Link This |
Email this |
Blog This |
Comments (0)
Metrologists gather for NCSLI regional meeting
February 1, 2008
On January 29, I attended a meeting of the
NCSLI New England region, held at the
Agilent office in
Andover, MA. Over fifty engineers and metrologists packed the conference room to hear three presentations on measurement.
Kevin Bull of Veriteq Instruments opened with a presentation on temperature and humidity measurement, with focus on sensors. He gave an overview of thermocouples, RTDs, and thermistors, explaining how they work, their relative accuracy, and typical applications.
Hats off to Kevin for correctly describing how thermocouples work. Too often, people incorrectly say that thermocouples create a voltage at the junction of two dissimilar metals. Bull correctly stated that the voltage occurs across a wire when its two ends are at different temperatures and that the voltage depends on the wire material, thus two wires of different materials will produce different voltages and the difference is related to temperature.
Bull spent the bulk of his presentation on humidity measurement, starting with Da Vinci’s hygrometer. Humidity measurement and control is important, particularly in calibration labs where small changes in humidity can affect the quality of a measurement. He explained why humidity measurements are so hard to make and why they are far less accurate than other parameters.
“Part of the problem,” he said, “is that humidity measurements are temperature dependent.” He discussed several methods of measuring humidity, with focus on capacitive RH sensors, the most common type. “Self heating can cause RH errors. A 0.3oC change in temperature can cause a 1% error.” Also, RH sensors must be in direct contact with the air, unlike temperature sensors that can work through materials.
Following Bull’s presentation, Bob Brown, chief metrologist at Agilent, gave a presentation on managing risk as it relates to calibration (Brown flew in from Loveland, CO for this meeting).
Brown explained how his lab, which houses “The Agilent volt,” a Josephson-junction array, is on par with the U.S. national volt at NIST. Agilent, NIST, and other labs engage in a “round robin” where they directly compare each other’s volt to verify their uncertainties. He also said that the metrologists in Loveland send a voltage reference to the company’s facility in Japan for comparison so that all calibration labs are working from the same reference voltage.
He told the audience about the hidden costs of setting your test parameters incorrectly or of not having equipment within specification. False failures result in good products being rejected. False passes can result in product recalls.
Risks occur when measurement change because of environmental conditions such as temperature and time. “You must remove the offsets,” said Brown. “But how to you know if the offsets change? That’s why it’s important to control the temperature in calibration labs and to characterize equipment so you know how it drifts over time. He showed how measurement uncertainties add with each step away from a reference value and how test-uncertainly ratios affect the quality of your measurements.
After lunch, the audience listened to a presentation from Dave Dwyer of Fluke about infrared temperature measurements. Dwyer shoed the audience how digital thermography is used in applications that range from animal diagnosis to predictive maintenance of electrical connections and mechanical devices. Electrical connections, for example, with increase in temperature as their resistance increases. That temperature rise can indicate impending failure.
Dwyer discussed emissivity and why infrared thermometers don’t work on some materials. The also explained how black-body sources are used to calibrate infrared thermometers.
Posted by Martin Rowe on February 1, 2008 | Comments (0)
