PIR sensors, weatherproofing

ZOR

Senior Member
I bought some small PIR pcb's a little while ago on Ebay. I want to make something to sense/scare pidgeons off my one and only plum tree. Every year they go on the tree and strip the branches of leaves.

Apart from cutting a hole the best I can in a plastic box and putting sealant around the freshnel? lense, has anyone seen any weatherproof equivelent devices that are small, as I want to use 3 in an arc to gain more field of sensing. OR if someone has seen a matching box cover for my units. I will try and add a picture of them bit later.

Thanks
 

hippy

Technical Support
Staff member
It might be worth taking a trip down to a garden centre or DIY store to see how outdoor PIR's are constructed, and you may be able to use those or modify off the shelf units to hold your own PCB's.

As best I can tell there is often little environmental protection for such units other than to keep rain from going into them though mains switching circuits seem to be better environmentally protected as would be expected.

Perhaps the best thing is to simply try it, see what outdoor exposure does to the assembly. Apply sealant around joins, put a lid on top to deflect rain and better protect any top seams, and see what happens.
 

BeanieBots

Moderator
The sealant as suggested sounds like a good option.
I have seen 'clingfilm aka saranwrap' used to good effect on other units. As the film is very thin it will probably have little effect on the PIR sensitivity.

Out of curiousity, once detected, what is the next step.
(I have a similar problem, tin foils sails is my current deterant which works quite well but strong winds destroy them).
 

neiltechspec

Senior Member
I have a few of those PIR modules, but only indoors in damp environments.
I used silicone sealant on the plastic boxes they are fitted in, still ok after over 12 months now.

Not sure about the UV stability of the 'lens' though for your application, they may 'rot' in a few months.
As they are so cheap, it's worth a go though, so seal them with silicone.

Neil.
 

ZOR

Senior Member
Many thanks for all the replies. I will look at all the suggestions. Did not think the units would work with anything over them (clingfilm) but guess as so thin. I did consider trying to find more commercial units designed for outside however want to stay away from mains, and use low voltages for a picaxe How to frighten them will be the task. I keep reading decoy birds (which I found did not work) don't work, so I might look at using a low voltage fire bell, removing the bell and making it bang on the inside of the case which will be on the tree. I also thought a small horn speaker emitting loud random geiger counter type clicks. So many ideas out there, frightening faces with LED eyes, water jets etc

At least when the basic bits work to sense the critters are there I can become an instigator get rid of them.
This can be a learning curve for the wretched Heron after the pond fish, and the starlings who wait until I cannot get out at the right time to eat/clear my small grape vine.

See how it goes, thanks again
 

The bear

Senior Member
@ ZOR,
You could try a rubber grommet or an 'O' ring to seal the lens.
Old CD discs hanging on nylon fishing line works for a while.
The secret is, to keep changing your scare devices.
Regards, Bear..
 

ZOR

Senior Member
Are you a real bear? Do you hire yourself out for garden duty? Thanks, the O rings a good idea, never thought of that one. Regards
 

The bear

Senior Member
@ ZOR,
Grizzly at times, I do my share of pest control.
Regards, Bear.

"When the lion feeds, something has to die."
 

geoff07

Senior Member
I have had an 'indoor' housed pir outside for a couple of years now. It still works and looks ok with no special treatment. Stuff intended for outside (floodlights with pirs for example, or those little led garden lights) can be so badly done that they look protected but still fill with water. Perhaps the most important thing is a small hole in the bottom. If you can put a little rain-proof housing over it then it will probably last ages.
 

ZOR

Senior Member
Thanks Geoff, I will have a go putting something together, don't know until I try. Thanks Bear, my life seems to attract pests.
 

Dippy

Moderator
A little protective 'hoody' would help I'm sure and would reduce sunlight on sensor. I hadn't thought of cling-film but a thin polythene bag may be an alternative , although anything that is in the 'beam' will affect performance.

Using o-rings , gaskets etc is all very well as long as there is something behind pressing on to the compliant material.

Looking at ready-made external housings may be useful but keep in mind that most cheapos will only be IP44 and should really be kept away from direct sploshing i.e. ideally mounted under shelter from heavy rain, for example, under the eaves of your house.

