Which automation system

[Otto-Mate]

Interesting thread guys, this has reminded me we need a list of HA technologies on Automated Home and I've been working on a new article to summarise these.

Thanks

M.

[b33k34]

I have a Loxone mini server here that I have been evaluating for a few weeks. The initial interest came after an enquiry about replacing a HomeVision system (now discontinued) + xAP gateway with something similar in terms of reliability, I/O , programming / ease of use plus would be easy to retrofit. I was so intrigued by the potential of this product that I sourced one myself to see if I could include xAP support for Loxone within the xAP Gateway and potential C-Bus integration.

From a quality viewpoint both the hardware and the software are top notch.

I agree that the mini server itself is attractively priced although some of the Extension modules (if required) seem quite expensive and also offer limited functionality sometimes. A very capable product yet the implementation / programming is still within the capabilities of HA enthusiasts. Importantly , even if you get an integrator to set up the system it can be easily be tweaked as your needs change - something not possible with the well know higher end end $$$ systems as the programming environment is not available to the user.

Another few days of research and I'm no further.

The more I look into Loxone the more I'm impressed and a miniserver plus 1 extension is only £800. That would give 20 digital inputs, 8 analogue inputs, 16 digital outputs and 8 analogue outputs. I reckon I could do all the lighting I want to with that as you can use the analogues to drive LED dimmers (I'd only want to switch on and off at fixed brightnesses as part of lighting scenes - presume that's possible)

Velbus looks would be close to £2k that to get 20 switched outputs plus software automation (about the same whether the Velbus hardware or basic Comfort non-alarm box). Loxone looks more powerful as well - Comfort looks like very old tech.

However, the big off-putting factor is the centralised and integrated nature of Loxone - it's a single point of failure. If my Loxone stopped working on Xmas eve I'd still be without any working lights. With further integration potentially without heating or access control. This is where their website analogy with a car fails. If your car breaks down you can quickly arrange another form of transport or a rental car. It's not so easy or cheap (or practical) to find somewhere else to live.

Integration into large units starts to look a bad idea from a maintenance point of view as well - if a single relay fails on the server unit presumably I'd need to buy a whole replacement unit (can't be without the control while it's repaired, if it's repairable). It would be easy enough to keep a spare relay unit for Velbus if they started failing. Not so cost effective to keep a spare server and extension for Loxone.

I suppose the other question is how easy it would be to move away from a Loxone install if they're not around when the server unit fails. I suppose I could run conventional multicore 240V cable from each switch to Zone0 to allow complete removal but that doesn't help in the short term when a module fails.

Comments or advice welcomed. Much as the heart likes the idea the head it telling me no at the moment.

Criteria end up being -

1) must be DIY configurable.
2) price has to represent value. In an age of £30 Raspberry Pi and £270 iPad mini bits of proprietary kit costing >£1000 don't look realistic.
3) must have some resilience

It's a new build so I'm starting from scratch - Ground source/UF heating, MVHR ventilation, Entryphone, probable alarm so theres a fair bit of stuff that could talk to each other.

[bailers]

Another few days of research and I'm no further.

The more I look into Loxone the more I'm impressed and a miniserver plus 1 extension is only £800. That would give 20 digital inputs, 8 analogue inputs, 16 digital outputs and 8 analogue outputs. I reckon I could do all the lighting I want to with that as you can use the analogues to drive LED dimmers (I'd only want to switch on and off at fixed brightnesses as part of lighting scenes - presume that's possible)

Velbus looks would be close to £2k that to get 20 switched outputs plus software automation (about the same whether the Velbus hardware or basic Comfort non-alarm box). Loxone looks more powerful as well - Comfort looks like very old tech.

However, the big off-putting factor is the centralised and integrated nature of Loxone - it's a single point of failure. If my Loxone stopped working on Xmas eve I'd still be without any working lights. With further integration potentially without heating or access control. This is where their website analogy with a car fails. If your car breaks down you can quickly arrange another form of transport or a rental car. It's not so easy or cheap (or practical) to find somewhere else to live.

Integration into large units starts to look a bad idea from a maintenance point of view as well - if a single relay fails on the server unit presumably I'd need to buy a whole replacement unit (can't be without the control while it's repaired, if it's repairable). It would be easy enough to keep a spare relay unit for Velbus if they started failing. Not so cost effective to keep a spare server and extension for Loxone.

