Smart/WiFi controlled Radiator Valves... self-build controller?

Think about and list what you NEED in the heating control system and then think about what you WANT in the system. Most of what ypu WANT is not essential to having a comfortable home.

I started a project to have individual heating controls for each room, while developing the hardware and software I found the cottage was comfortably warm and efficient in terms of gas used with the very basic controls that were put in as a temporary system.

The more components ( WiFi , routers, internet access etc etc ) that are in the system the more there is to go wrong. And teh harder it is to fault find and repair when it does go wrong.


It would be very interesting Bernard if you could explain what your original plan was for the heating control, and what you eventually ended up doing. (Just interested in the home automation aspect as you evidently have experience in electronics, from your posts.)
 
The initial concept for heating was a temperature sensor in each room on a RS485 multidrop serial link to a custom designed PIC based controller. Each radiator radiator would have a motorised valve operating on 12 volts from the controller. These were standard zone valve bodies with custom built drive units. Supply on one pin energises the motor to rotate the vlave spindle until the valve is open ( detected by reed switch ). Supply on a different pin energises the motor to rotate the spindle until the valve is closed ( a second reed switch ). They can also move the valve to a mid point but experience is that the flow with the valve in mid position is not half the flow. Gate valves would have provided better control of varable flow rates.

Hot water supply is from two hot water cylinders with the same temperature sensors and controlled from the controller. One for the bathroom and the other for kitchen and utillity room.

It was intended that individual rooms would have their own temperatures set by a 7 day program but with manual overides. If I decided to get up earlier than normal then a push button on the bedside table would bring the bedroom heating on immediately. Or if i had to go out un-expectedly a single tap on a button by the front door would put the heating to low.


The systems was designed, all the components assembled and bench tested. It worked, but during that time the reality of the heating requirement for the cottage became apparent. It was comfortable with fewer and smaller radiators than the original heat loss "calculations" had predicted were necessary. Thatched roofs confuse heat loss calculations.

Also maintaining different temperatures in different rooms was not really feasible given the construction of the buillding and restrictions on what changes can be made to a Listed Building. . With these changes much of the custom hardware is not needed. ( but as an engineer it was a good interesting project and I have some very good temperature sensor modules )

More later if interested
 
The initial concept for heating was a temperature sensor in each room on a RS485 multidrop serial link to a custom designed PIC based controller. Each radiator radiator would have a motorised valve operating on 12 volts from the controller. These were standard zone valve bodies with custom built drive units. Supply on one pin energises the motor to rotate the vlave spindle until the valve is open ( detected by reed switch ). Supply on a different pin energises the motor to rotate the spindle until the valve is closed ( a second reed switch ). They can also move the valve to a mid point but experience is that the flow with the valve in mid position is not half the flow. Gate valves would have provided better control of varable flow rates.

Hot water supply is from two hot water cylinders with the same temperature sensors and controlled from the controller. One for the bathroom and the other for kitchen and utillity room.

It was intended that individual rooms would have their own temperatures set by a 7 day program but with manual overides. If I decided to get up earlier than normal then a push button on the bedside table would bring the bedroom heating on immediately. Or if i had to go out un-expectedly a single tap on a button by the front door would put the heating to low.


The systems was designed, all the components assembled and bench tested. It worked, but during that time the reality of the heating requirement for the cottage became apparent. It was comfortable with fewer and smaller radiators than the original heat loss "calculations" had predicted were necessary. Thatched roofs confuse heat loss calculations.

Also maintaining different temperatures in different rooms was not really feasible given the construction of the buillding and restrictions on what changes can be made to a Listed Building. . With these changes much of the custom hardware is not needed. ( but as an engineer it was a good interesting project and I have some very good temperature sensor modules )

More later if interested

Thanks for that Bernard,yes,very interesting; obviously it was going to be a very comprehensive installation. I think any purpose built self install of this type is fascinating, to see how a problem was approached and solved in an individual way. I wonder what you actually used for heating control finally then?
 