You'll probably find silicone sealer and polythene sticky-tape will work for a while - remebering the fresnel (no 'h') is made from polythene so actual adhesion is difficult.


Have you actually experimented to see what detection range you get with pigeons on a warm day?
I agree that protecting your plums is very important. My plums suffer more from wasps and other similar menaces.
 

AllyCat

Senior Member
Hi,

... remebering the fresnel (no 'h') is made from polythene so actual adhesion is difficult.
Yes, certainly you need to be very careful about the choice of a "transparent" material to use in front of PIRs. Window glass, for example, is completely opaque for PIRs, in practice.

However, a quick Google found this link and the final two paragraphs are worth copying to here:

An IR transparent weatherproof shield or window can be made of polyethylene which is transparent to IR radiation in the 5 ~ 15 micrometer wavelength range. Polyethylene can be identified by burning a small piece. It will burn with a mostly blue flame and melt like wax while other plastics that are opaque to IR will burn with a smoky red flame and leave a black ash. The only problem with polyethylene is that it is waxy and most adhesives will not bond to it but you can hold it in place with scotch tape, silicone rubber or hot melt glue.

Plastic milk bottles are made of polyethylene, and although it is whitish and not visibly transparent, it is transparent to IR with some small loss in detection range. The polyethylene milk bottle is thin, easily cut with scissors and makes a good IR window. Many food containers and plastic sheeting are also made of polyethylene. A thinner window will have less absorption loss.


So a milk bottle may offer a solution, but I don't know where the author of that webpage is located, nor if milkbottles are made of the same material worldwide.

Cheers, Alan.
 

AllyCat

Senior Member
Hi,

Although I posted the original link, I do have some doubts about the suitability of a plastic milk bottle. Certainly polyethylene is a plastic with "unusual" characteristics, such as an exceptionally low dielectric loss (unlike PVC for example). So it is used as the insulation in ultra-high frequency circuits such as coaxial aerial cable and even special copper-clad PCB material (a nightmare to solder! ). That it may appear "white" to visible light is not significant, but I am concerned about the surface texture of a milk bottle, which might be considered similar to "ground glass", or the type of glass used for bathroom windows.

The fresnel lens is an optical device and may be as much affected as trying to use a telescope through bathroom window glass. Generally the surface roughness of an optical device needs to be "small" compared with the wavlength of the "light"; PIR may have a longer wavlength than visible light, but it's still only ~0.01 mm. Incidentally, although a "corridor PIR" may use a single fresnel lens (to project a long, single "beam"), the majority of PIRs use a compound lens with typically 20 separate beams, arranged as two or three horizontal "fans" for different elevations. The PIR is triggered if a "warm" object enters or leaves any one of the beams.

So it might be necessary to find a smooth-surfaced sheet/box made of polyethylene, or polish the surface(s) of a milk bottle to make a "clear" window. Polyethylene is quite soft so this might be possible using sandpaper or emery cloth and then a "grinding powder", perhaps metal polish or toothpaste?

Cheers, Alan.
 

ZOR

Senior Member
Thanks Alan. I had to wait for the 1 litre milk bottle to empty. I put the sensor inside and it seemed to get triggered ok on a bench test. As Rossko57 suggested, I will try making the latest technology version of a ship in a bottle! I will do some further testing so thanks all
 

ZOR

Senior Member
Thanks Jeff, makes my project/climate look like childs play compared to getting something out of the arctic. Couldn't see any plastic milk bottles though?
 

Dippy

Moderator
Maybe they have a slightly larger budget than you :)

Don't forget that even though people say "transparent" it doesn't mean that it doesn't have a refractive index.
Therefore shape distortions will bend the 'light' to a certain degree - so for a particular range of wavelengths a rough surface is analagous may be like taking a photo through frosted glass.
Controlled distortions are called lenses ; fresnel being a variety used in PIRs and many IR thermometers and, by the way, some microwave systems.

You want minimal distortion, therefore to act like a 'window' you will need it as flat and thin as possible - hence a film or bag as mentioned before. It won't be perfect of course.


Have tried (on a dry day) to see what range the PIR can actually detect our feathered friends?
(i.e. maximum range without a window). If it is only a metre then it may be back to the drawing board.
 

