Extreme-range pressure sensor boasts 1% total error

Cresskill, NJ--Measuring crushing pressures at hellish temperatures spells trouble for engineers. If transducers can be had to survive those conditions, the necessary corrections for temperature, hysteresis, drift, etc. vastly complicate system-controls design.

The "twin-column technology" pressure transducer can change all that. Simple in concept, the sensor consists of a horizontal base with two inlet ports connected to identical vertical tubes topped by independent housings for fiber-optic emitters and detectors. Electronics include a source of collimated light for the emitter and photodiodes, and signal conditioning for the detector (see the accompanying diagram).

Capable of measuring gage, absolute, or differential pressure, the transducer relies on the relative change in length between the tubes as pressure is applied to one or both of them, regardless of temperature. As one tube lengthens, the light incident on the two photodetectors changes in proportion. Amplified photodiode output, for now an analog voltage or current, records the change in pressure.

Set for introduction this quarter, the Model DPT will record pressure to 5 kpsi. It measures 1 x 2 x 4 inches and weighs just 14 oz. Over its operating temperature range of 40F to 160F, it exhibits a total combined accuracy of 1% full scale.

"Our goal was to eliminate the confusion concerning accuracy," explains Charles Liucci, president of LEL Corp. and inventor of the transducer. With over forty years experience working with and designing sensors, Liucci had grown frustrated with compensation tables and "best-fit" calibration curves. "With this design, no matter how you do your error analy-sis, worst-case error is n0.5%," he says. "And it's as linear as a taut string."

Although the concept is simple, the devil lies in the details. Working with engineers Don Ball and Steve Voto, Liucci spent four years developing the manufacturing processes necessary to make the sensor easy to assemble within the required tolerances.

Those tolerances are quite small. Although the exact figures are proprietary, Liucci says the tubes are parallel to one another and perpendicular to the base to within microinches. Similarly, the alignment between fiber-optic source and detectors is as precise as that attained on optical benches. Yet the design is rugged, capable of withstanding 50G shocks and 100G accelerations.

Features that minimize error start with the twin tubes. They're machined from the same piece of Ni-span C alloy 902 and cut to the same length to within similar tolerances used elsewhere in the design. With a coefficient of thermal expansion of 9 x 10⁷ inch/inch/C, the material's dimensional changes with temperature are small to begin with, but the symmetrical design minimizes any relative changes in tube length. Because the alloy is chemically inert, the transducer's mechanical performance won't change over time even if only one tube is exposed to wet or dry corrosive media.

The common-mode measuring technique also reduces error. Output from the twin photodetectors goes to a bridge circuit. If the LED source changes with time or supply voltage, the transducer output depends on the relative change in photodiode output, not the absolute brightness of the light they receive. Still, for safety, a feedback detector on the source side monitors the LED, signaling if its output falls beyond prescribed limits.

Additional details...Contact Charles Liucci, LEL Corp., 5 Burns Pl., Cress-kill, NJ 07626, (201) 569-8641.