpendicular direction has to be
weighted by the cosine. The effect of this misalignment is mostly compensated if the
mounting of the photometer was unchanged between its calibration as reference and the
transfer to the transfer standards within the CCPR comparison. Provided this
misalignment contributes significantly to the combined uncertainty then it has to be
mentioned in the model of evaluation and in the uncertainty budget.
NPL: The alignment of the NPL photometer was not changed between the calibration using
the reference lamps and the measurements of the comparison lamps; therefore it is not
necessary to include an uncertainty component for misalignment of the photometer
aperture.
NMIA:This uncertainty component (tilt) is included in the derivation of the uncertainty in the
aperture area, although those details were not included in our uncertainty budget.
However, it is true that a component due to this effect when using the aperture has not
been included. Using a retro-reflected laser, we have estimated the possible tilt in the
aperture when it is mounted in our usual way. The tilt is estimated to be approximately
0.08 degrees, so the contribution to the uncertainty is negligible. For consistency with
our comments below, we could submit a revised budget with this term included (and
set to zero) if participants considered it would add value.
NRC: This uncertainty is included in the ‘Measurement repeatability’ component.
NIST: Using a mirror and retro laser along with mechanical alignment the aperture alignment
off axis is very small. Less than 0.002% as captured in Table 5 of “Yuqin Zong, Maria
E. Nadal, Benjamin K. Tsai, and C. Cameron Miller, “Photometric Calibrations,” NIST
Special Publication 250-95. (2018). https://doi.org/10.6028/NIST.SP.250-95 “
3. It turns out that the variety of measurement budgets, and the components mentioned in there, is
quite large. However, when I received the document to review the uncertainty budgets, I was
really surprised that obviously only very few (3 of 11) participants have sent their model for
evaluation which they used to establish their distinct associated uncertainty budget.
According to GUM, the uncertainty budget of a measurement must be based on a measurement
model which clearly connects input and output parameters by means of a physical equation to
show the interdependencies and the sensitivities of the various uncertainty contributions.
It would have been nice, if every participant would have used the uncertainty determination
according to the Technical Report CIE 198:2011, where a clear GUM compliant example for the
determination of the uncertainty of luminous intensity is given, but only IO-CSIC, METAS and
PTB followed that route.
At least for me, it is not possible to judge about the legitimacy of a stated uncertainty
contribution of the other participants, where no information about the model of evaluation and
the measurement process is given.
If we take the first of the eleven budgets as an example:
NMISA simply copied your example of the measurement uncertainty from the Technical
Protocol – which I supposed to be only an example to show the difference between what you
call “Systematic” and “Random” effects. (BTW, I was not in favour of this chart because it is
not strictly according to GUM). In case of NMISA the selection of possible uncertainty
components from the (already) condensed Technical-Protocol-example might be good for a
rough uncertainty estimation, but not for a meaningful demonstration of metrology at the
high level of a CCPR comparison. May be that, e.g., stray-light is the most important
environmental issue at NMISA but at least spectral mismatch of used photometer is always
an issue and may not be neglected. May be that this contribution is hidden elsewhere, but
without further information such a kind of uncertainty budget without model of evaluation is
not sufficient. It is simply not possible to judge about the correctness of the stated
uncertainties – and this is valid for all those participants showing only condensed budgets
without models of evaluation.
Moreover, also the differentiation between “random” and “systematic” effects seems not to be
generally understood in the same way. Different participants subsume different type of
components under these classifications.
Therefore, at the current stage, and without further information, I can only agree with the
uncertainty budgets from IO-CSIC, METAS and PTB.
Answers:
PTB: In the Technical Protocol for this CCPR comparison chapter 6.1.1 the GUM is
explicitly claimed as reference for any statement of measurement uncertainty.
Additionally, the chapter 6.1.2 refers to the document CIE 198 as example for
modeling combination and presentation. The protocol itself gives in Appendix A.5 an
example for an abbreviated presentation. Thus, the model of evaluation is an essential
part in the documentation and has to be stated individually by each participant as well
as the complete uncertainty budget from CIE 198 as an intermediate step for the
summarized presentation recommended in Appendix A.5 to simplify the comparison of
individual contributions.
NIM: The uncertainty assessment should be carried out in accordance with GUM, as
mentioned in 6.1.1 of the technical protocol of CCPR-K3.2014. Although CIE
198:2011 give us a good example of uncertainty assessment for luminous intensity,
other approaches in accordance with GUM should also be accepted.
Our uncertainty assessment is consistent with GUM. The protocol didn’t require
submission of mathematical models and analysis procedure. We only submit the
uncertainty budget table which condenses the procedure of uncertainty assessment.
Some insignificant uncertainty components are not listed, such as temperature effect,
etc.
NPL: See attached pages (NPL response) for NPL answers.
NMIJ: See attached pages (NMIJ response) for NMIJ answers.
NMIA:One of the more valuable aspects of comparisons is the diversity of uncertainty
budgets, allowing a full range of components to be identified by the metrology
community. If all participants were to use an identical methodology for constructing
their budgets, the possibility of identifying effects that should be included would be
reduced. We therefore believe that uniform use of CIE 198 would be a backward step.
The logic that we used in distinguishing between random and systematic components
followed the requirements of the protocol. It described systematic components as
producing their unknown values from one measurement to the next, adding that they
will probably be the same for a complete round of measurement. Given that
measurements on the comparison artefacts were performed by transfer from a set of
working standard lamps over a short period of time using a common set of
instrumentation, the majority of effects are labelled as systematic. The random
components are those associated with the complete realignment of the comparison
lamps between measurements, lamp reproducibility and noise.
NIST: See Sections 1 and 3 of “Yuqin Zong, Maria E. Nadal, Benjamin K. Tsai, and C.
Cameron Miller, “Photometric Calibrations,” NIST Special Publication 250-95. (2018).
https://doi.org/10.6028/NIST.SP.250-95 “
NMISA:
Model of evaluation:
See attached page (NMISA response) for NMISA Uncertainty Budget Matrix (UBM).
𝐾𝑚 𝑑2 𝐹𝐼𝑐
𝐼= 𝑆𝐴
where
I is the luminous intensity
Km is the luminous efficacy
D is the distance from the lamp filament to the photometer
F is the spectral mismatch factor
Ic is the current, determined for the gain of the amplifier and the voltage as
measured for the LMT photometer
S is the responsivity of the LMT photometer
A is the area of the LMT photometer
Spectral mismatch: We corrected for spectral mismatch and therefore did not include it
in the model of evaluation.
Lamp alignment: We allowed for 1° uncertainty in the alignment of the lamps, as you can see
in the model of evaluation.
NMISA
The lamp alignment uncertainty seems quite large, although it is listed as Type A uncertainty. Is
there a specific effect that produces this large uncertainty?
We allowed for 1° uncertainty in the alignment of the lamps, as you can see in the model of
evaluation.
NMIA
The NMIA budget seems too complicated. It’s a bit hard to understand it without the facility and
measurement procedure description. I hope this description will appear in the Draft A report.
We have attempted a comprehensive evaluation of all the effects that could influence our
measurements, and their associated uncertainties. Quite a few of them have been evaluated as zero
(to the number of significant digits in our budget) but we considered it worthwhile to leave them in
the budget since they had been considered. It is true that the budget could also have been simplified
by replacing groups of related components with single combined values, but we believe that it was
better to provide a detailed breakdown.
We acknowledge that a description of the measurement facility and process is an important part of
assessing the budget. We have written our report and will be very happy to modify or extend it as
required to provide the information requested by other participants during the relevant part of the
report preparation.
NMIJ
The random effects section includes a lamp ageing component which is quite large and dominates
this part of the budget. Could you explain what this means?
See attached pages (NMIJ response) for NMIJ answers.
IO-CSIC
It is not stated which components are considered as random and which are systematic in the context
of the Appendix A6 table. Could, for example, an extra column be added to the Appendix A5 table
giving that classification?
IO-CSIC has submitted a revised Appendix A.5 in which an extra column has been added. From
column 3 of the original table, all Type A (2) have been labelled as Random and all Type B have
been labelled as Systematic. See Appendix A of this comparison report.
LNE-CNAM
The lamp alignment uncertainty seems quite large, although it is listed as Type A uncertainty. Is
there a specific effect that produces this large uncertainty?
METAS
It is not stated which components are considered as random and which are systematic in the context
of the Appendix A6 table. Could the components numbers in the Appendix A5 table be listed as
random or systematic (or is that what the asterisks in that table indicate)?
The answer is yes, the * indicates the random effects.
NPL
1. Is it possible that the ‘stabiliser current control’ contributes to the ‘test lamp repeatability’
component, meaning that there is some double counting in the random effects?
Yes, it is impossible to isolate the effect of ‘stabiliser current control’ from ‘test lamp
repeatability’ so there is potentially a small element of double counting in the random effects.
However since the test lamp repeatability component is intended primarily to allow for lamp
alignment variations and is treated as a worst case estimate, we have chosen to ignore this small
element of double counting. The effect on the final uncertainty is insignificant.
2. If they know the ageing rate of the lamp, it is better to correct the luminous intensity according to
the ageing rate. If they do the correction, the uncertainty will be smaller than 0.125%.
We do not know the actual change in luminous intensity due to ageing for each individual
reference lamp used. Each reference lamp has been used for a different length of time since the
original calibration and will also age at a slightly different (unknown) rate. We therefore do not
correct for ageing effects. The uncertainty estimate is a conservative allowance, which is based
on measurements on other lamps of the same type operated at the same correlated colour
temperature and under the same conditions coupled with knowledge of the maximum length of
time for which the reference lamps have been used since the original calibration.
3. Some expressions are not consistent with the requirements of GUM “Evaluation of measurement
data-Guide to the expression of uncertainty in measurement”, such as “value ±”, “ui”,
“Photocurrent measurement accuracy”.
- “ui”should be “ui”
- “Photocurrent measurement accuracy” should be “Photocurrent measurement”
We apologise for these typing mistakes, which were due to importing the table from an Excel
file. We have provided a corrected version of the report to the pilot laboratory.
4. The uncertainty of lamp alignment is not in the list.
The uncertainty due to lamp alignment is included under ‘Test lamp repeatability’ as described in
section 5.10 of our report.
PTB
Please clarify the two lines labelled “variance” at the end of the Appendix A5 table. Do they indicate
the distinction between the random and systematic uncertainties in the context of the comparison?
Similarly, could the components in the Appendix A5 table be labelled somehow (or a separate list be
given) as contributing to the random and systematic uncertainties?
a) According to the GUM all entries in Appendix A.5 are labeled in column 6 with “A” for
“statistical” or “B” for any “other determination”. These types of entries are combined and
listed separately for each lamp. The list was send to the pilot for an additional explanation
and mean values u(A) = 0.12% and u(B) = 0.13% are indicated. Thus, the combined
standard uncertainty for the transfer by only one lamp is 𝑢(𝐼) = √𝑢(A)2 + 𝑢(B)2 =
0.18%.
b) At the bottom of the table Appendix A.5 two values labeled udev = 0.12% for random
(“dev” for devise) and uinst = 0.14% for systematic contributions (“inst” for
instrumentation) are included. These numbers, their meaning and the evaluation are
explained in all details in the publication CIE 198-SP1.2:2011 (see chapter/example 2.13).
