Text preview for : tpoint.pdf part of Fluke tpoint Fluke 5700a tpoint.pdf
Back to : tpoint.pdf | Home
Updated Evaluation of Calibration Test Point Selection
for Fluke 57XX products
Speaker: Jeff Gust
Fluke Calibration
6920 Seaway Blvd
Everett, WA 98203
Chief Corporate Metrologist
Fluke Corporation
(425) 446-5471
[email protected]
Authors: Jeff Gust, Steven Haynes, Bill Britz, Randy Lemon, Neil Faulkner, Greg Tarolli,
Fluke Calibration
1.0 Abstract
The first Fluke 5700 series calibrator was manufactured in the late 1980's. The metrology and
associated test point selection for calibrating the product was rigorous by any standards, then or
now. However, the calibration procedure that has been in place essentially since its initial
release pre-dates significant advances in hardware, such as the Fluke Calibration 792A and
5790A, as well as ISO 17025, and the popularization of voluntary calibration laboratory
accreditation.
Fluke Calibration has re-visited the selection of test points in the calibration procedure to ensure
that the result of any 5700 series calibrator would clearly demonstrate that all specifications have
been verified, and that traceability for all functions is ensured. This was accomplished by an
event that brought together representatives from Design Engineering, Test Engineering, Service
Engineering, Service Metrology and Corporate Metrology. The group evaluated each
specification, considered the internal architecture of the product, and reviewed and improved
upon the test points. This paper presents the test points for the 5700 series calibrators and the
technical information as to why they were chosen.
2.0 Learning objectives
The attendee will understand why the test points selected for the Fluke 5700 series calibrators
verify all product specifications and ensure metrological traceability for all functions of the
product.
2013 NCSL International Workshop and Symposium
3.0 Introduction
The Fluke 5700 family of multifunction calibrators is one of the cornerstone reference standards
for electrical calibration laboratories around the world. This product is used in a wide array of
applications and establishments, from National Measurement Institutes to field calibration
laboratories; from factory production lines to onboard Navy ships. While some improvements
have been made in performance and internal circuit design over the years, much of the basic
design is the same as its original form. The 5700 series calibrators remain the world's most
accurate multifunction calibrator.
For Fluke to ensure that each manufactured product meets all published specifications, a
thorough metrology evaluation occurs during the design process. However, for the 5700 series,
the evaluation took place in late 1980's. There have been significant developments in quality
standards associated with metrology since then. The design and initial manufacture of the 5700
series predates ANSI/NCSL Z540-1-1994, ISO 17025, the European co-operation for
accreditation, the ILAC Mutual Recognition Arrangement, and the CIPM Mutual Recognition
Arrangement.
Over the years, accreditation bodies have questioned the test point selection associated with the
calibration of the 5700 series product. While Fluke has always satisfied technical inquiries
associated with our test point selection, we felt that it was time to revisit this issue. The first
reason is that some of original designers of the product are nearing retirement age, and
improving documentation associated with the calibration process is one of our knowledge
capture functions. The second reason is to provide a succinct document for both customers and
accreditation bodies to explain Fluke's test point selection strategy, so that this information can
be shared and presented for peer review. Lastly, Fluke has a strong commitment to continuous
improvement, and since the world of metrology has changed since the introduction of this
product, we felt that it was important to apply our own critical eye on the original test point
selection with regards to improvement opportunities.
Satisfactory selection of test points for the calibration of any test instrument requires intimate
knowledge of the product's design and a sound background in metrology. It also helps to have
an understanding of relevant quality standards. In June 2012, Fluke applied one of our lean
manufacturing management tools, a Kaizen event, to help us evaluate and select the 5700 series
product specifications. The Kaizen event brought together associates from Design Engineering,
Test Engineering, Service Engineering, Corporate Metrology and Service Metrology to evaluate
the present test points associated with calibration, to ensure that they adequately verify product
specifications. One of the principles of lean manufacturing is to eliminate tasks that do not add
value to the process. The team's goal was to ensure that the appropriate test points needed to
2013 NCSL International Workshop and Symposium
validate the product were selected, but not to add unnecessary test points. These are the test
points that are reported on the certificate of calibration, whether the calibration was produced
during its initial manufacture, or if the product is returned for recalibration. It is important to
note that there are many additional test points that the factory checks during the initial
manufacturing of the instrument. These are one-time tests to ensure functionality of new circuit
boards and are not required for routine recalibration. The results of the Kaizen for the calibration
are presented below. This set of test points can be used to determine performance of the
calibrator against the 24-hour, 90-days, 180-days or one-year specifications, or for the 99 percent
or 95 percent level of confidence, because the formatting of these specifications is identical; the
only difference is the test limits for each specification.
4.0 Discussion
4.1 DC voltage function
It is critical to perform thorough testing of the dc voltage function as it is interrelated to other
outputs of the instrument. Table 1 presents the test points that are required to verify the product
specifications of the 5700 series calibrators for dc voltage.
The instrument is tested at the minimum output at all ranges for compliance to specification. For
negative dc voltage values, a relay is employed to reverse the polarity of the voltage generation
circuitry, so Fluke also tests the "negative zero" output, which is the minimum negative voltage
value that can be sourced to ensure that the zero output on the negative range also performs to
the product specifications. When a zero output is selected, the default position of the relay is in
the positive voltage position. Note that the 1100 volt range has a minimum output of 100 volts
because the high voltage range is sourced by rectifying the output of the high voltage step up
transformer and is controlled by the 11 volt range of the digital-to-analog Converter (DAC)
assembly, so testing at zero volts for this range is not possible.
The linearity of the DAC is verified on the 22 volt range where the instrument is tested from zero
to positive and negative 19 volts. The same DAC circuitry is used for all voltage output
functions, so a test of the linearity of this range is all that is required to test the linearity for the
rest of the dc voltage output functions, making it only necessary to test a reduced set of points on
the other ranges. Due to the very accurate specifications of the 5700 series calibrators, Fluke
tests the rest of the ranges at outputs that are multiples of 10, as most calibration laboratories
achieve their best measurement uncertainties at these points. Even though this output is only
approximately 50 percent of the range output, since the linearity has been verified previously,
this satisfactorily validates the product specification.
2013 NCSL International Workshop and Symposium
Range Test Points
220 mV 0 mV
-10 nV
100 mV
-100mV
2.2 V 0V
-100 nV
1V
-1 V
11 V 0V
-1