I suppose the other question is how easy it would be to move away from a Loxone install if they're not around when the server unit fails. I suppose I could run conventional multicore 240V cable from each switch to Zone0 to allow complete removal but that doesn't help in the short term when a module fails.

Comments or advice welcomed. Much as the heart likes the idea the head it telling me no at the moment.

Criteria end up being -

1) must be DIY configurable.
2) price has to represent value. In an age of £30 Raspberry Pi and £270 iPad mini bits of proprietary kit costing >£1000 don't look realistic.
3) must have some resilience

It's a new build so I'm starting from scratch - Ground source/UF heating, MVHR ventilation, Entryphone, probable alarm so theres a fair bit of stuff that could talk to each other.

Just to add to your Loxone considerations you can use the 0-10v analogue outputs to control dim levels by using an appropriate dimmer module with 0-10v control. I have not done this yet however am looking to go down that road once I find a suitable din rail mounted module than represents better value than the Loxone dimmer extension which in my opinion is rather pricey for 4 outputs. I am really happy with my miniserver though, the possibilities are endless with the config software.

In terms of reliability I have had no issues after 6 months of use and have seen no evidence of hardware failure on their forum. I do however understand your worries as it is rather expensive to carry a backup spare in case of failure.

[b33k34]

Just to add to your Loxone considerations you can use the 0-10v analogue outputs to control dim levels by using an appropriate dimmer module with 0-10v control. I have not done this yet however am looking to go down that road once I find a suitable din rail mounted module than represents better value than the Loxone dimmer extension which in my opinion is rather pricey for 4 outputs.

The dimmer extension does look expensive. I don't think I'd need one though - it's a new build and I'm planning for all of my lighting, except perhaps some legacy desk lamps, to be LED. As far as I can work out you can dim the Loxone LED driver (which looks like a rebadged bit of kit from someone else so suspect there are alternatives) directly from the Analogue outputs of the miniserver or extensions. Given that the extension is cheaper than the dimmer and has 8 digital outputs as well as the 4 analogue it seems a no-brainer.

The DMX extension possibly makes even more sense - up to 128 dimmable circuits off the back of that (suddenly it's inputs that become more of a limitation than outputs and I was planning to reduce the number of switches per room through the use of scenes anyway).

As you can tell, having slept on it I'm coming back round.

[b33k34]

A quick google brings up this 8 channel low voltage dimmer (ie for LEDs or incandescents, but presumably requiring a seperate power supply - for £175 and this single channel for £48 (the Loxone one has 3 outputs but they're all dimmed to the same level.

If using the DMX extension then these dimmers from AliExpress (ie direct from China) are a third of the price of the Loxone branded one - about £25 each (potentially plus duty and VAT)

I'm pretty sure I read that the LED controller should be situated near the LEDs which makes the DMX connections rather more complex and means you'd need a separate local 24V power supply.

[Kevin]

b33k34 - Your thoughts about single point of failure are interesting and something often overlooked by HA users. Actually ignored or tolerated is probably a better word as failures are rare. Most centralised control systems introduce this problem although the cost to swap out defective modules will vary based on the complexity of the units and the number of integrated channels. With larger installs then spare units are less of an overall cost. With Loxone you might be able to implement a redundant server approach fairly easily, maybe even with auto fall over using dual mini servers - with the associated extra costs. One other strategy would be to divide lighting in rooms across the mini server and the Extension(s) so that if one Extension failed then you still have some lighting available. BTW I have no experience yet on what happens if an Extension fails on a system, I'm assuming the remainder stays functional.

By using additional mains wiring , switches and relays you can provide a fallback to manual operation , especially if you are using relay rather than dimmer channels. At a basic level this could be just some switches that bypass relays (placed near the Loxone system) to allow manual light control. A more flexible approach would be to bring the switched live back from a traditional wall switch as well as a three core feed to the light. This can actually be done with just one 4 core star wire back from existing lights eg ceiling rose. Then you can use each switched live to switch a bank of individual DIN mount changeover relays to provide the dry contact input to the Loxone system. You can get a lot of these ultra slim relays on one DIN rail and they are around £7 each. Now should you need to revert to manual you can just bridge the switched live back to the relay output and your wall switches will work as before. I have used this in a slightly different way (low voltage) on some of my lights , but this was because I had retrofit problems. It will also help in removing the HA quickly should I move home. In other areas I have C-Bus and some IDRANet, so I do not have the ability to override easily if they fail although it would tend to be more localised due to the distributed nature of the control.