The combination of these numbers to determine the uncertainty of the whole batch for the
transferred value of intensity is explained in great detail in CIE 198-SP1.1:2011 example
1.11.
It turns out that the instrumentation for the two rounds at PTB was stable and the properties of
the PTB-transfer-standards (WI41/G) are uniform. A separation in types A and B or
“random” and “systematic” gives no real difference. So, the uncertainty u(PTB) associated
with the luminous intensity value transferred by the batch with a number of 6 PTB-
transfer standards will be determined by the pilot laboratory from
2
𝑢𝑑𝑒𝑣
𝑢(𝐼)PTB = √𝑢𝑖𝑛𝑠𝑡
2
+ = 0.15%.
6
NIST
The values for the systematic component given in the second table (±0.20%) appear to derive only
from the illuminance unit realisation. Should the long-term drift of the photometers be included in
the systematic component?
As shown in Table 8 of “Yuqin Zong, Maria E. Nadal, Benjamin K. Tsai, and C. Cameron
Miller, “Photometric Calibrations,” NIST Special Publication 250-95. (2018).
https://doi.org/10.6028/NIST.SP.250-95“, we do have a longterm drift component but these
measurements were done within 1 month of the scale realization so there is no longterm drift.
NRC
There is no lamp alignment component – is it included in the lamp reproducibility term? Please also
clarify the meaning of the term labelled as “Lamp Maintenance / aging”.
The lamp alignment component is included in the lamp reproducibility term.
The term ‘Lamp Maintenance/aging’ is an estimate of the uncertainty in the luminous intensity
values of the lamps used for the comparison since the time that they were calibrated. This is
predominantly an estimate of the aging of the lamps due to use since they were calibrated.
Random effects**:
Electrical parameters:
- power supply fluctuations, upsf B 0.02
Photometer noise (25 readings), unoise A 0.001
Discrepancy between three photometers (random), upd A 0.05
Independent measurement reproducibility *** , urep A 0.05
**** Uncertainty associated with reproducibility (urep) partly includes uncertainties associated with
other random effects. Therefore, the Total random uncertainty is calculated as
Measurement parameters given in this table are suggested. Please modify and itemize according to
your particular situation. See Section 6.2 for explanation of the various items.
Note that if lamps are used as the laboratory working standards, a group of uncertainties would need
to be included in the above table to account for their behaviour.
The RMS total refers to the usual square root of the sum of the squares of all the individual
uncertainty terms.
1. The lamps used for the comparison were calibrated directly against NPL’s primary reference
standard luminous intensity lamps, which are of exactly the same type as the comparison
lamps. Any reflections from the inside edges of baffles or shutters are therefore common to
both the reference and comparison lamps and the effects cancel; no correction is necessary.
Extensive investigations into stray light effects (including light scattered, reflected or
diffracted by apertures and baffles) were carried out during the realisation of the luminous
intensity scale and assessed to be less than 0.01 % - this is included in the uncertainty
budget for NPL’s realisation of the candela.
2. The alignment of the NPL photometer was not changed between the calibration using the
reference lamps and the measurements of the comparison lamps; therefore it is not
necessary to include an uncertainty component for misalignment of the photometer
aperture.
3. NPL did not follow the model given in CIE 198:2011 since this is not how we usually structure
our uncertainty budget. We did, however, provide a detailed description of each of the
uncertainty contributions included in our uncertainty budget, which we believe gives the
information necessary to judge the legitimacy of each of these. For completeness, our
measurement equation is given below (this has also been added to our measurement
report):
𝐼v,𝑡 = 𝐶cal 𝑉𝑡 (1 + 𝐶d,𝑡 )(1 + 𝐶J,𝑡 )(1 + 𝐶p,𝑡 )𝐹SM,𝑡 (1 − 𝐶stray,𝑡 )(1 + 𝐶align,𝑡 ) (1)
where
(𝐼v,𝑟 +𝐶age,𝑟 )
𝐶cal = 𝑉𝑟
(2)
and
current for the reference lamp to the specified value is included in the uncertainty
budget for the luminous intensity of the reference lamp)
𝐶p,𝑡 is the error in luminous intensity for test lamp t due to differences in amplifier gain
and DVM sensitivity between measurement of the photocurrent produced by the
reference lamp and that produced by the test lamp
𝐹SM,𝑡 is the spectral mismatch correction factor for test lamp t
𝐶stray,𝑡 is the error in luminous intensity for test lamp t due to differences in stray light
between the reference and test lamps
𝐶align,𝑡 is the error in luminous intensity for test lamp t due to misalignment of the lamp
(the uncertainty due to misalignment of the reference lamp is included in the
uncertainty budget for the luminous intensity of the reference lamp)
Note all of the C terms listed above have an expected value of zero and an associated uncertainty
that has been estimated as described in our measurement report.
1. Yes, it is impossible to isolate the effect of ‘stabiliser current control’ from ‘test lamp
repeatability’ so there is potentially a small element of double counting in the random
effects. However since the test lamp repeatability component is intended primarily to allow
for lamp alignment variations and is treated as a worst case estimate, we have chosen to
ignore this small element of double counting. The effect on the final uncertainty is
insignificant.
2. We do not know the actual change in luminous intensity due to ageing for each individual
reference lamp used. Each reference lamp has been used for a different length of time since
the original calibration and will also age at a slightly different (unknown) rate. We therefore
do not correct for ageing effects. The uncertainty estimate is a conservative allowance,
which is based on measurements on other lamps of the same type operated at the same
correlated colour temperature and under the same conditions coupled with knowledge of
the maximum length of time for which the reference lamps have been used since the
original calibration.
3. We apologise for these typing mistakes, which were due to importing the table from an
Excel file. We have provided a corrected version of the report to the pilot laboratory.
4. The uncertainty due to lamp alignment is included under ‘Test lamp repeatability’ as
described in section 5.10 of our report.
The following equations are the physical model of uncertainty of luminous intensity at NMIJ.
λ2
K (d + ∆d1 ) V0
I1 = m 1
2
∫λ Φ
1
e ,λ ( λ ) V ( λ ) dλ
(1 + ct ) (1 + c1 ) (1)
λ
A G
∫λ Φ ( λ ) s e ( λ ) dλ
2
e ,λ
1
V2 (d 2 + ∆d 2 )
2
I 2 = I1 k c 2
(1 + ci ) (1 + ca ) (1 + c2 ) (1 + c3 ) (2)
V1 d2
Equation (1) is the model to determine the luminous intensity of the standard lamp. Equation (2) is the
model to transfer luminous intensity from the standard lamp to the transfer lamp. The meanings of each
variable are listed below.
d1 : Distance between the standard lamp and the standard photometer. Constant. No uncertainty.
negligible.
V (λ ) : Luminous efficiency function. No uncertainty.
se (λ ) : Spectral responsivity of the standard photometer. Uncertainty of this factor consists of two parts
in the budget. One is "Spectral responsivity of the silicon photodiode measured with the cryogenic
radiometer", and another is " Illuminance responsivity of the standard photometer with respect to the
spectral responsivity of the silicon photodiode".
ct : Deviation of the standard photometer responsivity by the room temperature.
c1 : Deviation of the luminous intensity measurement for the standard lamp set on and removed from the
lamp mount in many times. Accumulated data.
kc : Colour correction factor between the standard lamp and the transfer lamp. Uncertainty negligible.
ca : Deviation of luminous intensity through the period of recalibraion-limit burning time. We take this
effect into the uncertainty without correction. So it is listed in "Random effects" because we cannot predict
what value a lamp will take at each burning.
c 2 : Deviation of the luminous intensity measurement for the transfer lamp set on and removed from the
lamp mount in many times. Accumulated data.
c3 : Fluctuation of lamp signal.
Random effects:
Lamp parameters:
- lamp ageing B 0.11
ca
- lamp alignment (*) B 0.06
c2
- lamp reproducibility included in (*)
- lamp output fluctuations B 0.02
c3
Electrical parameters:
- power supply fluctuations included in (*)
Photometer noise included in (*)
(Measurement Set standard deviation of mean)
The effect of baffles is regarded as negligibly small. We expect that that effect can be as small as 0.007 %,
which is negligible in the NMIJ's uncertainty budget.
Metrologist
Type & Serial
Description: CCPR-K3 Luminous Intensity Intercomparison Range:
Number
Mathematical Model:
▼ Unit Under Test / Calibration (Uncorrelated) ▼ NOTE! ONLY CHANGE BLUE CELLS - All OTHER CELLS (WHITE) ARE PROTECTED
UBM V5.03 - CSIR-National Metrology Laboratory - dc Low Frequency (Designed by BvO) Final UB for CCPR-K3.a Uncertainty Budget Matrix
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report, Appendices C,D,E,F Information
CCPR-K3.2014: Luminous Intensity
Draft B Report
2020-October-15
Information for Appendices C-F
Summary of Measurement Data and Analysis
(k=1) values are used for all calculations
Appendix C
Summary of Participant Lamp Luminous Intensity Values
This worksheet contains the Luminous Intensity values for all the participant lamps
The values for Round#1, Round#2, and the final values have been determined as discussed in the Draft B report Section 4.2 "Participant Lamp Data"
The worksheet shows the calculations for the average NMI Luminous Intensity relative standard uncertainty
The lamp final data is linked to subsequent worksheets and calculations
Appendix D
Summary of Pilot Measurements of Participant Lamps
This worksheet combines the 'final' NMI luminous intensity values (from Appendix C) with the Pilot measurements of each lamp
The final NMI value for the comparison photometer responsivity (cd/V) is calculated from all the NMI lamps and Pilot measurements
Uncertainties are calculated as discussed in the Draft B report Sections 4.1 to 4.3
Appendix E
Calculation of the KCRV and the DoE
The data from worksheet Appendices C and D is gathered for the calculation of the KCRV and Unilateral DOE
The calculations are discussed in the Draft B report Section 4.4
If any changes are made, such as to the Mandel-Paule factor s,
-the Chi-square values, KCRV value, DOE values and uncertainties will all change.
