Because a PIR detects movement across its zones far better than movement towards it - so the detector should face across the likely route, not along it.

The segmented lens creates a fan of zones. Detection happens when a warm body crosses from one to another.

Someone walking straight at the detector stays within a single zone for much of the approach and may get close before being detected.

Corner mounting is the usual answer because it makes almost every route across the room a crossing route.

Mounting height matters too: too high and the zones spread so wide that a person crosses few of them, too low and coverage is short.

Follow the manufacturer's height range - it is what the lens was designed around, not a general recommendation.

Moving heat that is not a person: sunlight, heaters, air conditioning, reflective surfaces and animals.

Direct sunlight tracking across a floor is the classic cause, and it is seasonal, so a detector installed in winter can start alarming in spring.

Heaters and air conditioning outlets create moving warm air, which the detector reads as a moving heat source.

Anything that moves and is warm - a swinging sign near a heat source, curtains over a radiator - will do the same.

Small animals are detected by any detector without a pet-immune lens, and pet immunity has weight and mounting conditions attached.

The fix is siting, masking or a different lens. Reducing sensitivity turns a visible problem into an invisible one.

The lens is shaped so that the zones close to the floor are absent or insensitive, so an animal below a certain size does not register.

Manufacturers quote a weight limit, and that figure assumes the detector is mounted at the specified height and angle.

Mounted lower, or tilted down, the immunity is lost and the animal is detected.

Animals that climb defeat it entirely. A cat on furniture is at human height and will trigger the detector.

Multiple animals together can also present enough signal to be detected even if each is individually below the limit.

Where pets have full run of a protected space, a different technology or a different protection strategy - perimeter rather than volumetric - is usually better than fighting the lens.

Yes. The detector works on the contrast between body heat and the background, so detection falls off as the room approaches body temperature.

In a hot environment the difference between a person and the wall behind them is small, and the signal is correspondingly weak.

That matters in warm climates, plant rooms and unventilated spaces in summer.

Coverage figures on the datasheet are measured at moderate room temperature and should be treated as a best case.

Clothing reduces the signal too, so a well-covered intruder in a warm room is the hardest case of all.

Where high ambient temperature is normal, dual-technology or a different sensing principle is a more reliable answer than a better PIR.

No. Ordinary glass blocks the infrared wavelengths the detector uses, so it cannot see through a window - and cannot be triggered through one either.

That is useful: a detector facing a window is not set off by people or vehicles outside.

It also means a PIR cannot be used to watch an area through glazing, which surprises people specifying detection into an atrium or a display window.

Sunlight through the window still matters, because the sun heats surfaces inside the room and those surfaces are what the detector sees.

For detection at a glazed line, a beam detector, a glass break detector or an external detector is the right choice.

The same principle applies to plastic sheeting, partitions and display cases: the detector sees the surface, not what is behind it.