The whole problem I found was the software which came with the valves has no instructions as to what it does, in my case using MiHome eTRV's and I found if set to go from 16°C to 20°C at 6 am by 8 am when I wanted 20°C it had not reached it, I had to set it to go to 24°C at 6 am then at 8 am back down to 20°C to get the temperature required.

Same going down, from 22°C at 9 pm had to set to 16°C even though I wanted 18°C for the valve to close enough, seemed I had to do this at every change, and it would also change due to sun light through bay windows, and the software built into the valves was too slow to react to keep the room from over heating, I was very disappointed, they worked, can't say they didn't, but not as wanted.

The biggest help to controlling the house, was not the eTRV but actually fitting a simple TRV on radiator where the wall thermostat was fitted, what it allowed was fast heating of hall when front door had been opened, but then slowed down the heating after 18°C so enough time for rest of house to heat up, before the wall thermostat switched off boiler at 19°C. Clearly not a problem with EvoHome as you don't need a wall thermostat, but had I done that to start with, the eTRV would not have been required.

My eTRV has the ITTT system, but found near impossible to write my own software, it would require reverse engineering to work out how to write it, so in real terms only option is to select an app already written from the MiHome web site.

So in theory you should be able to write a program so if the target is above the current on any eTRV then switch a relay on so powering up boiler, the relays are made and sold, so hardware wise it can be done, however software wise I could not find the ready made software to do this, or enough info to allow one to write ones own software app.

I can't even find out where each bit of the program is stored, we have the valve head, the hub, and the internet connected server, one would hope if set to change from 16°C to 22°C at 6 am that would be in the valve, but one does not know if this is the case or not, the web page has last seen at moment says 1 minute ago, but I have looked some times to set 2 days, as in the summer when that was reported don't know what the valve would have done. However if I was doing it again, I would not use WiFi connected eTRV's I would use local set type, these would work just as well for me, as although it tells me temperature of room, I have no way of using the info.
 
Thanks for that Bernard,yes,very interesting; obviously it was going to be a very comprehensive installation.
It was going to be but the original concept was too "comprehensive" and over elaborate for the task of heating control.
I wonder what you actually used for heating control finally then?
One timer and one thermostat in the dining room was all that was needed to control heating of the whole cottage and keep it comfortable. The zone valves are manual because it works and the motivation to "automate" them is low. All the necessary cabling is in place, I just need to find time to do the work.
 
It's been interesting reading the thread, I too have some ideas if I ever have the time !
One consideration though is supportability. If I'm not around, even basic off the shelf automation stuff is well beyond the comprehension of the majority of plumbers who would take one look and declare that the only way of fixing it is to rip it all out and fit a combi.
 
the only way of fixing it is to rip it all out and fit a combi.
That is the problem when systems get over complicated.

I feel that many of the features on the top end "home automation" systems are more a sales feature than a feature that is useful in the management of the home
 
In the cases I was thinking of, it wouldn't even be "complicated". I have some ideas which aren't actually complicated (stuff that any half competent plumber should be capable of dealing with) - but I know that most plumbers would not even try and understand them, they'd just "<suck through teeth>What you want is a combi". I just hate to think what some of them would make of something as "complicated" as a radio receiver switching the boiler based on feedback from electronic TRVs :LOL:
If you don't believe me, have a search in the plumbing section for "thermal store" and see the sort of responses you get from people who claim to be plumbers. There's one or two who almost take it as a badge of honour that they would not under any circumstances consider recommending or installing one - or put another they, they are proud of the fact that they will deliberately make suboptimal recommendations to some people :evil:
 
they'd just "<suck through teeth>What you want is a combi". ~~~~~~ they are proud of the fact that they will deliberately make suboptimal recommendations to some people :evil:

I doubt they are proud of it. They may have to make a suboptimal recommendation because either they have no knowledge of anything better than that equipment which they have been trained on or are they are contractually committed to recommending equipment from that manufacturer.