Worksheet "DOE.plt"
Plot of the DOE values, uncertainties, and KCRV uncertainty, data from Appendix E
This is similar to Figure Two of the Draft B report
Appendix F
Calculation of the Bilateral DoE
The data from Appendix E is used to calculate the bilateral DOE as described in the Draft B report Section 4.4.3
CCPR-K3.2014: Luminous Intensity Draft B Page 1 of 15 CCPR-K3-2014DraftBAppendicesCDEF.xlsx
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report, Appendices C,D,E,F
Appendix_Cv2.1
A B C D E F G H I J K L M N O P Q
1 CCPR-K3.2014: Luminous Intensity Fractional Split of NMI
2 Draft B Report random uncertainties
3 2020-October-15 into uncorrelated and
4 Appendix Cv2.1 correlated components
5 Summary of Participant Lamp Luminous Intensity Values for combining lamps
6 fraction f SQRT(1-f^2)
7 0.5 0.866
8
9
10 NMI: NMISA
11
12 Lamp# Round# Data ID Lamp Electrical Lamp CCT NMISA Lamp Luminous Intensity (cd) Calculations for NMISA weighted mean
13 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
14 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
15 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
16
17 “24” 4595 PTB 09 R#1 “24” 4595 PTB 09R#1 5.824 30.242 2841.0 269.000 0.156% 0.653% 0.680%
18 R#2 “24” 4595 PTB 09R#2 5.824 30.254 2838.0 268.700 0.156% 0.653% 0.680%
19 final “24” 4595 PTB 09final 5.824 30.248 2839.5 268.850 0.110% 0.653% 0.663% 0.006625 22782.27568 0.250000 0.000138 0.001651
20
21 “39” 4596 PTB 09 R#1 “39” 4596 PTB 09R#1 5.892 30.816 2853.0 283.900 0.156% 0.653% 0.680%
22 R#2 “39” 4596 PTB 09R#2 5.892 30.826 2849.0 284.400 0.156% 0.653% 0.680%
23 final “39” 4596 PTB 09final 5.892 30.821 2851.0 284.150 0.110% 0.653% 0.663% 0.006625 22782.27538 0.250000 0.000138 0.001651
24
25 “42” 4597 PTB 09 R#1 “42” 4597 PTB 09R#1 5.880 30.713 2848.0 274.600 0.156% 0.653% 0.680%
26 R#2 “42” 4597 PTB 09R#2 5.880 30.725 2844.0 277.100 0.156% 0.653% 0.680%
27 final “42” 4597 PTB 09final 5.880 30.719 2846.0 275.839 0.110% 0.653% 0.663% 0.006625 22782.26294 0.250000 0.000138 0.001651
28
29 NSI 10 R#1 NSI 10R#1 5.890 31.962 2854.0 314.400 0.156% 0.653% 0.680%
30 R#2 NSI 10R#2 5.890 31.944 2869.0 317.200 0.156% 0.653% 0.680%
31 final NSI 10final 5.890 31.953 2861.5 315.788 0.110% 0.653% 0.663% 0.006625 22782.26349 0.250000 0.000138 0.001651
32
33 Average NMISA Luminous Intensity relative standard uncertainty sum: 91129.07749 1.00000
34 NMISA 0.028% 0.660% 0.661%
35 Final NMISA average relative standard uncertainty
36 u-uncorr u-corr uf
37 NMISA 0.0275% 0.6602% 0.6608%
38
39
40 NMI: NIM
41
42 Lamp# Round# Data ID Lamp Electrical Lamp CCT NIM Lamp Luminous Intensity (cd) Calculations for NIM weighted mean
43 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
44 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
45 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
46
47 NIM-01(Wi41/G-96) R#1 NIM-01(Wi41/G-96)R#1 5.794 29.846 2837.0 253.145 0.035% 0.167% 0.171%
48 R#2 NIM-01(Wi41/G-96)R#2 5.794 29.828 2837.0 252.660 0.057% 0.167% 0.176%
49 final NIM-01(Wi41/G-96)final 5.794 29.837 2837.0 253.012 0.030% 0.167% 0.170% 0.001696 347681.05909 0.197256 0.000029 0.000333
50
51 NIM-02(Wi41/G-152) R#1 NIM-02(Wi41/G-152)R#1 5.818 30.013 2829.0 263.580 0.032% 0.167% 0.170%
52 R#2 NIM-02(Wi41/G-152)R#2 5.818 30.020 2829.0 263.925 0.057% 0.167% 0.177%
53 final NIM-02(Wi41/G-152)final 5.818 30.016 2829.0 263.660 0.028% 0.167% 0.169% 0.001693 349030.11258 0.198021 0.000027 0.000334
54
55 NIM-03(Wi41/G-164) R#1 NIM-03(Wi41/G-164)R#1 5.807 29.781 2841.0 275.150 0.008% 0.167% 0.167%
56 R#2 NIM-03(Wi41/G-164)R#2 5.807 29.773 2841.0 275.600 0.048% 0.167% 0.174%
57 final NIM-03(Wi41/G-164)final 5.807 29.777 2841.0 275.164 0.008% 0.167% 0.167% 0.001672 357668.81541 0.202922 0.000008 0.000339
58
59 NIM-04(Wi41/G-180) R#1 NIM-04(Wi41/G-180)R#1 5.804 29.954 2839.0 265.172 0.020% 0.167% 0.168%
60 R#2 NIM-04(Wi41/G-180)R#2 5.804 29.947 2839.0 265.620 0.043% 0.167% 0.172%
61 final NIM-04(Wi41/G-180)final 5.804 29.950 2839.0 265.251 0.018% 0.167% 0.168% 0.001680 354380.97063 0.201057 0.000018 0.000337
62
63 NIM-05(Wi41/G-189) R#1 NIM-05(Wi41/G-189)R#1 5.780 29.730 2840.0 269.570 0.024% 0.167% 0.169%
A B C D E F G H I J K L M N O P Q
64 R#2 NIM-05(Wi41/G-189)R#2 5.780 29.711 2840.0 269.415 0.034% 0.167% 0.170%
65 final NIM-05(Wi41/G-189)final 5.780 29.721 2840.0 269.520 0.019% 0.167% 0.168% 0.001681 353828.27608 0.200743 0.000019 0.000337
66
67 Average NIM Luminous Intensity relative standard uncertainty sum: 1762589.23378 1.00000
68 NIM 0.005% 0.168% 0.168%
69 Final NIM average relative standard uncertainty:
70 u-uncorr u-corr uf
71 NIM 0.0049% 0.1681% 0.1681%
72
73
74 NMI: NMIA
75
76 Lamp# Round# Data ID Lamp Electrical Lamp CCT NMIA Lamp Luminous Intensity (cd) Calculations for NMIA weighted mean
77 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
78 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
79 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
80
81 S7 R#1 S7R#1 5.780 31.724 2856.0 298.759 0.009% 0.153% 0.153%
82 R#2 S7R#2 5.780 31.736 2856.0 298.716 0.008% 0.151% 0.151%
83 final S7final 5.780 31.730 2856.0 298.735 0.006% 0.152% 0.152% 0.001520 432773.37472 0.202861 0.000006 0.000308
84
85 350 LI3 R#1 350 LI3R#1 5.794 31.741 2856.0 298.447 0.016% 0.153% 0.154%
86 R#2 350 LI3R#2 5.794 31.751 2856.0 298.684 0.019% 0.151% 0.152%
87 final 350 LI3final 5.794 31.746 2856.0 298.551 0.012% 0.152% 0.153% 0.001526 429240.69680 0.201205 0.000012 0.000307
88
89 318 SI2 R#1 318 SI2R#1 5.781 31.722 2856.0 305.807 0.021% 0.153% 0.154%
90 R#2 318 SI2R#2 5.781 31.736 2856.0 305.845 0.018% 0.151% 0.152%
91 final 318 SI2final 5.781 31.729 2856.0 305.829 0.014% 0.152% 0.152% 0.001524 430298.79892 0.201701 0.000014 0.000307
92
93 306 S15 R#1 306 S15R#1 5.858 32.078 2856.0 308.499 0.017% 0.153% 0.154%
94 R#2 306 S15R#2 5.858 32.096 2856.0 308.601 0.016% 0.151% 0.152%
95 final 306 S15final 5.858 32.087 2856.0 308.551 0.012% 0.152% 0.152% 0.001524 430540.03093 0.201814 0.000012 0.000307
96
97 288 SI4 R#1 288 SI4R#1 5.786 31.672 2856.0 301.606 0.053% 0.153% 0.162%
98 R#2 288 SI4R#2 5.786 31.668 2856.0 301.514 0.048% 0.151% 0.159%
99 final 288 SI4final 5.786 31.670 2856.0 301.555 0.036% 0.152% 0.156% 0.001561 410497.16280 0.192419 0.000034 0.000298
100
101 Average NMIA Luminous Intensity relative standard uncertainty sum: 2133350.06417 1.00000
102 NMIA 0.004% 0.153% 0.153%
103 Final NMIA average relative standard uncertainty:
104 u-uncorr u-corr uf
105 NMIA 0.0041% 0.1528% 0.1529%
106
107
108 NMI: NMIJ
109
110 Lamp# Round# Data ID Lamp Electrical Lamp CCT NMIJ Lamp Luminous Intensity (cd) Calculations for NMIJ weighted mean
111 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
112 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
113 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
114
115 #37 R#1 #37R#1 5.756 29.069 2800.0 242.145 0.090% 0.256% 0.271%
116 R#2 #37R#2 5.756 29.064 2800.0 242.155 0.090% 0.256% 0.271%
117 final #37final 5.756 29.067 2800.0 242.150 0.064% 0.256% 0.264% 0.002638 143743.73352 0.200000 0.000064 0.000524
118
119 #40 R#1 #40R#1 5.794 29.550 2800.0 250.505 0.090% 0.256% 0.271%
120 R#2 #40R#2 5.794 29.544 2800.0 250.285 0.090% 0.256% 0.271%
121 final #40final 5.794 29.547 2800.0 250.395 0.064% 0.256% 0.264% 0.002638 143743.73173 0.200000 0.000064 0.000524
122
123 #51 R#1 #51R#1 5.736 29.264 2800.0 240.850 0.090% 0.256% 0.271%
124 R#2 #51R#2 5.736 29.262 2800.0 240.565 0.090% 0.256% 0.271%
125 final #51final 5.736 29.263 2800.0 240.707 0.064% 0.256% 0.264% 0.002638 143743.73045 0.200000 0.000064 0.000524
126
A B C D E F G H I J K L M N O P Q
127 #52 R#1 #52R#1 5.765 29.167 2800.0 241.500 0.090% 0.256% 0.271%
128 R#2 #52R#2 5.765 29.160 2800.0 241.490 0.090% 0.256% 0.271%
129 final #52final 5.765 29.163 2800.0 241.495 0.064% 0.256% 0.264% 0.002638 143743.73365 0.200000 0.000064 0.000524
130
131 #58 R#1 #58R#1 5.610 29.970 2800.0 244.280 0.090% 0.256% 0.271%
132 R#2 #58R#2 5.610 29.965 2800.0 244.505 0.090% 0.256% 0.271%
133 final #58final 5.610 29.967 2800.0 244.392 0.064% 0.256% 0.264% 0.002638 143743.73190 0.200000 0.000064 0.000524
134
135 Average NMIJ Luminous Intensity relative standard uncertainty sum: 718718.66125 1.00000
