Three ideal pressure sensors observe the same imagined chamber at the same time. One reports 130 kPa, another reports 30 kPa, and a third reports 10 kPa. Before deciding that two of them are wrong, ask what each reading is compared with. The first measures from an ideal vacuum reference. The second compares the chamber with surrounding air at 100 kPa absolute. The third compares it with another pressure point at 120 kPa absolute. Their calculations are 130−0, 130−100 and 130−120. All three readings fit the stated situation.

The names for these measurements are absolute, gauge and differential pressure. Manufacturers distinguish them by their references. [1][2] The useful question is therefore not just “What number did the sensor show?” It is “Pressure relative to what?” This article follows invented values and ideal sensors to answer that question. It does not ask you to build, connect, open or test pressure equipment. We will keep the physical pressures and their labels visible, then change one assumption at a time. That lets us explain a disagreement without mistaking different measurement definitions for a fault, or treating a calculation as evidence that a real instrument is accurate.

Let a chamber have absolute pressure PA=130 kPa, local ambient pressure Pamb=100 kPa and a second comparison point PB=120 kPa. An ideal absolute measurement returns PA=130 kPa. A vented gauge measurement returns Pg=PA−Pamb=30 kPa. A differential measurement with the declared direction A minus B returns ΔP=PA−PB=10 kPa. These values represent different measurands associated with the same applied chamber pressure. Equality of their numerical values is not the criterion for consistency.

The reference distinction is documented in NI's pressure-measurement explanation and TE's measurement-type definitions. [1][2] We use it to separate three questions that are often merged: which physical point is sensed, which reference defines the reported pressure, and how an electrical signal represents that quantity. The examples assume simultaneous ideal measurements of stated pressures, with no uncertainty, transient or installation error. Real comparisons require those additional considerations. No pressure experiment or instrument calibration was performed. The aim is an interpretation method: a discrepancy between two displayed numbers becomes a meaningful accuracy question only after the pressure quantities, units, points and conditions have been made comparable.

A pascal is a unit of pressure: one newton of force per square metre. A kilopascal is one thousand pascals. NIST lists the pascal's force-per-area definition and the meaning of the kilo prefix. [4] Changing 30 kPa to 30,000 Pa changes the numerical scale, not the physical quantity. If the first number was gauge pressure, the second is still gauge pressure. Converting units cannot silently change its reference to a vacuum.

A complete label needs both pieces of information. “130 kPa absolute” and “30 kPa gauge, relative to 100 kPa ambient” tell us enough to recognize the same chamber in our example. “30 kPa” on its own does not. Pressure sensor documentation may use abbreviations, suffixes or separate configuration fields for these distinctions, so their meaning needs to be read in context. [1][2] Think of the unit as describing the size of each numerical step, while the reference tells you where the comparison begins. Two records can share kPa units and still represent different quantities. Conversely, two records in different units can represent the same quantity after a proper scale conversion. Match the reference as well as the unit before comparing or combining them.

The SI relation is 1 Pa=1 N/m², and the decimal prefix gives 1 kPa=10³ Pa. [4] A unit transformation rescales the value of a specified quantity. It does not replace PA with PA−Pamb. In particular, 30 kPa gauge is 30,000 Pa gauge, whereas its corresponding absolute pressure in our initial state is 130,000 Pa. The latter calculation requires a reference pressure in addition to a unit conversion.

Keep the measurand definition separate from the unit symbol. Writing pressure-reference words explicitly avoids relying on a vendor's suffix convention. NIST notes more generally that a unit name alone does not fully define the measured quantity; the same dimensional unit can serve different quantities. [4] Here the relevant distinction is the boundary against which a pressure is reported. Metadata should therefore preserve both a recognized unit and an explicit reference type or reference value. A database column that records only a floating-point value and “kPa” has lost information needed to reconcile the three readings above. Extra decimal places cannot reconstruct that missing definition, and matching numeric ranges cannot establish that two channels share a reference.

Many pressure sensors contain a flexible diaphragm. Pressure loading changes its shape, and the device turns that change into an electrical signal. NI describes strain-gauge designs in which deformation changes electrical resistance, as well as capacitive designs in which movement changes capacitance. [1] The electronic method explains how a signal is produced. It does not, by itself, tell us which pressure reference the device uses.

For a gauge arrangement, the reference can communicate with local atmospheric pressure. An absolute arrangement uses a vacuum reference. A differential arrangement accepts two pressure inputs and measures their difference. TE describes these distinctions in its measurement types. [2] The pressure on the comparison side matters even when the user mainly notices the side connected to the chamber. That is why changing a number's label in software cannot physically change a sensor's reference arrangement. A valid software conversion can calculate a different quantity if the required reference information is also available, but merely renaming a gauge channel “absolute” adds no such information. The mechanism and the calculation must agree on what was sensed and what was subtracted.

