IR detector problem - affected by high winds.

[Terry Casey]

We have an IR detector light outside our front door. The detector is angled so that it is not triggered by anybody walking by and the light is angled so that it only illuminates our front garden.

But ...

A week or so ago I walked out of the house via the back door, as usual, down the drive and into the street. As I reached the far boundary of our property, the light turned on! (The light and detector focus on the area I'd already walked past, essentially behind me!)

I looked back at the garden but no sight of a cat or squirrel, so carried on. When I returned, I walked past the house slowly and right next to the front wall. The light stayed off.

This was the beginning of the last spell of high winds, mainly blowing directly at the front of the house, and there were further instances of the light turning on unexpectedly. Had the wind altered the position of the detector? I wondered.

Then the winds abated and the light has been working normally.

Today, the wind returned and the light started misbehaving again. I went out early evening, leaving by the back door as usual and walking down the drive. The light was already on. I walked down the street and it stayed on.

A couple of hours later I returned. In the meantime there had been a heavy downpour and the wind had dropped. I approached the house - no light - walked close to the wall as I passed the house - no light - walked down the drive and in the back door as usual - no light.

Has anybody got any ideas why an IR detector should play up in the presence of high winds and then work normally when they drop?

[ss]

Mine can sometimes play up from the exhaust from the gas boiler depending on which way the wind blows, I also have an extractor fan in a downstairs toilet which if it is cold can set the PIR off when the toilet is used when the extractor fan kicks out hot air from the toilet.

[Fredxx]

<snip>

In high winds most surfaces would become uniform in temperature. Your coat would quickly become the same the wind temperature.

IR sensors rely on seeing a low frequency temperature variation through their multiple lenses. If anything is a uniform temperature the sensor will not provide any output.

[Thomas Prufer]

... specifically, the field of view is separated into lines or spots by the lenses. A warm thing moving through the field of view is, as it moves, focused on the sensor, then not, then focused again, etc. This results in a modulation of the sensor signal, the frequency depending on the speed and spacing of the lenses. The sensor is pretty sensitive, to detect a 29° C cat moving across a

25° C background.

Wind alone would not trigger the sensor, if it is all the same temperature (unless it moves, say, branches, which shade and unshade a warm wall, a streetlight,

But wind moving a plume of hot air, or itself swirling with subtly different temperatures could give a similar signal as a warm thing moving across the field of view, and trigger the sensor.

Thomas Prufer

[Brian Reay]

We had a similar problem in a previous house. It was a s*d to diagnose as there was a gate involved. If the gate was closed, no problem. No significant wind, no problem. Warm outside, no problem. Eventually I noticed you needed all three- gate open, wind, cold outside and when the boiler fired up after dark, the security light came on.

As the gate was intended to be closed the fix was, at least, easy.

[Chris Hogg]

<snipped stuff about an erratic PIR detector>

I had a similar problem at a previous property. There was a rather open and scruffy hedge a few feet in front of the detector, and as the wind blew (which it did a lot there; on the coast and very exposed), the hedge thrashed around and triggered the detector. Removing the hedge solved the problem.

[Brian Gaff (Sofa 2)]

The problem is that as far as I am aware PIR devices detect movement by objects of differing IR emission changing between segments of the sensor. So warm air if it moves about in front of it can set it off. The same applies to inanimate objects if they are different temperatures and moving. In a wind such changes are far more likely to occur as has been mentioned on here before.

All you can do is adjust sensitivity really, there is no magic solution since there are so many reasons for differing temperatures. Brian

[Brian Howie]

Mine is positioned above the back door. If I leave it open, the hot air coming out of the house can trigger it. Perhaps it's warm air currents that's doing it.

Brian

[Brian Gaff (Sofa 2)]

No the lack of output idea does not roll. I have tested this on portable alarms. Its the movement of differing temperature things that sets them off, but many have some kind of logic to try and stop the issue. I remember one particular occasion where a small shrub was partly across a window. The pir could see the window from its position. As long as there was no wind, all was well as the window was warmer but always the same place, introduce a wind though, and the evergreen shrub swayed in front of the window making the false trigger occur. Re aim the sensor to not see the shrub and all was well again. Brian

[Andy Burns]

splintered front lens? wind annoys the spider hiding inside?

[alan_m]

Hot air on its own has not much of a heat signature for the PIR. The hot air would need particulates in it to be detected hence detecting the exhaust from a gas boiler etc. In a high wind condition perhaps more dust is being picked up or the wind chill effect on different surfaces of protrusions provides the trigger.

[alan_m]

Positioning PIR detectors can be more of a PITA if they are used for (indoor) bugler alarm detectors. When I fitted a PIR to detect entrance into the house via the front door I thought it a good idea to place it facing the door. Anyone entering the front door would trigger it and I would then have xx seconds to enter the code so that the alarm wouldn't sound. Unfortunately what also triggered it was anything being posted through the letter box, especially on a cold day when the post was cold and the inside of the house was centrally heated.

I had to reposition the PIR sensor to above the front door so the first detection happened when anyone was around half way along the hall-way or just climbing the stairs.

Similar problems can occur if placing the PIR detector facing a window where anyone walking in front of the window can trigger it - such as the postman taking a shortcut across your driveway to also deliver to your neighbour.

[David]

I also have had problems with an internal PIR triggering from movement outside a closed window. But other people have told me that's impossible. What's the actual mechanism?

[Martin Brown]

They are designed to ignore stationary warm objects but if the assembly is moving in the breeze and there is something warm in its field of view then its differential detector will see an apparent movement even though it is the detector swaying in the breeze. It may only take quite a modest deformation to trigger it. You can sneak up on a PIR if you move towards it slowly enough and/or wear aluminised clothes.

The other one to check is that there are no spiders hiding inside the sensor trying to stay warm and dry at this time of year. They are by far the most common cause of erratic PIR detectors misbehaving.

[charles]

It may be impossible, but it happens. This is why some manufacturers tell you to mount them so they can't see out of a window, That's how all mine are mounted.

[Chris J Dixon]

Does it warn you if they are about to play Reveille?

Chris

[Jeff Layman]

In particular, the second and third paragraphs.

[Tricky Dicky]

We have a PIR on the garage facing a window and it has never been triggered by movement inside the house. In fact it has a remote control which can be used to override the settings or to alter the settings and even that will not work through the window.

Richard

[Jeff Layman]

Really? I just tried the IR remote for the TV in the kitchen through the double-glazed patio door into the conservatory. Absolutely no problem changing the volume or channels.

[alan_m]

It depends on the IR band

Remote controls are in the near IR band up to 1.1 um and can been seen through a glass lens of a digital camera.

PIR detectors operate in the 8 to 14 um IR band. Most glass is usually opaque in this band. However the sun shining through the window can heat the net curtains and passing clouds cause a change in this temperature triggering the PIR.

Once you go outside of the visible band it may be hard to predict what may be causing a PIR to trigger. I used to work with and design military grade IR cameras. A matt varnish on a wood surface could act as a perfect mirror in IR and items such as black plastic rubbish sacks are completely transparent.