136 NMIJ 0.014% 0.262% 0.262%
137 Final NMIJ average relative standard uncertainty:
138 u-uncorr u-corr uf
139 NMIJ 0.0142% 0.2618% 0.2622%
140
141
142 NMI: IO-CSIC
143
144 Lamp# Round# Data ID Lamp Electrical Lamp CCT IO-CSIC Lamp Luminous Intensity (cd) Calculations for IO-CSIC weighted mean
145 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
146 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
147 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
148
149 Wi95A R#1 Wi95AR#1 5.836 30.807 2869.0 278.405 0.007% 0.310% 0.310%
150 R#2 Wi95AR#2 5.836 30.930 2869.0 278.184 0.007% 0.310% 0.310%
151 final Wi95Afinal 5.836 30.868 2869.0 278.294 0.005% 0.310% 0.310% 0.003100 104032.63337 0.200005 0.000005 0.000620
152
153 Wi95B R#1 Wi95BR#1 5.836 31.083 2868.0 285.515 0.004% 0.310% 0.310%
154 R#2 Wi95BR#2 5.837 31.127 2868.0 284.322 0.007% 0.310% 0.310%
155 final Wi95Bfinal 5.836 31.105 2868.0 285.188 0.003% 0.310% 0.310% 0.003100 104045.14922 0.200029 0.000003 0.000620
156
157 Wi95C R#1 Wi95CR#1 5.832 30.790 2862.0 286.693 0.003% 0.310% 0.310%
158 R#2 Wi95CR#2 5.832 30.860 2862.0 286.268 0.007% 0.310% 0.310%
159 final Wi95Cfinal 5.832 30.825 2862.0 286.637 0.002% 0.310% 0.310% 0.003100 104051.55216 0.200042 0.000002 0.000620
160
161 Wi95D R#1 Wi95DR#1 5.836 30.587 2868.0 271.761 0.004% 0.310% 0.310%
162 R#2 Wi95DR#2 5.836 30.633 2868.0 270.688 0.004% 0.310% 0.310%
163 final Wi95Dfinal 5.836 30.610 2868.0 271.222 0.003% 0.310% 0.310% 0.003100 104050.01451 0.200039 0.000003 0.000620
164
165 A454 R#1 A454R#1 25.500 12.247 2844.0 433.167 0.013% 0.310% 0.310%
166 R#2 A454R#2 25.501 12.267 2844.0 434.636 0.013% 0.310% 0.310%
167 final A454final 25.500 12.257 2844.0 433.899 0.009% 0.310% 0.310% 0.003101 103969.75394 0.199885 0.000009 0.000620
168
169 Average IO-CSIC Luminous Intensity relative standard uncertainty sum: 520149.10320 1.00000
170 IO-CSIC 0.001% 0.310% 0.310%
171 Final IO-CSIC average relative standard uncertainty:
172 u-uncorr u-corr uf
173 IO-CSIC 0.0011% 0.3100% 0.3100%
174
175
176 NMI: LNE-CNAM
177
178 Lamp# Round# Data ID Lamp Electrical Lamp CCT LNE-CNAM Lamp Luminous Intensity (cd) Calculations for LNE-CNAM weighted mean
179 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
180 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
181 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
182
183 926 R#1 926R#1 5.690 29.010 2796.0 234.400 0.220% 0.220% 0.311%
184 R#2 926R#2 5.690 28.970 2796.0 233.800 0.240% 0.220% 0.326%
185 final 926final 5.690 28.990 2796.0 234.125 0.162% 0.220% 0.273% 0.002733 133868.05548 0.342036 0.000277 0.000893
186
187 936 R#1 936R#1 5.690 29.150 2799.0 241.800 0.230% 0.220% 0.318%
188 R#2 936R#2 5.690 29.100 2799.0 241.200 0.290% 0.220% 0.364%
189 final 936final 5.690 29.125 2799.0 241.568 0.180% 0.220% 0.284% 0.002844 123649.59621 0.315928 0.000285 0.000852
A B C D E F G H I J K L M N O P Q
190
191 A430 R#1 A430R#1 25.000 11.950 2815.0 397.300 0.220% 0.220% 0.311%
192 R#2 A430R#2 25.000 11.960 2815.0 397.400 0.240% 0.220% 0.326%
193 final A430final 25.000 11.955 2815.0 397.346 0.162% 0.220% 0.273% 0.002733 133868.13157 0.342036 0.000277 0.000893
194
195 Average LNE-CNAM Luminous Intensity relative standard uncertainty sum: 391385.78326 1.00000
196 LNE-CNAM 0.048% 0.264% 0.268%
197 Final LNE-CNAM average relative standard uncertainty:
198 u-uncorr u-corr uf
199 LNE-CNAM 0.0485% 0.2638% 0.2682%
200
201
202 NMI: METAS
203
204 Lamp# Round# Data ID Lamp Electrical Lamp CCT METAS Lamp Luminous Intensity (cd) Calculations for METAS weighted mean
205 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
206 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
207 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
208
209 506 R#1 506R#1 5.760 30.559 2855.7 276.229 0.038% 0.312% 0.315%
210 R#2 506R#2 5.760 30.558 2855.7 276.193 0.038% 0.312% 0.315%
211 final 506final 5.760 30.558 2855.7 276.211 0.027% 0.312% 0.314% 0.003135 101728.56832 0.166724 0.000022 0.000522
212
213 684 R#1 684R#1 5.680 30.687 2854.4 277.966 0.048% 0.312% 0.316%
214 R#2 684R#2 5.680 30.686 2854.4 277.881 0.038% 0.312% 0.315%
215 final 684final 5.680 30.686 2854.4 277.914 0.030% 0.312% 0.314% 0.003138 101525.34990 0.166390 0.000025 0.000522
216
217 841 R#1 841R#1 5.860 30.341 2858.3 280.875 0.038% 0.312% 0.315%
218 R#2 841R#2 5.860 30.336 2858.3 280.311 0.037% 0.312% 0.315%
219 final 841final 5.860 30.339 2858.3 280.587 0.027% 0.312% 0.314% 0.003135 101726.38529 0.166720 0.000022 0.000522
220
221 1060 R#1 1060R#1 5.850 30.325 2841.0 272.256 0.038% 0.312% 0.315%
222 R#2 1060R#2 5.850 30.338 2841.0 272.986 0.038% 0.312% 0.315%
223 final 1060final 5.850 30.332 2841.0 272.627 0.027% 0.312% 0.314% 0.003136 101713.53497 0.166699 0.000022 0.000522
224
225 1063 R#1 1063R#1 5.900 30.557 2854.5 283.982 0.038% 0.312% 0.315%
226 R#2 1063R#2 5.900 30.568 2854.5 284.402 0.038% 0.312% 0.315%
227 final 1063final 5.900 30.562 2854.5 284.197 0.027% 0.312% 0.314% 0.003136 101714.96298 0.166701 0.000022 0.000522
228
229 1064 R#1 1064R#1 5.900 30.679 2854.8 287.908 0.037% 0.312% 0.315%
230 R#2 1064R#2 5.900 30.692 2854.8 288.563 0.037% 0.312% 0.315%
231 final 1064final 5.900 30.686 2854.8 288.232 0.026% 0.312% 0.313% 0.003135 101754.35247 0.166766 0.000022 0.000522
232
233 Average METAS Luminous Intensity relative standard uncertainty sum: 610163.15394 1.00000
234 METAS 0.006% 0.313% 0.313%
235 Final METAS average relative standard uncertainty:
236 u-uncorr u-corr uf
237 METAS 0.0056% 0.3133% 0.3133%
238
239
240 NMI: NPL
241
242 Lamp# Round# Data ID Lamp Electrical Lamp CCT NPL Lamp Luminous Intensity (cd) Calculations for NPL weighted mean
243 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
244 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
245 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
246
247 A644 R#1 A644R#1 25.360 12.505 2850.0 451.78 0.082% 0.158% 0.178%
248 R#2 A644R#2 25.360 12.500 2850.0 451.97 0.082% 0.158% 0.178%
249 final A644final 25.360 12.503 2850.0 451.87 0.058% 0.158% 0.168% 0.001683 353032.54775 0.208864 0.000061 0.000346
250
251 A647 R#1 A647R#1 25.310 12.510 2850.0 459.43 0.082% 0.158% 0.178%
252 R#2 A647R#2 25.310 12.533 2850.0 459.63 0.082% 0.158% 0.178%
A B C D E F G H I J K L M N O P Q
253 final A647final 25.310 12.522 2850.0 459.53 0.058% 0.158% 0.168% 0.001683 353032.54762 0.208864 0.000061 0.000346
254
255 PA758 R#1 PA758R#1 25.220 12.743 2850.0 460.33 0.082% 0.158% 0.178%
256 R#2 PA758R#2 25.220 12.751 2850.0 460.70 0.082% 0.158% 0.178%
257 final PA758final 25.220 12.747 2850.0 460.51 0.058% 0.158% 0.168% 0.001683 353032.54284 0.208864 0.000061 0.000346
258
259 877 R#1 877R#1 5.818 30.013 2853.0 276.34 0.082% 0.158% 0.178%
260 final 877final 5.818 30.013 2853.0 276.34 0.082% 0.158% 0.178% 0.001780 315576.87453 0.186704 0.000077 0.000323
261
262 890 R#1 890R#1 5.804 29.871 2853.0 273.93 0.082% 0.158% 0.178%
263 final 890final 5.804 29.871 2853.0 273.93 0.082% 0.158% 0.178% 0.001780 315576.87453 0.186704 0.000077 0.000323
264
265 Average NPL Luminous Intensity relative standard uncertainty sum: 1690251.38726 1.00000
266 NPL 0.015% 0.169% 0.169%
267 Final NPL average relative standard uncertainty:
268 u-uncorr u-corr uf
269 NPL 0.0151% 0.1686% 0.1692%
270
271
272 NMI: PTB
273
274 Lamp# Round# Data ID Lamp Electrical Lamp CCT PTB Lamp Luminous Intensity (cd) Calculations for PTB weighted mean
275 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
276 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
277 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
278
279 759 R#1 759R#1 5.650 29.123 2800.0 236.210 0.120% 0.130% 0.180%
280 R#2 759R#2 5.650 29.123 2800.0 236.220 0.130% 0.130% 0.180%
281 final 759final 5.650 29.123 2800.0 236.215 0.088% 0.130% 0.157% 0.001571 405267.17848 0.164702 0.000073 0.000248
282
283 791 R#1 791R#1 5.650 29.564 2800.0 247.550 0.120% 0.130% 0.180%
284 R#2 791R#2 5.650 29.565 2800.0 247.530 0.120% 0.130% 0.180%
285 final 791final 5.650 29.565 2800.0 247.540 0.085% 0.130% 0.155% 0.001552 414937.75913 0.168632 0.000072 0.000252
286
287 793 R#1 793R#1 5.650 29.387 2800.0 245.970 0.120% 0.130% 0.180%
288 R#2 793R#2 5.650 29.387 2800.0 246.000 0.130% 0.130% 0.180%
