Integrating Home Automation & Alarm

I have been thinking lately about improving my approach to home automation, and particularly in making better use of the various technologies I’ve got. I think one of the mistakes made in home automation generally is neglecting the ‘automation’ part. I’ve seen plenty of approaches which really just substitute one set of switches for another (albeit electronic). I can’t see much point in in switching lights on and off using a phone or remote when it’s much easier to do so using the light switch! So really what you want is a largely automatic system which anticipates your needs and does what you want without user intervention.

So far as I have written about before I already have a mixture of systems with MAX for heating (currently controlled using FHEM) and Z-Wave for which I have been using a Vera Lite. I have been using a 3rd party plugin which provided an interface between the Texecom panel and the Vera, but in spite of this I’ve never been that happy with it as a controller. The interface is rather old fashioned, the newer updates (so called UI7) have had a decidedly mixed reception and I’ve never really felt I understood how to program it properly. Whilst the UI looks pretty it was also rather clunky when you tried to use it.

I’ve written before about FHEM and whilst I know it’s not everyone’s cup of tea I have really got to like it because although it looks unappealing it gives you a huge amount of information and lets you do pretty much anything you want with a few command lines. I also think that it helps you understand a lot more about the devices you are using. So the obvious thing to do was to bring all my Z-Wave devices into FHEM and try to integrate heating and lighting and also build integration with the Texecom panel.

The key to all this is a USB Z-Wave adapter for my main server. These don’t come cheap and there is some choice. I eventually bought an Aeotec Z-Stick Gen5, which came recommended but pretty expensive at about £40.

aeotec-z-wave-usb-z-stick-526

 

There are others available, including the Z-Wave.Me UZB which seems rather cheaper. This should also work with FHEM and other software although the Aeotec seems slightly better established. The other and much cheaper option is to use an experimental method whereby CULFW devices can be set up as Z-Wave controllers. However this lacks some of the benefits of the proper controllers (including battery back up etc) and probably is not for ‘production’ systems just yet.

So far all has gone well. FHEM recognised the controller and configured it correctly. I had to ‘exclude’ all the devices from the Vera to reset them and then ‘include’ to the Z-Stick. You can do this either by physically taking the stick to the device and pressing the button on the stick, or else simply setting the Z-Stick into ‘inclusion mode’ from FHEM and then doing whatever is needed on the device.

I would recommend doing the latter, as then FHEM will automatically pick up and configure the devices for you. I now have a variety of Fibaro devices and a Minimote which are all available as devices in FHEM with a lot of detail about them. I have recreated a few scenes which I was using before (such as one which turns on a few table lamps in the front room).

Now I need to find a better way of integrating with the alarm. I am using an Arduino with one of the outputs from the panel to report whether the alarm is set or unset and therefore whether the house is occupied or not. I’m using this at the moment to trigger a ‘notify’ event in FHEM which sets the heating into low power mode and now sends Z-Wave events which turns off all the lights. What I would really like is to get FHEM to interface directly with the panel to read out the status of the sensors etc directly. I have been able to get some information about this so I am going to see if I can learn how to do this.

As always I’d be very interested to hear from anyone else who is exploring this kind of thing.

Intelligent Heating Control – First Steps

As promised I’m now turning my attention to a different project. For a long time I have wanted to handle control of my heating system much more intelligently than just relying on a single thermostat. My house is a fairly large Victorian terrace spread over three floors, and the rooms are large with high ceilings. At the moment the thermostat is in the hall, and this is supposed to take account of the heat in all the rooms of the whole house. It seems a very blunt instrument and I’m sure I’m wasting a lot of heat.

This sort of thing is topical and there have been products such as the much-vaunted Nest thermostat. However, I’m really unimpressed with this – because no matter how fancy it is it only samples heat from one place. If you live in a small flat then that’s fine – but it doesn’t help my situation.

As well as having more sampling points, you should also have more control over how the heat is distributed. Nest etc only control whether the boiler / heat source is on or off – not where the heat is going. My house uses standard gas-fired heating with radiators and thermostatic radiator valves (TRVs) but I’ve never been very satisfied with these. Also of course they don’t provide any control of the boiler.

It seems I’m not alone in this – there are a small number of products out there which address this issue, and the most recent and most consumer friendly is the Honeywell EvoHome system.

honeywell_evohome_connected_thermostat_pack_atp921r3100_hr924uk_radiator_controllers

This does look very interesting but I’m not pursuing it for various reasons (mostly because I don’t want to have to start again with a new system). There does seem to be a lot of discussion about it in the AutomatedHome forum

The basic idea is that you have one or more replacement TRVs which are motorised and have a temperature sensor. The valves report back their temperature to a central control point, which is then able to control the boiler to switch on as the need to heat each radiator arises. This feedback is the big plus over a normal TRV (which really only regulate maximum heat) as it allows them to regulate minimum heat too. When heat is called for, only the radiators which need it will be on and so the heat is much more efficiently transferred to where it is needed.