AllyCat

Senior Member
Hi,

Have tried (on a dry day) to see what range the PIR can actually detect our feathered friends?
Yes indeed. The principle of operation of a PIR is to detect (changes in) the heat which "leaks" from animate objects, particularly bare (human) skin (face and hands). Birds in particular are "designed" NOT to leak heat (they're usually covered in feathers).

Furthermore, PIRs are normally designed to avoid "false triggering" by small objects such as birds and animals. However, I'm not sure why my (outdoor light) PIR does often get triggered by animals such as cats and foxes. But I guess that either their "insulated" coat is obstructing the constant "background" heat radiated by the ground (but not the sky), or that they have warm noses!

Cheers, Alan.
 

Dippy

Moderator
Well that's the problem; these PIRs detect radiated heat changes relative to the background. That's why, obviously, the range is greater on a cold day/night and/or with a radiated heat shield (clothes and feathers).

Hence my suggestion to try it without any enclosure that will attenuate the signal.
If that scenario is inadequate under 'best' conditions then we're wasting our time covering it in milk bottles :)
 

SAborn

Senior Member
Personally i find adding lead to the pest works best, its a guaranteed result, with no milk bottle required, or even a picaxe or PIR needed.

I do like most birds, but i have a dislike for pigeons or Rats of the sky as i view them.

Since i reduced the local pigeon populous, i have been overwhelmed with more native birds within my yard, which i am more than happy to have, and enjoy seeing.
 

geoff07

Senior Member
So, given that pirs trigger on ir changes, due to the movement of warm bodies, and are not greatly sensitive at so doing, why do they also trigger on a windy day? I would expect the moving branches etc. that they might see to all be at very similar temperatures.

I have had to install secondary triggering for my outside light to stop the circuit switching all night long when it is windy. The control unit is of course based on a picaxe.
 

JimPerry

Senior Member
Most PIRs use a Fresnel lens to split the signal into several beams - the triggering is caused by change in signal between adjacent beams - move across a unit and it triggers easily - move towards the unit and triggering is not as sensitive
Wind moves across the unit and causes inbalance between received background IR
:cool:
 

geoff07

Senior Member
Correct. But where is the temperature difference coming from when all is at or close to the same ambient temp?
 

AllyCat

Senior Member
Hi Geoff,

Are you sure the PIR sensor itself is in an adequately "air-sealed" enclosure? Perhaps the wind is cooling the sensor (or its case) directly. See the "Infrared window" section near the foot of the link I posted in #12.

Or could the PIR "see" any (cold) sky between the branches?

Cheers, Alan.
 

hippy

Technical Support
Staff member
I would guess that PIR sensors are quite sensitive and likely have some form of AGC (Automatic Gain Control) so a small change from average can trigger a detection. This sensitivity is reduced by requiring multiple or continual changes and detections before it is raised as an actual detection.

It probably depends on exactly which PIR is used and how sensitive it is. It takes practice but it is possible to move slowly in a burglar alarm's PIR field of view without setting it off.
 

ZOR

Senior Member
My IR sensor seems happy either having a window made of a plastic milk bottle, or a clear Pringle plastic lid (crisps tube). Seems to be unaffected by its placement behind however doing tests outside it's not so easy to establish real effects caused by the material in front of the lens.
I am going to send one channel of RF back (433mhz) each time it senses a movement so I can look out of the window each time the buzzer goes to see if a Pigeon is there.

Thanks Jim, I wondered how the lens worked (splitting into bands) I was going to take the lens off at one point so now know it would have messed up things.
Thanks beb101, I did a search and found your unit on the UK Bay site. If all else fails might go down that road.
Wish I knew a bit about radar (doppler), another avenue.
Thanks all
 

Dippy

Moderator
Definitely. And whilst Hippy goes for the ninja approach for breaking & entering I prefer donning my spacesuit for my burglaries :)
Microwave is probably better.

There are dozens of different PIR sensors and circuits inlcuding pulse-counting.
The lens is there for the purpose outlined by jims, hence differnet lens designs for different purposes.

