Reed Switches & PIRs

[chris_j_hunter]

the Positeks look interesting ... any idea of price ?

they offer various models, with range to be specified when ordering ... - they don't look cheap !

[Paul_B]

I had a thought about this in the early hours of this morning (don't ask!).

Would it be be possible to use the door handles as a trigger when someone touches them. I'm thinking along the lines of how a touch-sensitive light works with change in capacitance to light-up?

The beauty of opening a closed door which is monitored by a reed switch is the light is on before you eneter the room. If you are using just a PIR then you need to enter the room before the light comes on. The downside of the PIR and reed switch combination is an issue with PIRs after a door is closed. So Cortex caters for potential false positives by delaying presence detection for 5 seconds (configurable) after a door is closed.

Could you use the touch on a door handle to turn the light on in the opposite room and assist in the presence detection? Of course ultimately it would be good if you could isolate the door handles either side of the door to feed the room presence.

Only issue I can currently see is how would you connect the capacity type sensor to IdraNET taking account of door movement?

Paul

[toscal]

Nice idea. What about a door that is ajar and you don't touch the door handle to open it.

[Paul_B]

I thought about that but I couldn't think of solution unless you wanted to paint your doors with a substance that you use to measure capacitance. I guess the use of touch sensitive handles could possibly address most users but you still need to account for people not using the handle and just the door itself. In this scenario your back to PIR or infra-red beams

Paul

[toscal]

If the rooms are carpeted why not use pressure sensor mats underneath. Something like these http://www.maplin.co.uk/Search.aspx?...0mat&source=15
They are quite cheap too.

[Karam]

Well I suppose one the more interesting facets of an integrated system is the ability to fuse together various sources of information, so more sources can perhaps improve accuracy, though how this information is managed is of course the tricky bit ...

Anyhow, regarding the speed of reaction of PIR only based detection - it depends how you place the PIRs. If a PIR is mounted close to a doorway and facing away from the door (or perhaps even above the door), then when the door is opened it can cause the PIR to trigger due to the sudden fluctuation of air temperatures in the vicinity of the sensor or perhaps even moving IR reflections/occlusions by the door body. Even if this doesn't happen, then with a PIR close to the door the person will trigger the PIR pretty soon after their body crosses the threshold. So reaction can be improved by PIR positioning. Of course one can also have multiple PIRs in a room.

[Paul_B]

Totally agree Karam. My most difficult room is a small ensuite. Without getting too base I've realised I probably need another PIR. For a male, movement is in one direction, whilst a female it is in the opposite direction (think about it guys ;o) My HA is aimed at saving me energy so my thresholds are aggresive, 10 seconds is the default lights after presence removal. Bearing in mind the PIR grace period of 5 seconds, I have potentially 5 seconds for the PIR to detect movement.

A PIR above the door still requires someone to enter the room before the light comes on, whereas opening a closed door with a reed switch the light is already on before entering. This may seem a minor point but to me it makes a huge difference to the experience of HA

Paul

[Karam]

No I think you're right Paul - what may seem like minor lags on paper of a few tenths of seconds can make quite a difference to the experience. Its a bit like you don't realise quite how sensitive we are to even sub degree temperature variations until you start monitoring and correlating with the data.