I’ve been looking at these systems for quite a few years now. When I first started about 5 years ago, the only thing out there was the FHT range. This is based on the FS20 protocol, which seems to be very well-established in Europe and particularly Germany. There is an enormous range of FS20 kit available, far more than any equivalent system used in the UK. However, it seems little known about here. The FHT system involved a separate valve motor (FHT 8V) and room thermostat. The motor adjusted the valve depending on the temperature, and a separate boiler interlock (FHT 8W) which listened to the settings of the valve motors, and when they reached a certain level it triggered the boiler. The FHT 8W is a remarkably overengineered piece of kit – I bought one at great expense (about £80) but I don’t think many other people in the UK did.

fht8w_heatrelay_1

I had this system (two or three thermostats and the boiler switch) in my old house which was much smaller, and it did seem to work quite well. I discussed this quite a bit at the time on various forums and blogs (see discussion with Jack Kelly who goes into the whole subject in some detail). However, when I moved house the FHT kit remained very expensive and I rather lost interest in the whole idea as other more pressing work took over.

The other thing which I did get from my first attempt was computer control of the system. I was not satisfied with just the kit, I wanted to monitor and control it much more closely. Once again, I found that the Germans were very far ahead of us. Through reading various forums (thank goodness for Google Translate) I learned about FHEM. This is very mature and well established home automation control system. I’ve been following home automation for years (and I have a developing Z-Wave system for lights etc which I’ll talk about some other time) and am aware of various software systems.

However FHEM stands out for a number of reasons – it is unashamedly geeky and flexible (hence it suits me well), it is extremely well supported (if you speak German) and has support for the widest number of different systems that I’ve ever seen. To get it to talk to the FS20 system I needed a CUL1101 device. At the time I got mine the only place you could get them was from busware which is another German company. It was rather expensive (about 50 Euros although they are more now) but gave me what I needed to interface FHT with FHEM and read out the data to plot some graphs. However I never got much beyond this and I was aware I hadn’t really made the most of it.

This is pretty much where I got up to before I picked it up again recently – which I’ll cover in the next post.

 

Premier Elite v3.00 Firmware

Visiting the Texecom website today I’ve made the exciting discovery that the firmware for the Premier Elite panels has had a major upgrade from v2.11 (which is what mine shipped with) to v3.00

Now I’m a firmware junkie – doesn’t really matter to me whether I want or need the new features, but I just want to be on the latest version. So whether it’s TVs, microwave ovens, dishwashers or guitar amps I’m always interested in getting the newest software release.

Texecom provide a very useful summary of the changes here

From a initial look through, much of it is expansion and improvements rather than major new features but one thing particularly caught my eye. Since I’ve been playing with these panels I’ve thought that I’d really like 3 COM ports – one for the GSM interface (when I finally get it working), one for integration with the Vera smart home controller (more of this in another post) and one for a permanent COMIP connection. This is apparently available on the most expensive panel but I didn’t think of that when I bought it.

Anyway it seems this new firmware allows COM3 to be accessed via the existing communication port – confirming my suspicion all along that it was just a COM port in disguise. It will require a breakout board which no doubt one will have to buy from Texecom:

comport

Although judging by this picture it will simply connect out a few pins – so maybe it will be possible to identify which pin is which and connect directly to it and I am worried that this new board will be expensive.

The other problem is that one cannot simply flash the new firmware using a COMIP or USB-COM which is a bit of a pain. The firmware is held in flash ROM, but they seem to have left the programming chip off the main board and so one has to buy the ‘CDH-001 Firmware Flasher’ which is going to be about £30 from the usual sources

Premier Elite Flasher Interface

Looking at this picture really makes me unwilling to pay £30 for the privilege – it is clearly just a MAX662A flash programming chip plus a couple of capacitors, a switch and a button which can’t be worth more than a few quid. However without knowing the pinout of the programming port and without being able to study one of these I’m a bit stuck. I’d love to be able to DIY one of these… but I suppose I may have to grit my teeth and stump up if I want to get the upgrade. I suppose if you are an installer you make your money back after the first job but for me the ‘cost per use’ is very high. If only they showed a photo of the underside!

To make matters worse the above adapter still needs the USB-COM to work, but at least I already have one of those.

If I do get one I’ll analyse it and post the results… or I could just stay at v2.11 but I’m not sure I can tolerate that now I know I’m out of date.

In other news I’ve now got all the bits from China so will try again with my ComGSM DIY effort and will post the results.