Skip to content
Website
🧱 CFC 238KNX home automation: topology, addressing and parameter settingChapter 7 of 10 in the section

Switching on right: presence detection, time delays, and scenes that work

This chapter guides you through fine-tuning presence detectors and lighting scenes on a KNX bus after commissioning. You will learn how to adjust coverage based on actual movement, calibrate integrated light sensors, lock in effective run-on times, and eliminate superfluous scenes to ensure comfort without user complaints.

Key takeaways

  • The radial coverage of a passive infrared sensor is always half its nominal tangential coverage.
  • Lux calibration is performed using a luxmeter on the work surface, never at room ceiling level.
  • Semi-automatic mode prevents nuisance switching by leaving turn-on to the user's manual command.
  • Three coherent lighting scenes per room guarantee user adoption without cluttering push-button keys.

Actual coverage, light threshold, and time delay: making detection reliable on site

Manufacturers' datasheets commonly advertise circular detection ranges of 8 to 12 metres for ceiling-mounted PIR detectors. On site, this figure corresponds solely to tangential movement, when a person crosses the lens beams perpendicularly. As soon as the occupant walks straight towards the sensor (radial movement), the usable range drops by half. For sedentary desk work, the so-called "seated presence" or micro-movement zone rarely exceeds a diameter of 4 to 5 metres at a standard mounting height of 2.80 metres. If the detector is mounted too far from the desk, the stationary worker is left in the dark every ten minutes.

  1. Delimiting coverage physically and logicallyIsolate stray transit areas. If a glazed door opens onto a busy corridor, disable the pyroelectric sector facing the opening in the ETS parameters, or clip the plastic masking shield under the dome to block disruptive beams.
  2. Calibrating mixed light measurementTurn off all luminaires in the room in the middle of the day away from direct sunlight. Measure the exact value with a luxmeter on the desk, then enter this setpoint into the KNX sensor's calibration object to recalibrate its internal scaling factor.
  3. Setting the run-on time to suit the roomNever use 60-second time delays in individual offices: they cause constant instability. Set the off-delay between 10 and 15 minutes in sedentary commercial spaces, compared to 2 to 3 minutes in a corridor or transit washroom.

Post-occupancy override and lighting scenes: eliminating overcomplicated setups

Automated lighting becomes intolerable the moment it strips humans of decision-making. In conference rooms or executive offices, "fully automatic" mode causes unwanted switch-ons every time someone briefly passes the open door. The professional solution is to switch to "semi-automatic" mode: the occupant turns on the light deliberately via the wall switch on arrival, and the KNX presence detector then handles only holding and turning off after the room is vacated.

KNX detector behaviour according to configured mode
Operating modeBehaviour on entryBehaviour on exitRecommended use
Fully automaticSwitches on upon entry if actual lux is below thresholdSwitches off after run-on time without detectionCorridors, washrooms, plant rooms, stairwells
Semi-automaticNo automatic switch-on, manual control requiredAutomatic switch-off after vacancy run-on timeIndividual offices, meeting rooms, bedrooms
Surveillance / AlarmSends cyclic telegram above trigger thresholdResets after intrusion endsIntrusion monitoring, out-of-hours movement alert

The post-occupancy override setting determines how the system reacts when a user manually turns off the lighting upon leaving the room. If the sensor takes back control immediately, it switches the room back on at the slightest step from the occupant closing the door. The dead time parameter (typically 15 to 30 seconds) blocks any redetection following a manual switch-off command, leaving sufficient time to leave.

The D+7 review visit: rectifying false triggers and actual thresholds

Initial commissioning of a KNX presence detector in commercial or major residential buildings relies on theoretical values entered into ETS: a standardised switch-off delay, passive infrared sensor sensitivity set to maximum to cover the space, and a lux threshold defined from lighting design recommendations. In practice, after five to ten days of actual room occupancy by users, these factory settings systematically create friction. The sensor detects a worker walking down the adjacent corridor through a glazed fanlight, responds to intermittent heat plumes from a fan coil unit, or plunges a seated worker into darkness every three minutes during reading tasks.

To prevent the operator from ultimately disabling the automation by forcing actuators permanently on via manual controls, an on-site fine-tuning visit is essential at the end of the first week of use. This intervention is not troubleshooting; it constitutes the normal completion of the telematics installer's configuration services. Analysing telegram logs on the bus line highlights rooms experiencing abnormal switching cycles, often exceeding fifteen to twenty switch-ons per hour without continuous occupancy.

Diagnosing detection anomalies and ETS parameter corrections during the inspection visit
On-site user observationPhysical or logical causeTelematics corrective action
Nuisance switch-on with open doorTangential coverage encroaching on circulation areaPhysical sector masking of PIR lens or software sensitivity reduction on access-facing sector
Nuisance switch-off while seated at deskTyping movements outside micro-movement detection zoneExtend run-on time from 5 to 12 minutes and switch sensor to high-sensitivity detection mode
On/off hunting under changing skiesLux measurement hysteresis too narrow for solar variationsWiden lux deadband and extend integration time of integrated sensor
Immediate switch-back-on after switch-offSudden cooling of heating element or luminaire perceived as movementEnable rearming dead time (post-switch-off pause of 2 to 4 seconds) in ETS parameters

Trimming mood scenes and configuring manual override recovery

The second focus of the post-occupancy visit covers wall push-buttons and lighting scene usability. During the design phase, specifications often multiply scenarios: meeting, video projection, reception, cleaning, focus, or relaxation modes. In reality, users adopt only two or three in daily operation. A push-button matrix overcrowded with triggers leads to mis-hits, general frustration, and complaints to construction management. The telematics installer uses the one-week feedback to remove ghost scenes from devices and reassign keys to clear, distinct functions.