289 final 793final 5.650 29.387 2800.0 245.984 0.088% 0.130% 0.157% 0.001571 405267.17807 0.164702 0.000073 0.000248
290
291 848 R#1 848R#1 5.700 28.573 2810.0 228.530 0.120% 0.130% 0.180%
292 R#2 848R#2 5.700 28.571 2810.0 228.540 0.120% 0.130% 0.180%
293 final 848final 5.700 28.572 2810.0 228.535 0.085% 0.130% 0.155% 0.001552 414937.75928 0.168632 0.000072 0.000252
294
295 851 R#1 851R#1 5.700 28.932 2815.0 233.490 0.120% 0.130% 0.180%
296 R#2 851R#2 5.700 28.931 2815.0 233.540 0.120% 0.130% 0.180%
297 final 851final 5.700 28.931 2815.0 233.515 0.085% 0.130% 0.155% 0.001552 414937.75792 0.168632 0.000072 0.000252
298
299 858 R#1 858R#1 5.700 28.561 2800.0 225.120 0.120% 0.130% 0.180%
300 R#2 858R#2 5.700 28.562 2800.0 225.010 0.130% 0.130% 0.180%
301 final 858final 5.700 28.561 2800.0 225.069 0.088% 0.130% 0.157% 0.001571 405267.17096 0.164702 0.000073 0.000248
302
303 Average PTB Luminous Intensity relative standard uncertainty sum: 2460614.80383 1.00000
304 PTB 0.018% 0.150% 0.151%
305 Final PTB average relative standard uncertainty:
306 u-uncorr u-corr uf
307 PTB 0.0177% 0.1500% 0.1511%
308
309
310 NMI: VNIIOFI
311
312 Lamp# Round# Data ID Lamp Electrical Lamp CCT VNIIOFI Lamp Luminous Intensity (cd) Calculations for VNIIOFI weighted mean
313 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
314 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
315 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
A B C D E F G H I J K L M N O P Q
316
317 3281 R#1 3281R#1 5.8800 29.952 2853.9 273.480 0.060% 0.240% 0.250%
318 R#2 3281R#2 5.8800 29.943 2853.9 275.060 0.060% 0.250% 0.260%
319 final 3281final 5.8800 29.948 2853.9 274.265 0.042% 0.245% 0.249% 0.002486 161765.29098 0.166687 0.000035 0.000413
320
321 3282 R#1 3282R#1 5.8000 30.547 2854.3 276.870 0.060% 0.240% 0.250%
322 R#2 3282R#2 5.8000 30.541 2854.3 276.880 0.060% 0.250% 0.260%
323 final 3282final 5.8000 30.544 2854.3 276.875 0.042% 0.245% 0.249% 0.002486 161746.98190 0.166668 0.000035 0.000413
324
325 N 01 R#1 N 01R#1 5.8800 30.419 2855.8 287.190 0.060% 0.240% 0.250%
326 R#2 N 01R#2 5.8800 30.413 2855.8 286.540 0.060% 0.250% 0.260%
327 final N 01final 5.8800 30.416 2855.8 286.864 0.042% 0.245% 0.249% 0.002487 161739.59856 0.166660 0.000035 0.000413
328
329 N 02 R#1 N 02R#1 5.9000 30.647 2854.1 285.880 0.060% 0.240% 0.250%
330 R#2 N 02R#2 5.9000 30.637 2854.1 285.180 0.060% 0.250% 0.260%
331 final N 02final 5.9000 30.642 2854.1 285.529 0.042% 0.245% 0.249% 0.002487 161739.00241 0.166660 0.000035 0.000413
332
333 N 03 R#1 N 03R#1 5.9200 30.594 2853.6 284.510 0.060% 0.240% 0.250%
334 R#2 N 03R#2 5.9200 30.583 2853.6 283.690 0.060% 0.250% 0.260%
335 final N 03final 5.9200 30.589 2853.6 284.099 0.042% 0.245% 0.249% 0.002487 161737.60655 0.166658 0.000035 0.000413
336
337 N 04 R#1 N 04R#1 5.8700 30.487 2856.6 284.040 0.060% 0.240% 0.250%
338 R#2 N 04R#2 5.8700 30.485 2856.6 284.050 0.060% 0.250% 0.260%
339 final N 04final 5.8700 30.486 2856.6 284.045 0.042% 0.245% 0.249% 0.002486 161746.97898 0.166668 0.000035 0.000413
340
341 Average VNIIOFI Luminous Intensity relative standard uncertainty sum: 970475.45939 1.00000
342 VNIIOFI 0.009% 0.248% 0.248%
343 Final VNIIOFI average relative standard uncertainty:
344 u-uncorr u-corr uf
345 VNIIOFI 0.0087% 0.2477% 0.2479%
346
347
348 NMI: NIST
349
350 Lamp# Round# Data ID Lamp Electrical Lamp CCT NIST Lamp Luminous Intensity (cd) Calculations for NIST weighted mean
351 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
352 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
353 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
354
355 NIST20100 R#1 NIST20100R#1 5.822 30.270 2855.0 283.000 0.149% 0.200% 0.249%
356 R#2 NIST20100R#2 5.822 30.260 2853.0 282.600 0.110% 0.200% 0.228%
357 final NIST20100final 5.822 30.265 2854.0 282.742 0.089% 0.200% 0.219% 0.002188 208939.93383 0.175990 0.000078 0.000377
358
359 NIST20101 R#1 NIST20101R#1 5.918 30.600 2856.0 287.300 0.178% 0.200% 0.268%
360 R#2 NIST20101R#2 5.918 30.600 2855.0 287.500 0.163% 0.200% 0.258%
361 final NIST20101final 5.918 30.600 2855.5 287.409 0.120% 0.200% 0.233% 0.002333 183709.21591 0.154738 0.000093 0.000349
362
363 NIST20102 R#1 NIST20102R#1 5.905 30.440 2855.0 288.500 0.136% 0.200% 0.242%
364 R#2 NIST20102R#2 5.905 30.430 2854.0 288.300 0.202% 0.200% 0.285%
365 final NIST20102final 5.905 30.435 2854.5 288.438 0.113% 0.200% 0.230% 0.002297 189581.00531 0.159684 0.000090 0.000355
366
367 NIST20103 R#1 NIST20103R#1 5.877 30.510 2858.0 286.600 0.136% 0.200% 0.242%
368 R#2 NIST20103R#2 5.877 30.500 2856.0 285.900 0.142% 0.200% 0.245%
369 final NIST20103final 5.877 30.505 2857.0 286.265 0.098% 0.200% 0.223% 0.002228 201384.13264 0.169626 0.000083 0.000369
370
371 NIST20104 R#1 NIST20104R#1 5.683 30.700 2858.0 272.700 0.136% 0.200% 0.242%
372 R#2 NIST20104R#2 5.683 30.700 2857.0 272.200 0.136% 0.200% 0.242%
373 final NIST20104final 5.683 30.700 2857.5 272.450 0.096% 0.200% 0.222% 0.002219 203045.65317 0.171025 0.000082 0.000371
374
375 NIST20105 R#1 NIST20105R#1 5.922 30.590 2859.0 291.400 0.163% 0.200% 0.258%
376 R#2 NIST20105R#2 5.922 30.590 2857.0 290.800 0.125% 0.200% 0.236%
377 final NIST20105final 5.922 30.590 2858.0 291.023 0.099% 0.200% 0.223% 0.002233 200565.53657 0.168936 0.000084 0.000368
378
A B C D E F G H I J K L M N O P Q
379 Average NIST Luminous Intensity relative standard uncertainty sum: 1187225.47744 1.00000
380 NIST 0.021% 0.219% 0.220%
381 Final NIST average relative standard uncertainty:
382 u-uncorr u-corr uf
383 NIST 0.0209% 0.2188% 0.2198%
384
385
386 NMI: NRC
387
388 Lamp# Round# Data ID Lamp Electrical Lamp CCT NRC Lamp Luminous Intensity (cd) Calculations for NRC weighted mean
389 Current(A) Voltage(V) K I(cd) Relative Standard Uncertainty Weights Relative Uncertainties
390 random systematic final lamp (uf) uf 1/(uf)^2 wi uncorrelated correlated
391 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) fractional normalised for combining lamps
392
393 NRC021 final NRC021final 5.661 30.356 2800.0 254.400 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
394
395 NRC022 final NRC022final 5.620 30.069 2800.0 251.600 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
396
397 NRC023 final NRC023final 5.635 30.211 2800.0 254.000 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
398
399 NRC026 final NRC026final 5.650 30.398 2800.0 252.200 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
400
401 NRC027 final NRC027final 5.665 30.461 2800.0 254.600 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
402
403 NRC030 final NRC030final 5.633 30.106 2800.0 253.800 0.100% 0.605% 0.614% 0.006135 26568.73431 0.166667 0.000083 0.001019
404
405 Average NRC Luminous Intensity relative standard uncertainty sum: 159412.40587 1.00000
406 NRC 0.020% 0.611% 0.612%
407 Final NRC average relative standard uncertainty:
408 u-uncorr u-corr uf
409 NRC 0.0204% 0.6115% 0.6118%
A B C D E F G H I J K L M N O P Q R S T U
1 CCPR-K3.2014: Luminous Intensity
2 Draft B Report
3 2020-October-15
4 Appendix Dv2.1
5 Summary of Pilot Measurements of Participant Lamps
6
7
8 Lamp uncertainties are for combining
9 individual lamp measurements Lamp uncertainties are for combining
10 NMI: NMISA all NMISA lamp measurements
11
12 Lamp# Round# Data ID NMISA Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NMISA+Pilot weighted means
13 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights Relative Uncertainties
14 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) uncorrelated correlated combined uT uT 1/(uT)^2 wi uncorrelated correlated
15 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
16
17 “24” 4595 PTB 09 final “24” 4595 PTB 09final 268.850 0.110% 0.653% 0.663% 85.224989 0.000429 0.000122 0.000889 0.099% 0.663% 0.670% 0.57096 3.067497 0.248890932 0.000246 0.001649
18 “39” 4596 PTB 09 final “39” 4596 PTB 09final 284.150 0.110% 0.653% 0.663% 85.421999 0.000429 0.000122 0.000752 0.087% 0.663% 0.668% 0.57085 3.068722 0.248990335 0.000216 0.001650
19 “42” 4597 PTB 09 final “42” 4597 PTB 09final 275.839 0.110% 0.653% 0.663% 85.339987 0.000429 0.000122 0.000143 0.045% 0.663% 0.664% 0.56681 3.112634 0.252553309 0.000114 0.001674
