An example of TEDS-enabled thermocouples is the Watlow Infosense line. Storing sensorspecific linearity and compensation data in an internal EEPROM is said to improve accuracy by a factor of 10 compared to that of RTDs at 600°C. Initial accuracy is also said to be three times that of Type-K thermocouples. The sensors automatically communicate calibration, identification, and traceability data back to measurement instrumentation.

An example of TEDS-enabled thermocouples is the Watlow Infosense line. Storing sensorspecific linearity and compensation data in an internal EEPROM is said to improve accuracy by a factor of 10 compared to that of RTDs at 600°C. Initial accuracy is also said to be three times that of Type-K thermocouples. The sensors automatically communicate calibration, identification, and traceability data back to measurement instrumentation.

The IEEE 1451.4 standard for smart transducer interfaces basically covers how individual sensors can store calibration and other information in an onboard memory, for use by electronic instrumentation.

Finalization of the TEDS spec could lead to, among other things, better sensors for gauging superhigh temperatures. An example of the trend comes from Watlow Electric Mfg. Co., in Richmond, Ill. (www.watlow.com). The firm now produces TEDS thermocouples called Watcouples that target such uses as heattreating furnaces and semiconductor fabrication processes. The claim to fame of these devices is much less drift and much longer life than for ordinary Type-K thermocouples, says Watlow. The accuracy and long life come from use of nonstandard thermocouple material that would have been impractical without the onboard correction made possible by TEDS.

Watlow's approach, likely to be mimicked by other sensor makers, is to substitute digital electronics for the pure sensor materials necessary in the past to get a linear relationship between temperature and output EMF. Materials used for thermocouples and resistive temperature detectors (RTDs) have historically been chosen for good linearity. But linearity has come at the expense of other properties that include accuracy and life.

The typical way of producing a sensor with linear properties is to add relatively small concentrations of metallic elements in the thermocouple metal. But relying on miniscule amounts of minor elements has a down side. The resulting sensors are susceptible to drift and other measurement errors caused by inhomogeneity.

Thermocouples comprised of noble metals such as gold, platinum, and palladium are much less prone to drift. But they tend to be expensive. Copper and copper-nickel thermocouples are another possibility but can contaminate some high-temperature processes as are common in semiconductor fabrication.

New high-temperature thermocouples avoid the contamination problem through the use of materials such as nickelsilicon and nickel-chromium alloy 600. The EMF-to-temperature relationships for these devices can be complicated, perhaps specified by a polynomial. But the polynomial coefficients can be easily encoded into TEDS memory on-board the sensor. This lets instrumentation make the necessary corrections to obtain accurate temperature readings. And the resulting information has more precision than available from ASTM conversion tables describing families of sensors.