Motion sensors are usually evaluated on sensitivity, as though a more sensitive sensor were a better one. In practice the parameter that decides whether a product is pleasant or maddening is the geometry of the Fresnel lens in front of the pyroelectric element, and that geometry is fixed at the tooling stage. Firmware can adjust how long a detection must persist before it reports; it cannot move a detection zone that the lens has already created.
Detection is a set of lobes, not a cone
A PIR sensor does not see a room. A lens segments the field into a series of zones, and the element responds when an infrared source crosses from one zone to another. That is why a person walking straight at the sensor can be missed while someone moving across it triggers instantly, and why a sensor that fires on an empty hallway is usually responding to a warm surface changing temperature — a radiator cycling, sun moving across a floor, an air-conditioning vent.
Any adjustment of so-called sensitivity is usually a change to the reporting threshold and the hold time, which trades false triggers against missed detections. It does not change which parts of the room generate zones.
Curtain, dome and long-range patterns
Three lens patterns cover most products. A dome pattern gives broad coverage for a room and is what ceiling-mounted sensors use. A curtain pattern produces a narrow vertical wall of zones, which suits doorways, window lines and perimeters where you want detection only along a boundary. A long-range pattern concentrates narrow zones for corridors, driveways or fences where the subject is far away and small in the frame.
Specifying the wrong one produces a product that fails in a specific, predictable way: a corridor sensor with dome optics detects the wall beside it and misses the far end; a perimeter sensor with long-range optics creates blind spots either side of a doorway. Neither is a firmware defect, and neither can be fixed in the field.
Pet immunity is a lens and algorithm claim
Pet immunity typically combines a lens pattern that ignores the lower zone with an algorithm that weights the size of the detection. It works best for animals below the stated weight, walking rather than climbing, and inside a defined distance. A cat on a table is at the height of a person and produces a person-sized signal. Buyers launching a pet-immune product in a market with a large number of climbing cats should expect a proportion of complaints regardless of the label, and should position the feature accordingly rather than promising a guarantee.
The material detail that ages badly
Lenses are moulded in different polymers, and not all of them hold their optical clarity equally under ultraviolet light or in high ambient temperatures. A sensor that performs in testing and degrades in service is usually a material story rather than an electronics story, and it is worth asking what the lens is moulded from and whether the part has been tested after accelerated ageing.
What to specify when ordering
The useful specification is: lens pattern, zone count and nominal coverage distance and angle, mounting height assumed in that coverage, pet-immunity limits with the conditions attached, lens material, hold time and reporting behaviour, and battery life at a stated number of detections per day. Those seven lines describe the product far better than a single sensitivity figure, and they are what a distributor needs to position it against the market.
XingLiao builds PIR motion sensors in dome, curtain and long-range patterns as OEM/ODM programmes, with lens tooling and coverage specified per market. Buyers can send the intended application to match a lens pattern before moulds are cut.
The short version
Choose the lens pattern first: dome for rooms, curtain for boundaries, long-range for corridors and perimeters. Treat sensitivity as a reporting threshold, not a detection improvement. Ask for the coverage, the assumed mounting height, the lens material and the pet-immunity conditions, because those describe how the product behaves in a house rather than on a test bench.
Top comments (0)