Pressure Sensors

What is a Pressure Sensor

A pressure sensor is an industrial measurement element that converts the mechanical pressure of a fluid (liquid or gas) acting on a sensing diaphragm into an electrical signal. It is one of the most common field instruments in process automation, alongside temperature, flow, and level as the four fundamental process variables. A qualified pressure sensor must maintain linear, repeatable, and traceable output under varying temperature, vibration, and media adhesion conditions throughout its service life, which fundamentally distinguishes it from an ordinary pressure gauge.

Structurally, a pressure sensor consists of three parts: (1) the pressure isolation unit, with a wetted surface made of stainless steel, Hastelloy, titanium, or ceramic diaphragm; (2) the sensing element, which converts the diaphragm's micro-deformation into an electrical signal via strain gauge, capacitive, piezoresistive, piezoelectric, or resonant principles; (3) the signal conditioning circuit, which performs amplification, temperature compensation, linearization, and digital communication. When the latter two are integrated and output standardized 4-20 mA / HART / Foundation Fieldbus signals, the industry calls it a "pressure transmitter," the term that appears on most factory purchase orders.

The industrial history of pressure measurement dates back to 1849, when French engineer Eugene Bourdon invented the Bourdon tube pressure gauge. In the 1930s, the strain gauge made pressure-to-electrical conversion possible. In the 1970s, Honeywell (USA) and Endress+Hauser (Germany) commercialized silicon piezoresistive chips, driving the modern transmitter. In 1985, the HART protocol overlaid digital configuration onto the 4-20 mA analog loop, forming today's most widespread industrial signal system. After 2000, MEMS (Micro-Electro-Mechanical Systems) shrank sensors to grain-of-rice size, entering automotive, consumer electronics, and medical devices.

In terms of application scale, pressure sensors span nine orders of magnitude: from 10^-6 Pa vacuum levels on spacecraft walls to 700 MPa ultra-high pressure in hydroforming presses. Each order of magnitude corresponds to different sensing principles and material choices. A "universal pressure sensor" does not exist; the essence of engineering selection is mapping process requirements to specific physical principles and mechanical structures.

Four engineering metrics determine pressure sensor quality: accuracy, long-term stability (annual drift), overpressure rating, and process temperature limit. These four collectively determine the total cost of ownership over the sensor's lifecycle. A cheap sensor has low initial purchase cost, but high annual drift and failure rate, requiring annual recalibration. Within three years, total cost often exceeds that of an industrial-grade product purchased upfront.

Pressure Measurement Types

Based on reference pressure, industrial pressure measurement falls into five types: absolute, gauge, differential, sealed gauge, and vacuum. Choosing the wrong type is the most common beginner mistake, causing an entire loop to read off by one atmosphere (approximately 1.013 bar), with especially severe consequences in low-pressure processes. The table below presents the core differences among the five types.

Absolute pressure uses vacuum as zero reference; the sensing element's back cavity is evacuated and permanently sealed. Barometers, altimeters, and chemical reactor headspace pressures require absolute units because atmospheric pressure itself fluctuates by plus-or-minus 30 mbar with weather and altitude, introducing systematic error in relative measurements. Absolute transmitters are more complex to manufacture; back-cavity weld hermeticity is the critical process challenge.

Gauge pressure uses local atmospheric pressure as zero reference; the sensing element's back cavity communicates with the atmosphere through a breather port. This is the most common type in industry: water pipe pressure, tank pressure, and tire pressure all use gauge. When atmospheric pressure changes with weather, gauge readings automatically follow, preserving the physical meaning of "how much above atmospheric." Note that the breather port needs a dust and moisture filter; otherwise, zero drift occurs in salt spray or dusty environments.

Differential pressure measures the difference between two ports and is the most complex and useful type in industrial instrumentation. Measuring pressure drop across a throttling element (orifice plate, Venturi tube) derives flow; measuring pressure drop across a filter inlet/outlet detects clogging; measuring pressure difference between tank top and bottom derives level. The key metrics for DP transmitters are static line-pressure rating and zero temperature drift, because common-mode high pressure on both sides amplifies zero offset.

Sealed gauge permanently seals factory atmospheric pressure in the back cavity without external communication. It suits subsea and deep-well applications where atmospheric reference is impractical. Readings shift with local atmospheric pressure changes, but the magnitude is negligible at engineering ranges. Vacuum is the negative extension of gauge, reading from 0 to -1 bar. It is physically still a gauge sensor with the low-range end optimized.

Mainstream Sensor Technologies

From the sensing element's physical principle, four mainstream pressure sensing technologies dominate industry: metal strain gauge, capacitive, piezoelectric, and MEMS piezoresistive. Each principle has its optimal range, accuracy, cost, and temperature envelope. There is no "universal" principle. The table below compares key engineering metrics of the four technologies.