Fred the plumber says to customer "" I can install a boiler from XXX and then you get a 10 year waranty but you want a boiler from YYY. That will have a 3 year warranty if I install it for your "" but fails to mention that Pete Plumber could install a boiler from YYY and offer a 10 year warranty.
 
If I'm not around, even basic off the shelf automation stuff is well beyond the comprehension of the majority of plumbers who would take one look and declare that the only way of fixing it is to rip it all out and fit a combi.

That's the answer to anything isn't it?
 
I look at the info page for my two eTRV's it shows target and current temperature which is all that is required to know if the boiler should run, they make a relay, so all it needs for my eTRV's to work nearly the same as the EvoHome is software. It is if this then that, so it could be done, however it seems it has not been written, at least for general use.

So my boiler is controlled by carefully balancing the standard TRV and the wall thermostat in the hall. I am told by plumbers that you don't fit wall thermostats and TRV's in the same room, however for me at least it works.

The TRV controls each room gradually turning more open or closed so maintaining a constant heat, the boiler flame goes up and down also maintaining the gradual changes in demand, once the boiler can't turn down any more, then it starts to cycle, and only after that do we actually want to fully switch off the boiler, and it seems it is this last function which is so hard.

In real terms this last function is done manually, at this time of the year my thermostat is set to defrost, in another 3 months I will need to reactivate the heating, however I set the stops on the thermostat, so all it needs is turning until it hits the stop and all should then start to work again.

So in real terms I did not need the eTRV's the standard ones were good enough, what I needed was a TRV in the hall, once I got it to work, it seemed so simple, hall is the room with the outside door, so wants to be able to heat up quickly, it is also the room which we want to be the coolest, the TRV kicking in just before the wall thermostat means in really cold weather the wall thermostat never gets warm enough to switch off, all control is done with the TRV's, but as the weather gets warmer the hall starts to warm up just that little bit more, and the wall thermostat kicks in.

Sounds simple, however although I can set it when living in the house, for a heating engineer to set it would be hard, it is a case of 1/16 of an inch turn every day until it works A1, they are simply not in the house for weeks on end, so near impossible for them to do what I have done. The same applies to lock shield valve setting. And once set it is a case of leave it well alone.
 
I didn't say anything about controlling when I'm not in the home... for me it's about controlling when I am in the home. Ideally without telling the internet every detail.

Remote control I'm less convinced about when you're not home, though they tend to use examples like if you come home from work early. I think many bits of this are gimmicky but if you did shift work or had an irregular schedule, being able to turn the heating on as you leave work might be nice. And only turning the heating on when any radiator actually needs heat seems a good idea compared to the heating coming on every evening even if the TRVs are shut.
The heating won't come on every evening if the room stat is satisfied. Trv are just for local control and have very limited use.
 
The problem with relying on a room stat is that you will always have a compromise. The room stat has to be in the space that "most often" needs heating - otherwise you'll find other spaces fail to get heat when needed. And the standard wax-capsule TRVs aren't all that good at regulating in the face of varying demands - there's a fair spread in temperature required between "completely shut" and "allowing full heating".
BUT - a combination of TRVs and a room stat is a huge improvement in control over "f**k all" :rolleyes:

That's where eTRVs and linked boiler controls really win. The eTRVs (by incorporating an integral element) can provide significantly better control (not to mention, room by room heating profiles*), and by linking them to the boiler you can get real "only when needed" firing of the boiler.

Having said that, in the flat I have a thermal store (so boiler decoupled from loads) and the heating runs with a modulating pump (Grundfoss Alpha4). Because the pump modulates down as flow reduces, the TRVs work much better and that is the only control apart from the timer - the tenants since I put that in have been very happy with it. It's also very quiet without all the hissing you get from a regular system with the pump fighting against almost closed valves.

* Depending on system abilities

But room-by-room control isn't new. Back in the early 70s, one of our neighbours had his new-build fitted out with a room stat in every room controlling a 2 port zone valve for that room. I can only assume that some later owner ripped all that "complicated nonsense" out.
 
Back
Top