The most sensitive point lies in handling recovery following manual intervention. When a user alters lighting at the push-button to force switch-off or dim the level, the KNX presence detector must not forcibly regain control at the slightest subsequent micro-movement. Such overbearing behaviour causes immediate irritation among staff. Two recovery strategies must be decided based on room typology:

  • Switching to semi-automatic mode: turning on is done exclusively by deliberately pressing the switch, while the detector only manages safety switch-off after prolonged vacancy (recommended for individual offices and conference rooms).
  • Override time delay: any manual action freezes the lighting state for a defined period (typically 45 to 90 minutes) or until the room remains empty for the entire basic run-on time, signalling the end of occupancy.

During this adjustment, the telematics installer ensures that as-built documentation faithfully reflects the configured reality. In accordance with the professional diligence requirements set out in the Code of Obligations, particularly art. 364 CO which subjects the contractor's care to the same rules as that of an employee, the finalised ETS project file must record every adjustment to coverage, run-on times, and group address mapping. This documented file is handed over to the network administrator or client to conclude the building automation integration phase definitively.

Handling complaints and on-site dispute protocol

When a client complains about flickering lights or a detector plunging a room into darkness during a work session, the fatal knee-jerk reaction is to disable automation or extend the run-on time to forty-five minutes. This drift turns a connected installation into an energy pit and creates lasting frustration. For an immovable work, the client has 60 days to give notice of defects under art. 367 CO. The telematics installer must therefore intervene methodically to eliminate malfunctions as soon as initial operating feedback is received.

  1. Diagnosis using ETS loggingConnect the USB or IP interface to the control bus and run the ETS group monitor for at least two hours during normal occupancy. Filter telegrams by switching and measured lux group addresses. This log immediately distinguishes a thermal false trigger caused by a convector heater from an erroneous cyclic command sent by a misconfigured device.
  2. Physical lens adjustmentCheck the detection zone and actual range of the passive infrared sensor. Fitting adhesive masking shutters provided by the manufacturer blocks sectors facing a busy corridor or glazed door, preventing nuisance switch-ons caused by people walking outside the room.
  3. Recalibrating time delays and lux levelsMeasure actual illuminance with a luxmeter on the working plane at 0.80 m above finished floor level. Readjust the light threshold in the ETS application to compensate for reflection drift caused by dark furniture. Set the run-on time between twelve and fifteen minutes for an individual office—an ideal value to bridge typing micro-movements.
  4. Simplifying scenes and unblocking overridesPurge complex scenes that are never called up, keeping only three clear moods on push-buttons. Configure automatic return to detection mode after manual override: a two-hour inactivity delay rearms the sensor, ensuring night-time switch-off even if the master off-command is forgotten.

Frequently asked questions

How do you distinguish a motion detector from a presence detector?

A motion detector detects broad movements and switches off the load as soon as action stops, which suits circulation areas. A presence detector incorporates high-resolution optics capable of capturing micro-movements, such as typing on a computer keyboard. In addition, it continuously measures ambient light to switch off the luminaire if natural daylight exceeds the set threshold, even while the user remains seated at their desk.

Why does the light turn off abruptly while the room is occupied?

This fault stems from an underestimated run-on time or insufficient actual coverage relative to the person's static position. If the run-on time is set below ten minutes, prolonged immobility while reading leads to switch-off. In ETS, adjust the switch-off delay to twelve or fifteen minutes and direct the detection zone towards the workstation, or set the sensor to high-sensitivity mode on channels dedicated to micro-movements.

What should be done when a manual override locks out automatic switch-off in the room?

This lockout occurs when the manual control object locks the sensor without an auto-release timer. To correct this behaviour in ETS, configure the post-occupancy override parameter on a rolling time basis. Following a manual input on the wall push-button, prolonged absence of movement for thirty minutes must override the local bypass, force the line off, and rearm automatic presence detection for subsequent users.

How many lighting scenes should be assigned per workspace?

On site, experience shows that multiplying mood scenes makes the interface incomprehensible to users, who ultimately end up using none. Limit yourself to three effective scenes clearly identified by engraving: a Work scene at 500 lux, a Meeting or Projection scene with low dimming, and an All-Off scene. Overly subtle intermediate scenes are systematically ignored and add unnecessary complexity to maintenance troubleshooting.

Texts and standards cited

  • CO art. 367after delivery, the client inspects the condition of the work as soon as feasible and gives notice of defects to the contractor; for an immovable work, the client has 60 days to give notice, and any shorter contractual period is void (para. 1bis, in force since 1 January 2026); each party may request an examination by experts (para. 2).

In the same section