Metal strain gauge bonds constantan or NiCr alloy foil to the back of a stainless steel diaphragm, forming a Wheatstone bridge. Pressure deforms the diaphragm, changing foil resistance with strain, and the bridge outputs millivolt-level voltage. Advantages include good linearity, shock resistance, and temperature coefficient compensable to plus-or-minus 0.02% FS/K. Disadvantages include sensitivity to bonding process and long-term creep at the 1% FS level, requiring welded or thin-film sputtered strain gauges for high-accuracy applications.

Capacitive uses two metal or ceramic electrodes as a variable capacitor; diaphragm deformation changes the plate gap, and capacitance change is converted to frequency or pulse width via an oscillator circuit. Capacitive sensing is extremely sensitive at low pressures (1 mbar full scale) and is the de facto standard for high-end DP transmitters. Representative products include Rosemount 3051, Yokogawa EJX, and E+H Cerabar S. Disadvantages include structural complexity, high unit cost, and sensitivity to media dielectric constant changes (wet gas applications require exclusion).

Piezoelectric uses quartz or PZT crystals that generate charge under pressure. It has no moving parts, response frequency can reach 100 kHz, and is the only choice for measuring engine combustion pressure, explosion shock waves, and water hammer transients. However, piezoelectric signals "discharge" to zero under static pressure, so it can only measure dynamic pressure, not steady-state process monitoring. Representative manufacturers include Kistler (Switzerland) and PCB (Germany).

MEMS piezoresistive implants boron ions into a single-crystal silicon diaphragm to form piezoresistors, then packages them as a chip. MEMS's greatest advantage is low batch manufacturing cost (under 5 USD per die), with sizes down to grain-of-rice level, making it the only practical choice for automotive TPMS, medical ventilators, and smartphone barometers. Disadvantages include the need for silicone oil or gel isolation (media cannot directly contact the silicon chip) and temperature limit typically at 125 degrees C or below.

Process Media and Materials

Media compatibility determines the metal grade selection for the wetted diaphragm and process contact parts. A mismatch leads to pitting, stress corrosion cracking, or even diaphragm perforation and leakage. Common wetted materials include austenitic stainless steel 316L, nickel-based alloys Hastelloy C-276 / C-22, Grade 2 titanium, alumina ceramic, tantalum, and Monel 400.

316L stainless steel contains 16-18% Cr, 10-14% Ni, and 2-3% Mo. Its low carbon content (0.03% max.) inhibits intergranular corrosion, making it the default choice for 80% of process industry applications. It is fully compatible with water, steam, air, light hydrocarbons, and dilute alcohols, and can serve long-term with moderate organic acids (acetic acid, formic acid at 50% or less). However, 316L does not withstand chloride-containing media (hydrochloric acid, wet chlorine gas, seawater above 200 ppm Cl-), with a pitting threshold around 30 degrees C + 1,000 ppm Cl-.

Hastelloy C-276 is a nickel-based alloy with 16% Cr, 16% Mo, and 4% W, offering five times or more the pitting resistance of 316L against hydrochloric acid, wet chlorine, ferric chloride, and seawater (PREN approximately 67). It is the corrosion-resistance first choice in chemical and pharmaceutical industries. Its work-hardening rate is high and weldability is inferior to 316L, with unit cost 6 to 8 times that of 316L. Hastelloy C-22 is an upgraded version with improved compatibility for oxidizing media (HNO3, HF).

Ceramic diaphragm (Al2O3 99.9%) has stable dielectric constant and extremely high hardness. Its HV1500 hardness makes it virtually immune to particle abrasion (cement slurry, mining slurry, coal powder). Ceramic also provides electrical insulation, avoiding cathodic protection circuit interference. The disadvantage is brittleness; it fractures suddenly under impact or mechanical stress, so it is typically embedded with metal seal rings rather than fully welded, requiring O-ring seals (FFKM or EPDM) for long-term hermeticity.

The table below is a quick-reference lookup for common media and recommended wetted materials. It is intended for initial selection only; before engineering implementation, always obtain the manufacturer's corrosion chart and verify specific concentration, temperature, and flow velocity.

Key Specification Parameters

Reading spec sheets is a fundamental skill for purchasing engineers. The same transmitter may list 12 to 30 parameters across different manufacturer spec sheets, but only eight truly drive selection decisions: accuracy, repeatability, hysteresis, temperature compensation, long-term stability, response time, overpressure rating, and output signal. Each is explained below.

