KNX Sensor

A sensor places its result on a group address and is done. Whether an actuator is listening there, three actuators or none at all, it does not learn and does not need to. This is the reason a KNX installation can be extended without touching existing devices.

The price of this freedom is care with the data point type. A temperature sensor sends a 2-byte floating point value, a push button a single bit. Anyone putting both on the same group address gets no error message, but wrong values.

Operating devices are the most common case: push buttons, button sensors, scene controllers. They turn a finger press into a telegram, often with several functions per rocker and a status light that in turn is fed by the actuator.

Measuring sensors deliver numbers rather than commands. Temperature, humidity, brightness, wind speed, CO₂. They send either at fixed intervals or only once the value has changed by a set amount, and the second setting determines the bus load.

Reporting sensors capture states: presence, movement, an open window, smoke. And finally the binary input brings conventional technology onto the bus by translating a potential-free contact into a telegram, such as a classic switch or a window contact.

Many devices belong to several groups at once. A weather station measures, reports threshold violations and supplies the time, and a room sensor combines operation and measurement in one housing.

Unlike with the actuator, the mounting location of a sensor is a technical decision and not a question of space. A temperature probe above a radiator measures the radiator. A presence detector above the dining table sees the table and not the door. A brightness sensor facing a white wall reports the wall.

This can later only be corrected with chasing work, which is why positioning belongs in the planning and not on the building site. The actuator, by contrast, can be swapped in the distribution board at any time.