In industrial applications, good repeatability often is more important then absolute accuracy. If process pressures vary over a wide range, pressure sensors with good linearity and low hysteresis are the preferred choice.

Ambient and process temperature variations also cause errors in pressure measurements, particularly in detecting low pressures and small differential pressures. In such applications, temperature compensators must be used.

Power supply variations also lower the performance of pressure transducers. The sensitivity of a transducer determines the amount of change that occurs in the output voltage when the supply voltage changes, with the measured pressure and the rated pressure of the transducer remaining constant.

Pressure transducers usually generate output signals in the millivolt range (spans of 100 mV to 250 mV). When used in transmitters, these are often amplified to the voltage level (1 to 5 V) and converted to current loops, usually 4-20 mA dc.

The transducer housing should be selected to meet both the electrical area classification and the corrosion requirements of the particular installation. Corrosion protection must take into account both splashing of corrosive liquids or exposure to corrosive gases on the outside of the housing, as well as exposure of the sensing element to corrosive process materials. The corrosion requirements of the installation are met by selecting corrosion-resistant materials, coatings, and by the use of chemical seals.

If the installation is in an area where explosive vapors may be present, the pressure sensor and its power supply must be suitable for these environments. This is usually achieved either by placing them inside purged or explosion-proof housings, or by using intrinsically safe designs.

Probably the single most important decision in selecting a pressure transducer is the range. One must keep in mind two conflicting considerations: the instrument's accuracy and its protection from overpressure. From an accuracy point of view, the range of a transmitter should be low (normal operating pressure at around the middle of the range), so that error, usually a percentage of full scale, is minimized. On the other hand, one must always consider the consequences of overpressure damage due to operating errors, faulty design (waterhammer), or failure to isolate the instrument during pressure-testing and start-up. Therefore, it is important to specify not only the required range, but also the amount of overpressure protection needed.

Most pressure instruments are provided with overpressure protection of 150% to 400% of range. These protectors satisfy the majority of applications. Where higher overpressures are expected and their nature is temporary (pressure spikes of short duration--seconds or less), snubbers can be installed. These filter out spikes, but cause the measurement to be less responsive. If excessive overpressure is expected to be of longer duration, one can protect the sensor by installing a pressure relief valve. However, this will result in a loss of measurement when the relief valve is open.

If the transmitter is to operate under high ambient temperatures, the housing can be cooled electrically (Peltier effect) or by water, or it can be relocated in an air-conditioned area. When freezing temperatures are expected, resistance heating or steam tracing should be used in combination with thermal insulation.

When high process temperatures are present, one can consider the use of various methods of isolating the pressure instrument from the process. These include loop seals, siphons, chemical seals with capillary tubing for remote mounting, and purging.