So it is possible to provide this fallback with not a significant materials cost, but labour in a new install would be more...

The Loxone dimmer does seem relatively expensive although it is well specified hardware supporting over 1KW and most loads interchangeably (leading/trailing) and CFL's. It also has 8 inputs for switches. My dissapointment was really the software implementation for ramping rates. It appears the dimmer , something I don't own BTW, doesn't implement these internally instead relying on the mini server to send successive level updates timed to simulate a ramp. Incorporating ramp rates within a scene doesn't appear too easy, and advanced aspects like coincident arrival at a level not handled at all. Interestingly when I originally specified my lighting I over estimated dimmer channels by almost double. I instead should have used relays as b33k34 - catchy name BTW - has also decided. With the demise of incandescents this is even more practical.

K

[toscal]

If in ten years Loxone are no longer here, the fact its a centralised system would mean replacing it with something else should be quite easy. In fact you would have more options.
If you are worried about reliability then I would set up a mission critical list for lighting, these are lights that need to work no matter what. I am sure you could use some extra mains relays and a couple of 2 way switches for these, so at least if the unit fails the switch in the room would still work. You would need to use another relay that turns on when the lights do so that the miniserver knows a light is on. I think this is what Kevin is suggesting.
We are currently designing a project for some garden lighting using all LED lights. The flexistrips will be using our 0-10V 3 ch dimming transformers (each channel can be controlled separately). The Loxone one actually uses a single 0-10V signal to control the 3 channels so you only get about 8 colours. This means that 1 to 2v gives red, 2.125 to 3 gives yellow etc. The LED spots will be controlled via the relay outputs. As will the LED flood lights and deck lights.
I would recommend you download the config software for the miniserver as you can play with this and then check it with the simulator mode. so you don't need the miniserver to test your config.

[b33k34]

Loxone depends on momentary switches though - how would this work with 'standard' switches? Wouldn't you need to swap the momentary switches for standard if there was a failure? I'm starting to think this is less of an issue -as an alternative to Kevin's suggestion of switching High Voltage and using relays to control Loxone with everything wired to a central location (but only CAT5 to the switches) the alternative would be to remove the automation kit by -
- replacing the momentary switches with conventional switches
- replacing the Loxone lighting kit with a bank of low cost relays
- running a 12V or 24V supply to operate the switching side

There might be one or two edge cases but for most of the lighting that should work.

Toscal - using 3 analogue outputs to control one section of lighting sounds complex. The DMX controller looks like it adds a lot more power and flexibility if you're doing this with lighting in more than a single location.

I did pick up today that (obviously when you think about it) more than one switch can be connected to a single Loxone digital input - ie replicating switches on either side of a room only uses one set of inputs. That frees up inputs for use by PIRs and other sensors (I was thinking that by using a DMX controller inputs would become the factor driving the need for further extension modules rather than outputs).

[toscal]

If you are using an RGB strip then having being able to control each channel separately makes more sense (well to us anyway), we have often found that some clients prefer being able to fine tune the shade of blue or purple to suit their needs. Also if you have 3 fully controllable channels you can have 3 controllable lighting zones. Which is kind of what we will be doing, so one of our 3 channels dimmers will be controlling 3 separate warm white strips. Not too sure how you would do this with one of the Loxone dimmers though, by using the analogue . OK you can go the DMX route, but at 269 euros for the DMX module, and then another 75 for the PWM dimmer just to control 3 channels seems a bit of over kill to us.
Forgot to mention that the Loxone dimmers need two 0-10V inputs one for colour as I mentioned above and the other is for overall brightness.

As far as I can tell you only need momentary switches if you want to do more than just on and off from the switch. Such as long press dims up or down or you could have a multi press set up,so one press is scene one, 2 quick presses scene 2 etc. And to get the dimming functions you would also need the dimming extension module if you are not using the PWM dimmers etc.