20 NSI 10 final NSI 10final 315.788 0.110% 0.653% 0.663% 85.824098 0.000429 0.000122 0.000215 0.048% 0.663% 0.664% 0.57019 3.075809 0.249565424 0.000120 0.001654
21
22 NMISA Summary sum: 12.32466 1.00000
23 NMISA-weighted mean: R(i)= 85.452603
24 Uncertainties Final NMISA + Pilot relative standard uncertainty:
25 NMISA 0.0275% 0.6602% 0.6608% u-uncorr u-corr uf
26 NMISA + Pilot (u(Ri)) 0.0366% 0.6626% 0.6636% NMISA + Pilot 0.0366% 0.6626% 0.6636%
27 NMISA_transfer 0.0614%
28 Lamp uncertainties are for combining
29 NMI: NIM all NIM lamp measurements
30
31 Lamp# Round# Data ID NIM Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NIM+Pilot weighted means
32 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
33 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
34 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
35
36 NIM-01(Wi41/G-96) final NIM-01(Wi41/G-96)final 253.012 0.030% 0.167% 0.170% 86.340464 0.000429 0.000122 0.000887 0.099% 0.170% 0.197% 0.16966 34.74019 0.218689827 0.000215 0.000372
37 NIM-02(Wi41/G-152) final NIM-02(Wi41/G-152)final 263.660 0.028% 0.167% 0.169% 86.225685 0.000429 0.000122 0.002458 0.249% 0.170% 0.302% 0.26016 14.77449 0.093005545 0.000232 0.000158
38 NIM-03(Wi41/G-164) final NIM-03(Wi41/G-164)final 275.164 0.008% 0.167% 0.167% 86.202028 0.000429 0.000122 0.001285 0.136% 0.168% 0.216% 0.18582 28.96062 0.182307413 0.000247 0.000306
39 NIM-04(Wi41/G-180) final NIM-04(Wi41/G-180)final 265.251 0.018% 0.167% 0.168% 86.090246 0.000429 0.000122 0.000221 0.048% 0.168% 0.175% 0.15083 43.95691 0.276709193 0.000133 0.000466
40 NIM-05(Wi41/G-189) final NIM-05(Wi41/G-189)final 269.520 0.019% 0.167% 0.168% 86.490125 0.000429 0.000122 0.000803 0.091% 0.169% 0.192% 0.16569 36.42377 0.229288022 0.000209 0.000386
41
42 NIM Summary sum: 158.85598 1.00000
43 NIM-weighted mean: R(i)= 86.269629
44 Uncertainties Final NIM + Pilot relative standard uncertainty:
45 NIM 0.0049% 0.1681% 0.1681% u-uncorr u-corr uf
46 NIM + Pilot (u(Ri)) 0.0472% 0.1688% 0.1753% NIM + Pilot 0.0472% 0.1688% 0.1753%
47 NIM_transfer 0.0494%
48 Lamp uncertainties are for combining
49 NMI: NMIA all NMIA lamp measurements
50
51 Lamp# Round# Data ID NMIA Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NMIA+Pilot weighted means
52 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
53 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
54 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
55
56 S7 final S7final 298.735 0.006% 0.152% 0.152% 85.962034 0.000429 0.000122 0.001042 0.113% 0.152% 0.190% 0.16298 37.64564 0.16301544 0.000184 0.000249
57 350 LI3 final 350 LI3final 298.551 0.012% 0.152% 0.153% 86.046121 0.000429 0.000122 0.000256 0.050% 0.153% 0.161% 0.13859 52.06672 0.225462458 0.000113 0.000345
58 318 SI2 final 318 SI2final 305.829 0.014% 0.152% 0.152% 86.140581 0.000429 0.000122 0.000835 0.094% 0.153% 0.179% 0.15455 41.86359 0.181280261 0.000170 0.000277
59 306 S15 final 306 S15final 308.551 0.012% 0.152% 0.152% 86.125138 0.000429 0.000122 0.000505 0.066% 0.153% 0.167% 0.14350 48.55967 0.210276009 0.000139 0.000321
60 288 SI4 final 288 SI4final 301.555 0.036% 0.152% 0.156% 86.200848 0.000429 0.000122 0.000120 0.045% 0.157% 0.163% 0.14031 50.79737 0.219965832 0.000098 0.000344
61
62 NMIA Summary sum: 230.93300 1.00000
63 NMIA-weighted mean: R(i)= 86.100187
64 Uncertainties Final NMIA + Pilot relative standard uncertainty:
65 NMIA 0.0041% 0.1528% 0.1529% u-uncorr u-corr uf
66 NMIA + Pilot (u(Ri)) 0.0323% 0.1537% 0.1571% NMIA + Pilot 0.0323% 0.1537% 0.1571%
67 NMIA_transfer 0.0360%
68 Lamp uncertainties are for combining
69 NMI: NMIJ all NMIJ lamp measurements
70
71 Lamp# Round# Data ID NMIJ Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NMIJ+Pilot weighted means
72 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
73 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
74 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
75
76 #37 final #37final 242.150 0.064% 0.256% 0.264% 86.638944 0.000350 0.000122 0.001108 0.116% 0.264% 0.288% 0.24994 16.00753 0.176760272 0.000205 0.000467
77 #40 final #40final 250.395 0.064% 0.256% 0.264% 86.682269 0.000350 0.000122 0.000056 0.035% 0.264% 0.266% 0.23093 18.75147 0.207059735 0.000073 0.000547
CCPR-K3.2014: Luminous Intensity Draft B Page 9 of 15 Appendix D
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report, Appendices C,D,E,F Appendix_Dv2.1
A B C D E F G H I J K L M N O P Q R S T U
78 #51 final #51final 240.707 0.064% 0.256% 0.264% 86.701927 0.000350 0.000122 0.000303 0.046% 0.264% 0.268% 0.23242 18.51144 0.204409186 0.000095 0.000540
79 #52 final #52final 241.495 0.064% 0.256% 0.264% 86.668812 0.000350 0.000122 0.000297 0.046% 0.264% 0.268% 0.23227 18.53537 0.204673479 0.000094 0.000540
80 #58 final #58final 244.392 0.064% 0.256% 0.264% 86.601076 0.000350 0.000122 0.000123 0.037% 0.264% 0.267% 0.23091 18.75488 0.207097328 0.000077 0.000547
81
82 NMIJ Summary sum: 90.56070 1.00000
83 NMIJ-weighted mean: R(i)= 86.659060
84 Uncertainties Final NMIJ + Pilot relative standard uncertainty:
85 NMIJ 0.0142% 0.2618% 0.2622% u-uncorr u-corr uf
86 NMIJ + Pilot (u(Ri)) 0.0267% 0.2640% 0.2654% NMIJ + Pilot 0.0267% 0.2640% 0.2654%
87 NMIJ_transfer 0.0408%
88 Lamp uncertainties are for combining
89 NMI: IO-CSIC all IO-CSIC lamp measurements
90
91 Lamp# Round# Data ID IO-CSIC Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for IO-CSIC+Pilot weighted means
92 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
93 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
94 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
95
96 Wi95A final Wi95Afinal 278.294 0.005% 0.310% 0.310% 86.199291 0.000429 0.000122 0.000807 0.091% 0.310% 0.323% 0.27881 12.86417 0.195292059 0.000178 0.000606
97 Wi95B final Wi95Bfinal 285.188 0.003% 0.310% 0.310% 86.162482 0.000429 0.000122 0.000959 0.105% 0.310% 0.328% 0.28224 12.55373 0.190579163 0.000200 0.000591
98 Wi95C final Wi95Cfinal 286.637 0.002% 0.310% 0.310% 86.229941 0.000429 0.000122 0.000322 0.054% 0.310% 0.315% 0.27150 13.56655 0.205955007 0.000110 0.000639
99 Wi95D final Wi95Dfinal 271.222 0.003% 0.310% 0.310% 85.980405 0.000429 0.000122 0.000633 0.076% 0.310% 0.320% 0.27474 13.24842 0.201125343 0.000154 0.000624
100 A454 final A454final 433.899 0.009% 0.310% 0.310% 86.239028 0.000459 0.000122 0.000123 0.048% 0.310% 0.314% 0.27078 13.63858 0.207048428 0.000098 0.000643
101
102 IO-CSIC Summary sum: 65.87145 1.00000
103 IO-CSIC-weighted mean: R(i)= 86.162792
104 Uncertainties Final IO-CSIC + Pilot relative standard uncertainty:
105 IO-CSIC 0.0011% 0.3100% 0.3100% u-uncorr u-corr uf
106 IO-CSIC + Pilot (u(Ri)) 0.0343% 0.3103% 0.3122% IO-CSIC + Pilot 0.0343% 0.3103% 0.3122%
107 IO-CSIC_transfer 0.0365%
108 Lamp uncertainties are for combining
109 NMI: LNE-CNAM all LNE-CNAM lamp measurements
110
111 Lamp# Round# Data ID LNE-CNAM Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for LNE-CNAM+Pilot weighted means
112 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
113 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
114 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
115
116 926 final 926final 234.125 0.162% 0.220% 0.273% 87.151593 0.000350 0.000122 0.000890 0.096% 0.274% 0.290% 0.25258 15.67491 0.373165218 0.000357 0.001021
117 936 final 936final 241.568 0.180% 0.220% 0.284% 88.029666 0.000350 0.000122 0.002348 0.237% 0.285% 0.371% 0.32629 9.392788 0.223609635 0.000531 0.000636
118 A430 final A430final 397.346 0.162% 0.220% 0.273% 87.423860 0.000459 0.000122 0.000171 0.049% 0.274% 0.278% 0.24298 16.93759 0.403225148 0.000197 0.001103
119
120 Summary sum: 42.00529 1.00000
121 LNE-CNAM-weighted mean: R(i)= 87.457723
122 Uncertainties Final LNE-CNAM + Pilot relative standard uncertainty:
123 LNE-CNAM 0.0485% 0.2638% 0.2682% u-uncorr u-corr uf
124 LNE-CNAM + Pilot (u(Ri)) 0.0669% 0.2761% 0.2841% LNE-CNAM + Pilot 0.0669% 0.2761% 0.2841%
125 LNE-CNAM_transfer 0.0936%
126 Lamp uncertainties are for combining
127 NMI: METAS all METAS lamp measurements
128
129 Lamp# Round# Data ID METAS Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for METAS+Pilot weighted means
130 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
131 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
