What is Humidity sensors?

Humidity sensors: measurement principles and industrial control

Humidity sensors: what the reading actually means

A humidity sensor responds to water vapour in air or another gas. For room monitoring and many industrial applications, the displayed quantity is relative humidity, written as %RH. That number is useful only when you know the temperature and what part of the process is being sampled. Relative humidity, absolute humidity and dew point are different quantities, not alternative names for the same reading. [1]

The sensing element is only one part of the instrument. A humidity transmitter combines the sensor with electronics and a usable interface; a controller or PLC can then make an operating decision. Typical applications include greenhouses, HVAC, storage monitoring and industrial drying. [5]

Why heating can make the percentage fall

Relative humidity compares the partial pressure of water vapour with the saturation vapour pressure at the same temperature. A reading of 60% RH does not mean that 60% of the air volume consists of water. [9]

RH (%) = 100 × pv / psat(T)
Use the same pressure units for the actual water-vapour partial pressure, pv, and the saturation value, psat(T). [1]

As an illustrative calculation, air at approximately atmospheric pressure and 25 °C / 60% RH falls to about 45% RH when heated to 30 °C at constant pressure without adding or removing water. The lower percentage is not evidence that water has been extracted. Pair the humidity channel with a representative temperature measurement. [9]

Absolute humidity is water-vapour mass per unit volume, commonly expressed in g/m³. Dew point is the temperature at which cooling the gas to saturation would allow condensation to begin at its existing pressure and water-vapour content. A surface below that temperature can collect condensation. For compressed gases, the relevant measurement pressure must also be stated. [1]

Three sensing approaches, with different roles

Capacitive sensing uses a moisture-sensitive polymer film. As the film exchanges water molecules with its surroundings, its dielectric properties change, producing a measurable change in capacitance. Electronics convert that response into a calibrated humidity value. The principle does not require visible droplets to form inside the sensor. [2]

Resistive sensing tracks the changing resistance or impedance of a moisture-sensitive material. The required excitation and temperature correction depend on the element. These descriptions refer to the sensing principle, not to whether the final instrument has an analog or digital output. A humidity element should not be confused with a capacitive proximity sensor used to detect objects.

Optical moisture analysis exploits the absorption of light by water vapour at selected wavelengths. TDLAS, for example, uses tunable laser absorption for gas analysis. Such process instruments are not simply room RH sensors with a different housing. [4]

Watch the output at 59% RH

The example below requests dehumidification at RH ≥ 60% and removes the request at RH ≤ 58%. Between those limits, the controller keeps its previous output state. This two-percentage-point hysteresis band reduces repeated relay switching around one threshold. Controller hysteresis is a deliberate operating setting; it is not the same as the sensing element’s measurement hysteresis.

Dehumidifier control: why two thresholds matter At 55 percent RH the command is off. At 60 it switches on, remains on at 59, and switches off at 58. Stop freezes the frame; Start restarts the sequence. Switch on: 60 %RH · Switch off: 58 %RH Relative humidity · constant temperature 50 58 60 70 %RH Dehumidification command 55 %RH OUTPUT OFF Below the lower threshold. Initial state: no demand. 60 %RH OUTPUT ON The upper threshold is reached. Dehumidification is requested. 59 %RH OUTPUT ON The previous state is retained. The lower limit has not been reached. 58 %RH OUTPUT OFF The lower threshold is reached. The demand is removed. Illustrative sequence: 55 → 60 → 59 → 58 %RH. Not live process data.

The four stages are deliberately fixed: 55 → 60 → 59 → 58% RH. At 59%, the output stays on because the lower switching point has not yet been reached. Start restarts the sequence; Stop freezes the displayed state. This is a teaching sequence, not live sensor data or a physical process simulation. The thresholds are illustrative, not universal humidity targets.

A real installation may also need minimum run and off times to protect the controlled equipment. The command state must be distinguished from the physical NO/NC contact state. PID humidity control is another option where the process and actuator support it; neither switching logic nor fault behaviour should be copied into machinery without application-specific design.

Connect the signal, then verify the scale

Check that the chosen 4–20 mA or 0–10 V output matches the receiving analog input. With Modbus RTU, the device address, serial settings, register map and numerical scale must agree. Not every transmitter provides every interface. Use the actual model’s drawing for power, commons and cable or connector assignments. [5]

For a transmitter configured specifically as 4–20 mA = 0–100% RH, the linear conversion is RH (%) = (I − 4) / 16 × 100, with I in mA. Therefore, 12 mA represents 50% RH and 13.6 mA represents 60% RH. Recalculate the scale when the transmitter range changes; those values are not universal defaults.

Missing data must not be accepted as dry air. Handle a broken loop, invalid register value or transmitter fault separately from a valid low-humidity reading. Recording both temperature and RH with a data logger and trending them in SCADA makes the sequence of events easier to interpret.

Read the specification beyond “0–100% RH”

Accuracy and resolution: A stated limit of ±2% RH means two RH percentage points under the specified conditions, not two percent of the displayed value. A fine display increment does not prove equally fine accuracy. Check temperature dependence, operating range and long-term drift, and distinguish typical figures from guaranteed limits. [6]

Response time: A specification such as t63 describes the time to respond to about 63% of a humidity step, rather than the time to settle completely. Filters, protective caps and air movement affect installed response. Polling a slow probe more frequently does not make its physical response faster. [3]

Location: Sample the air you actually need to control. Nearby electronics, a warm motor, sunlight or a heated wall can change the probe’s local temperature. Temperature compensation cannot make a badly positioned sensor representative of a different location. [8]

Environmental limits: Consider condensation, solvent exposure, dust and cleaning agents. An IP rating alone does not establish chemical compatibility or measurement accuracy while wet. Select suitable filters and sensor accessories without unnecessarily restricting vapour exchange at the sensing surface. [3]

Calibration is a comparison; adjustment changes the instrument

Calibration establishes measurement error by comparison with a traceable reference. Adjustment changes the instrument’s correction when needed. Let the reference and test probe reach equilibrium in the same conditions before comparing them. Set the service interval using manufacturer guidance, process risk and historical drift; factory calibration is not a lifetime guarantee against change. [7]

Two checks that avoid misleading tests

Can a Pt100 measure humidity? A Pt100 is a temperature element. It may provide the temperature channel in a humidity instrument, but a separate humidity-sensitive mechanism is still required.

Is breathing on the probe a calibration test? No. Breath changes temperature as well as water-vapour conditions, and provides no known reference value. Deliberately wetting a probe is not a valid accuracy test either and may disturb the measurement. Use a suitable reference and follow the manufacturer’s cleaning instructions rather than touching or chemically treating the sensing film. [3]