If you haven't already, done so, download the config software and have a play. Also watch the tutorial videos, they are very good starting points.


Otto- like the idea of an article to summarise what's available as regards HA tech.

[Wexfordman]

Loxone depends on momentary switches though - how would this work with 'standard' switches? Wouldn't you need to swap the momentary switches for standard if there was a failure? I'm starting to think this is less of an issue -as an alternative to Kevin's suggestion of switching High Voltage and using relays to control Loxone with everything wired to a central location (but only CAT5 to the switches) the alternative would be to remove the automation kit by -
- replacing the momentary switches with conventional switches
- replacing the Loxone lighting kit with a bank of low cost relays
- running a 12V or 24V supply to operate the switching side

There might be one or two edge cases but for most of the lighting that should work.

Toscal - using 3 analogue outputs to control one section of lighting sounds complex. The DMX controller looks like it adds a lot more power and flexibility if you're doing this with lighting in more than a single location.

I did pick up today that (obviously when you think about it) more than one switch can be connected to a single Loxone digital input - ie replicating switches on either side of a room only uses one set of inputs. That frees up inputs for use by PIRs and other sensors (I was thinking that by using a DMX controller inputs would become the factor driving the need for further extension modules rather than outputs).

On the centralised sytem, a similar comparison perhaps to a similar scenario I had a few years aog ith a home audio system might be worht considering. I initially installed a multiroom audio system in my house which was a centralised system. It consisted of a 6 zone 6 source system with distrubuted sepakers and keypads in each room. Had it about 4 years, and a bad lightning storm one night blew it. I was then left with no audio system, it was not repairable, and a replacement was going to cost me about 9k euro. In fact it was no bad thing, becaue I went for a distributed sonos system, which while also a much superior system, it was also made up of indivudual room components, not reliant on any single unit, if one failled, then my system would still be okay, and cost me a few hundred euro to replace, as oppossed to a few thousand!

I would consider a similar scenario with a home automation system controlling lights etc, except light and heat being a bit more critical. I would be cautious about any device which is a single point of failure.

With regards the topology of your cabling system, I htik the wiring structure is similar, whether it be velbus, loxone, cbus, idratek (i may be incorrect on this), in that thay also work of the same principle of cat5/twisted pair to light switches, and live/nuetral et wired ack from load to relay unit in the mcb.

When I was rewiring my house, I considered whther or not to leave 220v wiring in my light switches or remove it completely. I decided in the end to remove it. Part of the reason was that I figured even if velbus did not work out long term for me, there were lots of similar systems that I could migrate to whcih had similar wiring topology.

So, I suppose I am sying from my point of view, I can still go for an alternative system if needs be from a wiring perspective, I could go loxone, cbus,idratek etc, and I think this should be an important consideration for you, but also a comfort with respect to relying on any one system. There are other options if you decide to move to something else in the future. Again, I am saying this with only a brief knowledge of the other systems, but my understanding is the wiring systems are compatible.

Going back to single point of failure, with regards my own scenario I simply dont have one. Each of my modules, work stand alone, there is no central controler which is critical for operation of lights and heating. My system will work as is even if my HA controller (comfort) dies a death. My system is capable of being controlled bby a number of different controllers from a number of different vendors, including comfort, velbus home server, linux systems etc, so I am not reliant on a single vendor. One velbus module failure will at worst, knock out 4 circuits, no more! (the 12v power supply is the only component at risk, but I have two of these installed and they are a standard component you can get from any electrical store anyway).


As someone on these forums said before, we are all a bit biased towards our own particular systems, and I certainly am too, so bear this in mind, my advise is definately biased towards the systems I use.


Ignoring velbus however, I think you might find many here who are very very pro comfort, and for good reason. It is an exceptionally relaiable and versatile system. I am now 13 years using comfort in my home and have never had a hardware failure (I did upgrade my system a few years back though). And again, the one peice of logic I just dont think can be beaten, is the combination of security and automation into one device, meaning easily shared sensors and integration between HA and security. I also dont think comfort could be considered dated in anyway, it is continuously improving both in hardware and software features etc. Dated, it certianly is not!