Accuracy is typically expressed as % FS (percent of full scale) or % SP (percent of set point). 0.5% FS on a 10 bar span means maximum error at any point is plus-or-minus 50 mbar; 0.5% SP at a 1 bar set point is only plus-or-minus 5 mbar, but at 0.1 bar becomes plus-or-minus 0.5 mbar. Note that manufacturers often use "accuracy" as a composite term for non-linearity, hysteresis, and repeatability combined, but IEC 60770 requires listing three independent metrics. When purchasing high-end transmitters, request itemized data.

Repeatability is the output scatter of multiple measurements of the same pressure under identical conditions, the core metric for evaluating sensor random error. A sensor with repeatability better than accuracy suits relative measurement (rate of change, flow totalization); a sensor with poor repeatability but high accuracy is limited to absolute-value readings. Hysteresis is the output difference at the same pressure point between ascending (loading) and descending (unloading) cycles. Metal strain gauge hysteresis is typically 0.05% FS or less; polymer diaphragms can reach 0.5%.

Temperature compensation specifies the upper limits for zero drift and span drift within a given temperature range. Quality transmitter specs typically read: less than 0.02% FS/K (zero) + less than 0.04% FS/K (span), within -20 to +80 degrees C. The compensated range is the temperature window within which the manufacturer guarantees accuracy; outside this range, accuracy is not assured. For critical applications, request measured thermal drift curves across the full operating temperature range (-40 to +85 degrees C).

Output signal is the interface to the control system. Five mainstream types:

• 4-20 mA: Two-wire current loop, most widespread, strong noise immunity, can transmit over 1 km. Default choice for 80% of industrial applications.
• 4-20 mA + HART: Overlays plus-or-minus 0.5 mA of 1,200/2,200 Hz digital modulation on the current loop, enabling remote configuration, diagnostics, and variable upload.
• 0-10 V: Three- or four-wire voltage output, common for PLC analog input cards, but long-distance voltage drop is problematic.
• Foundation Fieldbus / PROFIBUS PA: Pure digital bus, multiple devices sharing one cable pair, suited for large DCS projects.
• IO-Link / Modbus RTU / Ethernet (PROFINET, EtherNet/IP, HART-IP): Smart factory trend, convenient for data centralization.

Overpressure rating is the maximum pressure the sensor can withstand without damage, typically 2 to 4 times the upper range limit. Burst pressure is the diaphragm's mechanical rupture limit, 2 to 5 times the overpressure rating. Water hammer, cavitation, and start-stop transients can instantaneously generate peaks far exceeding working pressure; insufficient overpressure rating causes permanent zero offset.

Selection Decision Factors

To apply the knowledge from the preceding five chapters to a specific model, follow the decision sequence below. Most selection mistakes occur not from a single wrong step, but from premature decisions at the wrong level. These eight steps can serve as a fixed RFQ template.

1. Pressure type and range: First decide absolute / gauge / differential / vacuum, then derive the appropriate range using "operating point sits within 30 to 70% of span." For differential service, additionally specify static line-pressure rating.
2. Accuracy class: Distinguish among loop control (0.5% FS is sufficient), custody transfer (0.1% FS or better), and reference standard (0.025% FS or better). Each accuracy tier roughly doubles the price.
3. Process connection and materials: Threaded (G1/2, NPT1/2, M20x1.5) / clamp (Tri-Clamp 1.5 inch, 2 inch) / flanged (DN50 PN40). Select diaphragm material per Chapter 4 media table.
4. Electrical connection and ingress protection: M12-A 4-pin / 7/8 inch / Hirschmann / cable gland. Housing IP65 / IP67 / IP68. Outdoor or washdown scenarios require IP67 or above.
5. Certifications: Process safety SIL2 / SIL3, explosion-proof ATEX / IECEx / NEPSI / FM, sanitary 3-A / EHEDG, metrology OIML R109, Pressure Equipment Directive PED 2014/68/EU.
6. Temperature environment: Media temperature (determines isolation approach), ambient temperature (determines electronics thermal drift), vibration level (IEC 60068-2-6 typically 4-150 Hz, 5g).
7. Output signal and protocol: 4-20 mA HART is default; new installations may consider PROFIBUS PA / Fieldbus / Ethernet-APL; OEM integration considers IO-Link / I2C / SPI.
8. Total cost of ownership (TCO): Purchase price + installation + calibration (0.5 to 1 man-hour per year) + spare parts + failure-downtime losses. A sensor that saves 500 USD upfront but drifts 1% FS per year can cause process waste within three years that far exceeds the purchase price difference.

One last commonly overlooked dimension is manufacturer serviceability: local spare parts inventory, field calibration service availability, HART DD file registration with FieldComm Group, and firmware upgradability. These seem irrelevant at the purchasing stage but determine repair response time after 5 to 10 years of production line operation. WIKA, Endress+Hauser, Emerson, Yokogawa, Honeywell, and ABB have all established calibration laboratories and spare parts centers in China, making them reliable choices for large projects.