You can buy Microwave or 'dual-tech' devices for better performance.
You can buy bare Microwave transceivers and do your own processing, though it can be time consuming and the pigeons may have died of old age before it's working ;)
It just depends on how much you value your plums.
 

geoff07

Senior Member
Or could the PIR "see" any (cold) sky between the branches?
I think Alan is probably right, though I'm still uncertain, as when it is windy the sky is usually cloudy and thus relatively warm, and the beam angle is pointing at the ground so as to see people. One of life's mysteries.
 

rossko57

Senior Member
Interesting potted PIR history
http://www.kube.ch/downloads/pdf/kube_irs2paper.pdf
with some detail on Fresnels and zoning. Note they can still detect guys in insulating spacesuits - its the contrast that counts, something cold moving against a warm tarmac background will do nicely. Or something at ambient moving against sky background.

Thinks; many fresnels are optimized to detect horizontal movements for people detection. Consider putting such sideways on to detect vertically moving birds?
 

Dippy

Moderator
Well I was having a little joke about spacesuits.

The theory is simple. The PIR sensor + circuit senses changes.
The fresnel lens is a multiple lens which amplifies sections so that a 'hot' object moving across causes high-low (on-off) due to focus. These signals (above a threshold which may be variable or AGC) are converted to pulses. These pulses are often counted to try and reduce false triggers in a security system. Sometimes the number-of-pulses-before-triggering discrimination is adjustable.
These parameters will vary with design and manufacture.

A stationary source with a shield waved will also produce the high-low signal from the sensor; e.g. a branch waving in front of a source or background. After all, if you remove a shield/absorber with a warmer background then the delta is +ve. Some sensor configurations can even give crude direction sensing. For edification I can only suggest buying a basic sensor and having a play.
The COTS devices don't care how the changes are generated they simply respond.

Anyone with a suitable anorak like mine can connect a sensor plus rotating mirror to a 'scope and wave your arms around an watch the scan level change - you can see your hotspots vary with the amplitude of the horizontal 'scope scan. Highly enterntaining if you find yourself at a very loose end and fed up with wall-papering :)

It ain't magic or even a mystery... unless you use the Mystic Meg PIR sensor.
 

beb101

Senior Member
I would be careful with the Chinese sensors. The Parallax sensor has the interface electronics. Otherwise, it's DIY,

http://hackaday.com/2013/08/14/making-the-electronics-for-a-doppler-motion-sensor/

Also, if range is important, this comment from the above link may be important,

"It looks like this is for the HB100, while the Paralax model uses
the GH100. The HB100 has a 20m range, while the GH100 is limited
to 9m. Good catch, though. I wonder if they’re swappable."
 

ZOR

Senior Member
Thanks, still waiting for the sun to rise in China for my query on number of units. Will do a bit more to find whats needed to make them work and then as low cost put an order in. On my normal house security using PIR's, I get lots of false alarms, maybe the doppler will be better in burglar detection. Thanks rossko57, interesting read on lenses, thanks
Is Mystic Meg still around Dippy?
 

AllyCat

Senior Member
Hi,

Personally, I read the ebay listing as being for 3 units, because the same seller offers single units at $3.60 each. But, I also have some concerns about the usefulness of those Chinese units: Although there is a vague reference to 20 metres range, the suggested applications appear to mainly concern detecting quite large lumps of metal (such as a car) at relatively short range. I rather doubt if it could / would detect birds, unless you can persuade them to wear metallised overcoats. :)

However, I'm not convinced that the extra hardware (which appears to be a simple level-slicer) of the Parallax unit would be of much / any benefit. To detect birds, I suspect that you may need to detect the "doppler profile" of flapping wings, for example. That might be a simple analogue high / bandpass filter, or often a micro / DSP-based FFT (Fast Fourier Transform) algorithm would be used. But I doubt if a FFT is within the capabilities of a PICaxe.

Cheers, Alan.
 

ZOR

Senior Member
Thanks Alan,
You are way above me.
Yes, the seller confirmed the deal is for 3 pieces. If I had detailed information to be able to build what electronics was needed and programming code, then I would consider them for repacing my PIR units on outside the house security which keep going off with rain droplets etc.
Maybe I should do what SAborn suggests and get a gun.
 
Top