Calibration is changing

Two significant issues have recently emerged, one driven by a document, the other by economics.

By Martin Rowe, Senior Technical Editor -- Test & Measurement World, 6/1/2010 12:00:00 AM

E-mail

The calibration and metrology profession hardly seems to change. In this business, engineers treat the latest measuring instrument with suspicion until it has a proven, documented track record. Yet, changes do take place, although perhaps more slowly than in other fields. Two significant issues have recently emerged, one driven by a document, the other by economics.

ANSI/NCSL Z540.3 section 5.3 (Ref. 1) addresses the risk of false acceptance. The probability of accepting an out-of-tolerance part can't exceed 2% (figure). Arriving at that number involves knowing the uncertainty of the measurement, which is directly tied to TUR (test uncertainty ratio).

Howard Zion, director of technical operation at Transcat, advises clients on how PFA (probability of false acceptance) affects their manufacturing procedures. "The conversation takes anywhere from 30 minutes to four hours," he said. "Engineers outside of the calibration community are starting to understand the importance of measurement uncertainty."

Compliance to ANSI Z.540.3 requires that the probability of false passing is less than 2%.

Ultimately, it's up to you to decide how much risk is acceptable. The amount of acceptable risk depends on the consequences of passing bad parts. For example, if a measurement should fall at its tolerance limit, then the probability that the measurement is out of tolerance is 50%. Depending on the part, you may need to add guardbands around your measurements to increase the probability that the actual value falls within tolerance. In some instances, a 2% PFA might be too high. (View links to papers that describe PFA and how to determine risk.)

The second shift has occurred as companies have moved calibration of measurement equipment closer to manufacturing. Measurement consultant Charles Motzko noted that the practice started at TRW (now Northrop Grumman) before being adopted by other aerospace companies and eventually by other industries. Moving calibration closer to manufacturing may mean less downtime, because the equipment is calibrated in place rather than returning to the cal lab.

This shift has changed the way companies view metrology, which has traditionally been treated as a separate technical function. "Moving calibration to the manufacturing floor means you must teach engineers and technicians how to measure," said Motzko. "Measurements can no longer be a specialty."

In addition, the use of automated calibration tools means that calibration may be performed by less skilled people than in the past. Even so, there's a concern that industry won't have enough people trained in calibration and metrology. "The skill set is down," said Steve Griffin, president of Workload Tools and director of the 2010 NCSLI Workshop and Symposium. The problem will likely increase as engineers with measurement expertise retire—they'll take their knowledge with them.

To address the problem, NCSL International has a program to pique interest in calibration and metrology. The NCSLI Committee on Education Liaison and Outreach works with colleges that offer programs in measurement. NCSLI has also distributed a DVD called "Find a Cool Career in Metrology" to its members. You can learn more about the program at www.ncsli.org. Click on "Learning and development."

Reference
1. ANSI/NCSL Z540.3-2006, "Requirements for the Calibration of Measuring and Test Equipment," NCSL International. www.ncsli.org.

More information about FPGA and how to determine risk

Deaver, David, and Jack Somppi, "A Study of and Recommentations for Applying the False Acceptance Risk Specification OF Z540.3," Measurement Science Conference 2010. www.msc-conf.com.

Castrup, Howard, "Risk Analysis Methods for Complying with Z540.3," 2007 NCSLI Workshop and Symposium. isgmax.com/Articles_Papers/Risk%20Analysis%20Methods%20for%20Z540.pdf.

Castrup, Howard, "An Examination of Measurement Decision Risk and Other Measurement Quality Metrics," 2009 NCSLI Workshop and Symposium. isgmax.com/Articles_Papers/Examination%20of%20Risks%20and%20Other%20Metrics.pdf.

Zion, Howard, "Metrology Concepts: k=3.9? . . . Why Not??," 2007 Measurement Science Conference. www.transcat.com/PDF/MetrologyConceptOOT.pdf.

Learn phase noise and harmonics "Comparing Spectral Purity in Microwave Signal Generators" from Giga-tronics give you an overview of how phase noise and harmonics add distortion to all microwave signals. The paper explains how to interpret these specifications. www.gigatronics.com/downloads/appnotes/AN-GT115A-Signal_Purity_in_Microwave_Signal_Generators.pdf. LTE test resources available online The Agilent Technologies Website offers a series of app notes, posters, Webcast CDs, and brochures on LTE testing. Topics include design and test fundamentals, stimulus response testing, and MIMO test. www.agilent.com/find/LTE-forward. CD explains measurements Keithley Instruments' "A Guide to Understanding Electrical Test and Measurement" CD contains app notes, white papers, and Webcasts about instruments such as multimeters, source-measure units, and waveform generators. www.ggcomm.com/Keithley/Mar10_NewsRelease_GeneralMeasCD.html.