A diaphragm-based transducer responds mechanically to loading across the sensing structure. Strain-gauge implementations can convert deformation to resistance changes in a bridge; capacitive implementations use geometry-dependent capacitance. NI describes both mechanisms and notes that bridge-based designs can serve absolute, gauge or differential measurements. [1] Thus “strain gauge” identifies a transduction approach, whereas “gauge pressure” identifies a pressure reference. The shared word does not make the classifications equivalent.

For the reference convention, an absolute device may incorporate a vacuum reference, a vented gauge device communicates with the local atmosphere, and a differential device accepts two pressure ports. [2] Actual packaging and compensation vary, so these concepts are not a construction specification. The important measurement equation is that the output represents a defined pressure or difference. An electronic scaling operation acts on that output; it does not alter the physical boundary on the other side of the diaphragm. Converting a measured gauge quantity to absolute pressure requires the appropriate ambient reference, while converting a two-port difference into either endpoint pressure requires additional information about the other endpoint. Neither quantity can be reconstructed by changing an axis label alone.

A sealed-gauge description adds another possibility. Instead of following the current local atmosphere through a vented reference, the instrument uses its documented sealed or fixed-offset reference arrangement. TE describes implementations that can use an absolute reference together with an electronic offset representing a standard atmosphere. [2] The exact construction and offset belong to the device documentation; “sealed” is not enough information to calculate them.

For a clean mathematical comparison, invent a reading defined as chamber absolute pressure minus a fixed 100 kPa reference. At 130 kPa absolute, that reading is 30 kPa. If local air changes from 100 to 95 kPa while the chamber stays at 130 kPa absolute, this fixed-offset reading remains 30 kPa. A vented gauge would instead read 35 kPa. The two devices are answering different reference questions. The fixed 100 kPa in this example is deliberately chosen; it is not a claim that all sealed gauges contain that pressure or keep a physical trapped gas perfectly unchanged. Likewise, subtracting one stored baseline in software does not automatically create a continuously atmosphere-referenced measurement. You need the actual reference definition and its behavior to know what the result means.

A sealed-gauge or gauge-like fixed-offset output must be interpreted through its documented reference implementation. TE distinguishes it from a vented gauge and describes an implementation using an absolute reference with an electronic standard-atmosphere offset. [2] That is not a universal claim about the contents or stability of every sealed reference cavity. A physical cavity and a compensated electronic offset can have different error mechanisms even when their nominal output conventions look similar.

Define a hypothetical reported quantity Ps=PA−Pref with Pref fixed at 100 kPa. For PA=130 kPa and Pamb=95 kPa, Ps=30 kPa but Pg=PA−Pamb=35 kPa. The difference Pg−Ps=Pref−Pamb=5 kPa follows directly from their references. A fixed tare on an absolute-pressure channel has the same algebraic character: it subtracts the saved value, not necessarily the current ambient pressure. Its usefulness depends on which quantity the application intends to report. Treating it as a live atmospheric reference after conditions change would change the measurement meaning without changing the arithmetic. Preserve the offset definition and its time or calibration basis instead of collapsing all gauge-like readings into one unlabeled pressure field.

A sensor might produce a voltage, current or digital value. TE lists several output types, including amplified voltage and digital interfaces. [2] The output format does not by itself reveal whether the underlying pressure is absolute, gauge or differential. Imagine a purely illustrative linear scale: 1 volt represents 0 kPa and 5 volts represents 100 kPa of the declared measured quantity. A 3-volt signal sits halfway along that scale, so it represents 50 kPa.

If that imaginary scale is absolute, the result means 50 kPa absolute. If it is gauge and the ambient reference is 100 kPa absolute, the result means 150 kPa absolute. The same numerical voltage and range labels can therefore need different physical interpretations. This is a calculation about a hypothetical transfer rule, not wiring guidance. A real device's permitted supply, output and measurement range require its own documentation. Similarly, a differential range of 10 kPa does not state the maximum pressure allowed at either port. Our 130/120 and 1,030/1,020 kPa pairs have the same difference but very different absolute levels. The value a sensor is intended to measure and the conditions it can withstand are separate parts of its specification.

For the invented linear transfer V=1 V+(4 V)Q/(100 kPa), Q is the explicitly declared pressure quantity. At V=3 V, the normalized fraction is (3−1)/(5−1)=0.5 and Q=50 kPa. If Q=PA, the inferred absolute pressure is 50 kPa. If Q=Pg with Pamb=100 kPa, PA=Q+Pamb=150 kPa. The transfer function supplies scaling; the reference definition supplies the meaning of Q. Neither can substitute for the other.

Manufacturers provide analog and digital output options and pressure-range information as different specifications. [1][2] A numerical transfer example cannot establish electrical compatibility, calibration uncertainty, media compatibility or pressure survivability. The same separation applies to differential and line-pressure limits. A small PA−PB does not bound PA or PB individually: adding any common offset preserves the mathematical difference. Real devices may also have common-mode effects and asymmetric port limits that are absent from our ideal equations. Before interpreting a differential output at a stated line pressure, consult the actual device contract. The calculation demonstrates why that information is necessary; it does not supply a safe operating range or a procedure for applying the example pressures.