132 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
133
134 506 final 506final 276.211 0.027% 0.312% 0.314% 86.258517 0.000429 0.000122 0.000121 0.045% 0.314% 0.317% 0.27337 13.38156 0.169279687 0.000075 0.000531
135 684 final 684final 277.914 0.030% 0.312% 0.314% 86.112881 0.000429 0.000122 0.000012 0.043% 0.314% 0.317% 0.27298 13.42002 0.169766281 0.000073 0.000533
136 841 final 841final 280.587 0.027% 0.312% 0.314% 86.381991 0.000429 0.000122 0.000413 0.060% 0.314% 0.319% 0.27587 13.1396 0.166218887 0.000099 0.000522
137 1060 final 1060final 272.627 0.027% 0.312% 0.314% 86.190786 0.000429 0.000122 0.000602 0.074% 0.314% 0.322% 0.27786 12.9524 0.163850777 0.000121 0.000514
138 1063 final 1063final 284.197 0.027% 0.312% 0.314% 86.229931 0.000429 0.000122 0.000732 0.085% 0.314% 0.325% 0.28030 12.72808 0.161013003 0.000137 0.000505
139 1064 final 1064final 288.232 0.026% 0.312% 0.313% 86.068002 0.000429 0.000122 0.000163 0.046% 0.314% 0.317% 0.27289 13.42833 0.169871366 0.000078 0.000533
140
141 Summary sum: 79.04999 1.00000
142 METAS-weighted mean: R(i)= 86.206253
143 Uncertainties Final METAS + Pilot relative standard uncertainty:
144 METAS 0.0056% 0.3133% 0.3133% u-uncorr u-corr uf
145 METAS + Pilot (u(Ri)) 0.0245% 0.3138% 0.3148% METAS + Pilot 0.0245% 0.3138% 0.3148%
146 METAS_transfer 0.0301%
147 Lamp uncertainties are for combining
148 NMI: NPL all NPL lamp measurements
149
150 Lamp# Round# Data ID NPL Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NPL+Pilot weighted means
151 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
152 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
153 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
154
CCPR-K3.2014: Luminous Intensity Draft B Page 10 of 15 Appendix D
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report, Appendices C,D,E,F Appendix_Dv2.1
A B C D E F G H I J K L M N O P Q R S T U
155 A644 final A644final 451.875 0.058% 0.158% 0.168% 86.471175 0.000459 0.000122 0.000566 0.073% 0.169% 0.184% 0.15895 39.58029 0.198756516 0.000145 0.000335
156 A647 final A647final 459.530 0.058% 0.158% 0.168% 86.503670 0.000459 0.000122 0.000293 0.054% 0.169% 0.177% 0.15338 42.50903 0.213463522 0.000116 0.000360
157 PA758 final PA758final 460.515 0.058% 0.158% 0.168% 86.629919 0.000459 0.000122 0.000516 0.069% 0.169% 0.182% 0.15795 40.0844 0.201287989 0.000139 0.000340
158 877 final 877final 276.340 0.082% 0.158% 0.178% 86.585864 0.000429 0.000122 0.000376 0.057% 0.178% 0.187% 0.16219 38.01411 0.190891815 0.000109 0.000341
159 890 final 890final 273.930 0.082% 0.158% 0.178% 86.559634 0.000429 0.000122 0.000243 0.049% 0.178% 0.185% 0.16023 38.95173 0.195600158 0.000096 0.000349
160
161 NPL Summary sum: 199.13957 1.00000
162 NPL-weighted mean: R(i)= 86.549260
163 Uncertainties Final NPL + Pilot relative standard uncertainty:
164 NPL 0.0151% 0.1686% 0.1692% u-uncorr u-corr uf
165 NPL + Pilot (u(Ri)) 0.0274% 0.1725% 0.1746% NPL + Pilot 0.0274% 0.1725% 0.1746%
166 NPL_transfer 0.0431%
167 Lamp uncertainties are for combining
168 NMI: PTB all PTB lamp measurements
169
170 Lamp# Round# Data ID PTB Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for PTB+Pilot weighted means
171 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
172 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
173 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
174
175 759 final 759final 236.215 0.088% 0.130% 0.157% 86.234506 0.000429 0.000122 0.000609 0.075% 0.158% 0.174% 0.15030 44.26968 0.158817616 0.000118 0.000250
176 791 final 791final 247.540 0.085% 0.130% 0.155% 86.189513 0.000429 0.000122 0.000138 0.045% 0.156% 0.162% 0.13972 51.22315 0.183763219 0.000083 0.000286
177 793 final 793final 245.984 0.088% 0.130% 0.157% 86.260429 0.000429 0.000122 0.000818 0.092% 0.158% 0.183% 0.15755 40.28643 0.144527687 0.000134 0.000228
178 848 final 848final 228.535 0.085% 0.130% 0.155% 86.239953 0.000429 0.000122 0.000435 0.061% 0.156% 0.167% 0.14425 48.06017 0.172416019 0.000105 0.000268
179 851 final 851final 233.515 0.085% 0.130% 0.155% 86.233833 0.000429 0.000122 0.000637 0.077% 0.156% 0.174% 0.14972 44.61168 0.160044535 0.000123 0.000249
180 858 final 858final 225.069 0.088% 0.130% 0.157% 86.204822 0.000429 0.000122 0.000096 0.044% 0.158% 0.164% 0.14101 50.29429 0.180430924 0.000079 0.000284
181
182 Summary sum: 278.74541 1.00000
183 PTB-weighted mean: R(i)= 86.225460
184 Uncertainties Final PTB + Pilot relative standard uncertainty:
185 PTB 0.0177% 0.1500% 0.1511% u-uncorr u-corr uf
186 PTB + Pilot (u(Ri)) 0.0267% 0.1566% 0.1589% PTB + Pilot 0.0267% 0.1566% 0.1589%
187 PTB_transfer 0.0491%
188 Lamp uncertainties are for combining
189 NMI: VNIIOFI all VNIIOFI lamp measurements
190
191 Lamp# Round# Data ID VNIIOFI Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for VNIIOFI+Pilot weighted means
192 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
193 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
194 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
195
196 3281 final 3281final 274.265 0.042% 0.245% 0.249% 85.590866 0.000429 0.000122 0.001309 0.138% 0.249% 0.285% 0.24353 16.86199 0.15936868 0.000220 0.000397
197 3282 final 3282final 276.875 0.042% 0.245% 0.249% 85.734964 0.000429 0.000122 0.001173 0.125% 0.249% 0.279% 0.23880 17.53544 0.16573364 0.000207 0.000413
198 N 01 final N 01final 286.864 0.042% 0.245% 0.249% 85.588783 0.000429 0.000122 0.000448 0.062% 0.249% 0.257% 0.21959 20.73790 0.19600125 0.000122 0.000488
199 N 02 final N 02final 285.529 0.042% 0.245% 0.249% 85.709032 0.000429 0.000122 0.001196 0.127% 0.249% 0.279% 0.23955 17.42616 0.16470088 0.000209 0.000410
200 N 03 final N 03final 284.099 0.042% 0.245% 0.249% 85.502599 0.000429 0.000122 0.002160 0.220% 0.249% 0.332% 0.28421 12.37983 0.11700620 0.000258 0.000291
201 N 04 final N 04final 284.045 0.042% 0.245% 0.249% 85.591162 0.000429 0.000122 0.000402 0.059% 0.249% 0.256% 0.21893 20.86361 0.19718936 0.000116 0.000491
202
203 Summary sum: 105.80493 1.00000
204 VNIIOFI-weighted mean: R(i)= 85.623532
205 Uncertainties Final VNIIOFI + Pilot relative standard uncertainty:
206 VNIIOFI 0.0087% 0.2477% 0.2479% u-uncorr u-corr uf
207 VNIIOFI + Pilot (u(Ri)) 0.0479% 0.2489% 0.2535% VNIIOFI + Pilot 0.0479% 0.2489% 0.2535%
208 VNIIOFI_transfer 0.0531%
209 Lamp uncertainties are for combining
210 NMI: NIST all NIST lamp measurements
211
212 Lamp# Round# Data ID NIST Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NIST+Pilot weighted means
213 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
214 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
215 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
216
217 NIST20100 final NIST20100final 282.742 0.089% 0.200% 0.219% 86.302305 0.000429 0.000122 0.000969 0.106% 0.219% 0.243% 0.21005 22.66487 0.156983006 0.000166 0.000344
218 NIST20101 final NIST20101final 287.409 0.120% 0.200% 0.233% 86.265433 0.000429 0.000122 0.000848 0.095% 0.234% 0.252% 0.21756 21.12687 0.146330347 0.000139 0.000342
219 NIST20102 final NIST20102final 288.438 0.113% 0.200% 0.230% 86.368920 0.000429 0.000122 0.000123 0.045% 0.230% 0.234% 0.20235 24.42365 0.169164733 0.000075 0.000389
220 NIST20103 final NIST20103final 286.265 0.098% 0.200% 0.223% 86.304520 0.000429 0.000122 0.000280 0.051% 0.223% 0.229% 0.19761 25.60772 0.17736598 0.000091 0.000396
221 NIST20104 final NIST20104final 272.450 0.096% 0.200% 0.222% 86.344087 0.000429 0.000122 0.000096 0.044% 0.222% 0.227% 0.19562 26.13128 0.180992258 0.000080 0.000402
222 NIST20105 final NIST20105final 291.023 0.099% 0.200% 0.223% 86.240820 0.000429 0.000122 0.000566 0.071% 0.224% 0.235% 0.20235 24.42349 0.169163677 0.000120 0.000378
223
224 Summary sum: 144.37788 1.00000
225 NIST-weighted mean: R(i)= 86.305732
226 Uncertainties Final NIST + Pilot relative standard uncertainty:
227 NIST 0.0209% 0.2188% 0.2198% u-uncorr u-corr uf
228 NIST + Pilot (u(Ri)) 0.0286% 0.2251% 0.2269% NIST + Pilot 0.0286% 0.2251% 0.2269%
229 NIST_transfer 0.0563%
230 Lamp uncertainties are for combining
231 NMI: NRC all NRC lamp measurements
CCPR-K3.2014: Luminous Intensity Draft B Page 11 of 15 Appendix D
CCPR-K3.2014: Luminous Intensity 2022-May-20
