New design for sensor and actuator network

[stuartiannaylor]

Don't know if anyone is interested but here goes.
If anyone has been involved in BMS, Home Automation or Lighting control you will know there are a plethora of standards.
I have a design that I have been playing with for ages that I am interested in providing as an Open Source design spec.
I have been working on the premise of the 50 pence network node which utilises a low-fi powerline carrier.
There is nothing out there that I can find that is suitable and it sort of sits above x-10 but doesn't bother with any of the full TCP/IP OS model.
Its main premise is to remove the need for control wiring in a local subnet in a similar way that Lin can sit on a CAN bus and CAn can (image there) sit on an IP bus.
The idea is to make network bridges from a simplistic net work has a lot in common with 1-Wire but can carry power (serveral amps not milli amps).
This means you could run sensors and actuators in a very similar scenario to 12Vac halogens (actuators) on a 64 node sub net. This then bridges to ZigBee or HomePlug to allow its to scale to a full IP network if you should want.

Anyone any interest ?

Stuart

[wookey]

Yes - that sounds very interesting indeed. So if I understand correctly you propose to use a low-voltage network, but modulate the bus signals over the power lines?

Is this currently a design rather than any actual hardware? And do you have docs we can look at?
Of course your solution to the plethora of standards is yet another. Open is good (very good) of course, but is there really much chance of getting any traction in this area? It's going to have to be very good _and_ cheap enough to make an impression.

[Dyslexicbloke]

I was just batting on about this sort of thing on another thread ... should have read this first I gues, as I was advocating essentally the same thing.

Pure coincidence not plagerisme I promise ....

For the record I think it a great idea, high power and relativly low speed is almost opersit to all current development that I am aware of.
You could end up with a simple to deploy robust network requireing nothing more complex than a few lengths of twin and earth and a screwdriver.

I havnt been able to find anything that will do this for me, in fact I was considdering playing with some small PIC's to see if I could design and build a protacol from scratch, but I'd rather work to an existing standard if yiou have one.

Were you thinking Master/Master or Master/Slave .... I had a few ideas about device priority multiplexing for handshaking if your interested.

Alistair

[stuartiannaylor]

Well strangely enough its is in a way very similar to the Moixa design but it works on 12V ac and has exactly the same installation layout as Hallogen lighting.

Zigbee and Homeplug nodes put a huge cost on any slave node. I was working on the fact that a network bridge acting like a standard electronic transformer would require zero changes to installation and wiring regulation.
If you have a SMPS then you have a major advantage over Homeplug as you have an active powerline of a known power that you can clean. This means very simple powerline comms can be produced and most of the cost is in the Bridge / SMPS design allowing very low cost slave nodes to be produced.

I was trying to aim at 50p microcontroller slave nodes with a High end DSP microcontroller providing power and an assynchronous comms. Most traffic is from master to slaves so cost and communication is mainly in the bridge unit. This just looks like an electronic transformer. The system was aimed at 12V ac / 12V dc and limited to 150 watts to provide a small area sub bus. This reduces protocol complexity and deals with voltage drop of long cable runs and provides a slave node count of approx 50 which works well in scope.

Its main application is LED lighting but any complex accutator / sensor network could be accomplished.

Task lighting and intelligent control can save up to a further 70% of your lighting energy bills the main reason this is not widespread is basically cost. I have had commercial interest but everytime I have tried to push the Open Source idea the conversations have fallen flat. In fact it has been fairly frustrating and I am beginning to lose interest after 5 years.

[fred bloggs]

Members

There is a low voltage power line signalling system which has been around for 23 odd years, it is mainly used for fire detection systems , known in the trade as "analogue addressesable". I used to run the test department for a UK owned fire systems manufacturer in the early 90's and we used to make all sorts of specials using the same basic 24V dc analogue loops for security, building control systems etc. we could get 4A at 24V DC out of a loop with a 1000m length of MIC (pyro). I cannot find the data I used to have but you could get small custom processors from the likes of Nittan which formed the basis of all of the detectors and controllers (sounders, lights). We built several systems in test as 'foreigners' to control machinery, heating systems, I could kick myself knowadays as I didn't keep a record of the circuits or software for the stuff we did would have be great to control my heating system at home. Try googling analogue addressable systems.

Fred