Final Report, Appendices C,D,E,F Appendix_Dv2.1
A B C D E F G H I J K L M N O P Q R S T U
232
233 Lamp# Round# Data ID NRC Lamp Luminous Intensity Pilot Measurements Combined Uncertainty u(R(i,j) ) Calculations for NRC+Pilot weighted means
234 I(cd) Relative Standard Uncertainty R(i,j) Relative Standard Uncertainty Relative Standard Uncertainty Weights
235 random systematic final lamp (uf) <cd/V> u-uncorr u-corr u-uncorr(lamp) u-uncorr u-corr combined uT uT 1/(uT)^2 wi wi*u-uncorr wi*u-corr
236 u-uncorr u-corr SQRT(u-uncorr^2 + u-corr^2) cd/V normalised
237
238 NRC021 final NRC021final 254.400 0.100% 0.605% 0.614% 86.710956 0.000303 0.000122 0.000663 0.073% 0.614% 0.618% 0.53582 3.48310 0.165944 0.000121 0.001018
239 NRC022 final NRC022final 251.600 0.100% 0.605% 0.614% 86.432845 0.000303 0.000122 0.000604 0.068% 0.614% 0.617% 0.53358 3.51242 0.167341 0.000113 0.001027
240 NRC023 final NRC023final 254.000 0.100% 0.605% 0.614% 86.667211 0.000303 0.000122 0.000168 0.035% 0.614% 0.615% 0.53266 3.52457 0.167920 0.000058 0.001030
241 NRC026 final NRC026final 252.200 0.100% 0.605% 0.614% 86.838748 0.000303 0.000122 0.000788 0.084% 0.614% 0.619% 0.53789 3.45635 0.164669 0.000139 0.001010
242 NRC027 final NRC027final 254.600 0.100% 0.605% 0.614% 86.676055 0.000303 0.000122 0.000294 0.042% 0.614% 0.615% 0.53312 3.51843 0.167627 0.000071 0.001029
243 NRC030 final NRC030final 253.800 0.100% 0.605% 0.614% 86.872708 0.000303 0.000122 0.000413 0.051% 0.614% 0.616% 0.53492 3.49476 0.166499 0.000085 0.001022
244
245 Summary sum: 20.98962 1.00000
246 NRC-weighted mean: R(i)= 86.699196
247 Uncertainties Final NRC + Pilot relative standard uncertainty:
248 NRC 0.0204% 0.6115% 0.6118% u-uncorr u-corr uf
249 NRC + Pilot (u(Ri)) 0.0250% 0.6136% 0.6141% NRC + Pilot 0.0250% 0.6136% 0.6141%
250 NRC_transfer 0.0534%
Appendix_Ev2.1
A B C D E F G H I J K L M N O P Q R S T U V W X
1 CCPR-K3.2014: Luminous Intensity
2 Draft B Report
3 2020-October-15
4 Appendix Ev2.1
5 Calculation of the KCRV and the Unilateral DoE
6 0.0006163 Draft Av1.0 value
7 0.0003400 Draft Av2.0 value
8 Mandel-Paule adjustment s: 0.0003100 (relative standard uncertainty)
9
10 Unilateral DoE
11 Relative Standard Uncertainties Uncertainty u(c,t,s) KCRV uncertainty calculation Outlier Test Chi-square Unilateral DoE ui(Di) for k=1
12 NMI R(i) u(NMI) uc(NMI) Pilot Transfer cutoff NMI+transfer+s KCRV weights wi u(R(i))=u(NMI)+u(t) wi*u(R(i)) R(i)-KCRV (R(i)-KCRV)/u(NMI) (R(i)-KCRV)^2/u(c,t,s)^2 Di correlation term ui(Di) k=1
13 cd/V no cutoff with cutoff u(t) relative standard cd/V 1/u(c,t,s)^2 normalised relative Standard cd/V cd/V cd/V 16.906950 (Ri-KCRV)/KCRV cd/V ui(Di)/KCRV Di/ui
14
15
16 1 NMIJ 86.659060 0.002622 0.002622 0.000408 0.002672 0.231546 18.652039 0.071599 0.002654 0.229982 0.016467 0.403221 1.77 3.032584 0.004675 0.007574 0.220850 0.002560 1.8258
17 2 NIM 86.269629 0.001681 0.001722 0.000494 0.001818 0.156878 40.632629 0.155976 0.001753 0.151194 0.023583 0.013790 0.10 0.007727 0.000160 0.007131 0.138511 0.001606 0.0996
18 3 PTB 86.225460 0.001511 0.001722 0.000491 0.001818 0.156726 40.711690 0.156280 0.001589 0.136982 0.021408 -0.030379 -0.23 0.037572 -0.000352 0.005865 0.127891 0.001483 -0.2375
19 4 METAS 86.206253 0.003133 0.003133 0.000301 0.003163 0.272672 13.449899 0.051630 0.003148 0.271359 0.014010 -0.049586 -0.18 0.033070 -0.000575 0.007604 0.263608 0.003056 -0.1881
20 5 NIST 86.305732 0.002198 0.002198 0.000563 0.002290 0.197679 25.590523 0.098234 0.002269 0.195860 0.019240 0.049893 0.26 0.063704 0.000578 0.007537 0.185152 0.002147 0.2695
21 6 NMIA 86.100187 0.001529 0.001722 0.000360 0.001787 0.153834 42.256726 0.162211 0.001571 0.135221 0.021934 -0.155652 -1.18 1.023773 -0.001805 0.005932 0.125736 0.001458 -1.2379
22 7 VNIIOFI 85.623532 0.002479 0.002479 0.000531 0.002554 0.218685 20.910394 0.080269 0.002535 0.217068 0.017424 -0.632307 -2.98 8.360223 -0.007331 0.007564 0.207391 0.002404 -3.0489
23 8 IO-CSIC 86.162792 0.003100 0.003100 0.000365 0.003137 0.270298 13.687225 0.052541 0.003122 0.268975 0.014132 -0.093046 -0.35 0.118499 -0.001079 0.007602 0.261155 0.003028 -0.3563
24 9 NPL 86.549260 0.001692 0.001722 0.000431 0.001802 0.155991 41.095974 0.157755 0.001746 0.151156 0.023846 0.293421 2.00 3.538205 0.003402 0.007209 0.138189 0.001602 2.1233
25 10 NMISA 85.452603 0.006608 0.006608 0.000614 0.006644 0.567723 0.000000 0.000000 0.006636 0.567105 0.000000 -0.803236 -1.42 0.000000 -0.009312 0.000000 0.570144 0.006610 -1.4088
26 11 NRC 86.699196 0.006118 0.006118 0.000534 0.006149 0.533124 3.518389 0.013506 0.006141 0.532446 0.007191 0.443357 0.84 0.691594 0.005140 0.007658 0.528486 0.006127 0.8389
27
28 median NMI uncertainty: 0.002339 SUM(wi): 260.505487 1.000000 u(KCRV): 0.058794
29 cutoff uncertainty: 0.001722
30
31 KCRV: 86.255839 cd/V
32 u(KCRV): 0.058794 cd/V
33 u(KCRV): 0.000682 (relative standard uncertainty)
34
35 for plot text box: Mandel-Paule adjustment s: 0.00031 (relative standard uncertainty)
36 u(KCRV): 0.00068 (relative standard uncertainty)
37
38 for plot: 0.0 0.000682 -0.000682
39 12.0 0.000682 -0.000682
40 0.000682 -0.000682
41 0.000682 -0.000682
42 0.000682 -0.000682
43 0.000682 -0.000682
44 0.000682 -0.000682
45 0.000682 -0.000682
46 0.000682 -0.000682
47 0.000682 -0.000682
48 0.000682 -0.000682
49 0.000682 -0.000682
50 0.000682 -0.000682
Appendix_Fv2.1
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z AA
1 CCPR-K3.2014: Luminous Intensity
2 Draft B Report
3 2020-October-15
4 Appendix Fv2.1
5 Calculation of the Bilateral DoE (uij, k=1)
6 ui and uij are k=1 uncertainties
7
8
9 j: IO-CSIC METAS NIM NIST NMIA NMIJ NMISA NPL NRC PTB VNIIOFI
10 i NMI Di ui Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij Dij uij
11 1 IO-CSIC -0.001079 0.003028 -0.000504 0.004433 -0.001239 0.003580 -0.001657 0.003859 0.000726 0.003494 -0.005753 0.004097 0.008234 0.007334 -0.004480 0.003577 -0.006219 0.006889 -0.000727 0.003503 0.006252 0.004021
12 2 METAS -0.000575 0.003056 0.000504 0.004433 -0.000735 0.003603 -0.001153 0.003881 0.001230 0.003518 -0.005250 0.004117 0.008737 0.007345 -0.003977 0.003600 -0.005715 0.006901 -0.000223 0.003526 0.006756 0.004042
13 3 NIM 0.000160 0.001606 0.001239 0.003580 0.000735 0.003603 -0.000419 0.002867 0.001964 0.002353 -0.004515 0.003180 0.009472 0.006864 -0.003242 0.002474 -0.004980 0.006386 0.000512 0.002365 0.007490 0.003082
14 4 NIST 0.000578 0.002147 0.001657 0.003859 0.001153 0.003881 0.000419 0.002867 0.002383 0.002760 -0.004096 0.003492 0.009891 0.007014 -0.002823 0.002864 -0.004562 0.006547 0.000931 0.002770 0.007909 0.003403
15 5 NMIA -0.001805 0.001458 -0.000726 0.003494 -0.001230 0.003518 -0.001964 0.002353 -0.002383 0.002760 -0.006479 0.003084 0.007508 0.006820 -0.005206 0.002349 -0.006945 0.006339 -0.001452 0.002234 0.005526 0.002982
16 6 NMIJ 0.004675 0.002560 0.005753 0.004097 0.005250 0.004117 0.004515 0.003180 0.004096 0.003492 0.006479 0.003084 0.013987 0.007147 0.001273 0.003177 -0.000465 0.006690 0.005027 0.003093 0.012005 0.003670
17 7 NMISA -0.009312 0.006610 -0.008234 0.007334 -0.008737 0.007345 -0.009472 0.006864 -0.009891 0.007014 -0.007508 0.006820 -0.013987 0.007147 -0.012714 0.006862 -0.014452 0.009042 -0.008960 0.006824 -0.001982 0.007104
18 8 NPL 0.003402 0.001602 0.004480 0.003577 0.003977 0.003600 0.003242 0.002474 0.002823 0.002864 0.005206 0.002349 -0.001273 0.003177 0.012714 0.006862 -0.001738 0.006385 0.003754 0.002361 0.010732 0.003078
19 9 NRC 0.005140 0.006127 0.006219 0.006889 0.005715 0.006901 0.004980 0.006386 0.004562 0.006547 0.006945 0.006339 0.000465 0.006690 0.014452 0.009042 0.001738 0.006385 0.005492 0.006343 0.012471 0.006644
20 10 PTB -0.000352 0.001483 0.000727 0.003503 0.000223 0.003526 -0.000512 0.002365 -0.000931 0.002770 0.001452 0.002234 -0.005027 0.003093 0.008960 0.006824 -0.003754 0.002361 -0.005492 0.006343 0.006978 0.002992
21 11 VNIIOFI -0.007331 0.002404 -0.006252 0.004021 -0.006756 0.004042 -0.007490 0.003082 -0.007909 0.003403 -0.005526 0.002982 -0.012005 0.003670 0.001982 0.007104 -0.010732 0.003078 -0.012471 0.006644 -0.006978 0.002992
DoE
0.015 +u(KCRV)
-u(KCRV)
0.010
0.005
DOE
0.000
-0.005
-0.010
-0.020
NMIJ NIM PTB METAS NIST NMIA VNIIOFI IO-CSIC NPL NMISA NRC