Wired Radiator valves - Honeywell MT4/MT8

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • Cube
    Automated Home Lurker
    • Jan 2019
    • 5

    Wired Radiator valves - Honeywell MT4/MT8

    I've just finished reading the monster thread called "Evohome is Noisy - please help me".

    Beside the relief of surviving the experience I learned about the MT4/MT8 Small Linear Thermoelectric Actuator.

    This seems like a miraculous device with lots of advantages: -
    • It is totally silent (no motors).
    • No batteries to replace.
    • It supports proportional control through PWM.
    • 24V version available (safe for bathroom use).
    • It responds immediately (although it can take 2 mins to fully open).
    • It won't lose connection due to interference.

    It does have a number of downsides as far as I can see. For one it doesn't easily slot into the Evohome infrastructure unless you are willing to buy a relay device for every valve and sacrifice the possibility of proportional control of the radiator flow. Secondly it has no built in thermostat or display but to be honest I don't see that as a problem. I see Honeywell recommend driving it with something called Excel 10 or 12 but don't have much detail yet.

    So has anyone driven their radiators with this instead of full Wireless TRVs like HRT92? If so I'd be interested to hear how easy it was to drive (especially if you used PWM).
  • DBMandrake
    Automated Home Legend
    • Sep 2014
    • 2361

    #2
    I've seen those before but the big stumbling block is for proportional control you'd need an external PWM controller (the Excel 10 or 12, whatever those are) and I'm not aware of any way to interface one of these with an Evohome system and maintain proportional PWM control.

    Honeywell seem to be missing a trick here by not providing a (remotely mountable) box that correctly interfaces these to an Evohome system whilst maintaining full PWM proportional control equivalent to an HR92. In other words you'd want this thing with it's remote controller box to look like an HR92 in a regular radiator zone as far as Evohome is concerned. Potentially this remote controller could be fully built into a 3 pin wall wort with low voltage wires to the actuator and wireless communications from the wall wort to the Evohome controller, making installation in areas where a free power socket is near the radiator relatively easy.

    The only way you could use one at the moment would be to use the two position version with a BDR91 and set it up as a "zone valve" zone. This would use the system wide 10 minute cycle TPI so would be proportional control of a fashion, but not nearly as good as actual continuous PWM control which would hold it at a precise partial opening rather than just fully opening and closing it every 10 minutes.

    A few problems with this approach:

    1) TPI would not give as good control as PWM, but would be "OK" and better than nothing. From a control perspective it would be no different than having a remote 2 port zone valve controlling the flow through that individual radiator with a BDR91 controlling it. However if it is very slow to open and close compared to a zone valve control would not be as good. One possible issue is that if the radiator flow is "noisy" for example has a lot of hiss, that hiss would keep stopping and starting every few minutes. This could also disturb sleep compared to a constant hiss. Some radiators are silent with full flow, some aren't, and there may be some hiss generated as it goes from closed to open or back.

    2) A BDR91 clicks quite loudly so there would be a loud click every 10 minutes or so which would defeat trying to make the room silent - so then you're talking about running a long mains voltage cable through a bedroom to a BDR91 mounted in the loft space, in a wardrobe etc to try to muffle the clicking noise to inaudibility. A silent actuator is not much use in a bedroom if the thing controlling it is noisy and in the same room!

    3) If you want safer low voltage 24V wiring for example bathroom or perhaps bedroom you still need 240v AC to the BDR91 and then an additional 24v power supply of some sort near the BDR91. The BDR91 does have isolated volt free relay contacts so this is possible, although it's a lot of hardware to stuff away somewhere safely and inconspicuously for just one radiator.

    4) No temperature sensor as you point out, so now you're paying for a DTS92 or similar for the zone as well. So the total cost of temperature sensor, actuator, BDR91 and possibly 24v supply is getting rather high compared to an HR92.

    5) While the actuator won't lose wireless comms the thing that controls it still needs wireless comms from some location in the house back to the controller. So comms loss is still a potential issue.

    I too am interested to know if anyone has actually used one of these yet and how they set it up. It's been mentioned several times as an option but nobody has come back and said they used one.
    Last edited by DBMandrake; 5 January 2019, 10:21 AM.

    Comment

    • Cube
      Automated Home Lurker
      • Jan 2019
      • 5

      #3
      You make some excellent suggestions.

      Originally posted by DBMandrake View Post
      ... the big stumbling block is for proportional control you'd need an external PWM controller.
      That's the problem in a nutshell.

      I've had a look at the Excel 10 range of controllers mentioned in the literature and it doesn't really help. You would need one Excel 10 controller per radiator. I was kind of hoping for something with 8 or more PWM outputs so you could run wires from it to each radiator in a "star" layout.

      So it looks like the only way to do this would be to make some Arduino based project using MOSFET drivers (to avoid the audible relay clicking). From the data sheet it looks like the devices expect 24V AC with a minimum cycle time of about 30 seconds. So if you output 24V for 15 seconds and nothing for a further 15 seconds the valve would be half open.

      If you used CAT5 cable leading from the controller to each MT4 you would only need the one PSU to the controller. You might also be able to wire up the switch in each MT4 so you could detect that the valve was physically working.

      It all sounds like a fun project but I think on reflection I'll go the HR92 route and pay for Honeywell's research leaving me to concentrate on doing something similar for the lighting controller

      Comment

      • martin@tinroof.me.uk
        • Jan 2019
        • 1

        #4
        Originally posted by DBMandrake View Post
        I've seen those before but the big stumbling block is for proportional control you'd need an external PWM controller (the Excel 10 or 12, whatever those are) and I'm not aware of any way to interface one of these with an Evohome system and maintain proportional PWM control.
        ...

        The only way you could use one at the moment would be to use the two position version with a BDR91 and set it up as a "zone valve" zone. This would use the system wide 10 minute cycle TPI so would be proportional control of a fashion, but not nearly as good as actual continuous PWM control which would hold it at a precise partial opening rather than just fully opening and closing it every 10 minutes.
        ...
        I actually did exactly this in our main bedroom. My wife was complaining about being woken up early by the HR92 whirring away every few minutes as the room warmed up in the early morning. I replaced the HR92 with a MT4 connected to BDR91, and the room set up as a zone-valve zone. The relay box is situated in a void behind our en-suite shower room, which is separated from the bedroom by an wooden access panel. This reduces the relay click sound to a level where it is no louder than the occasional clicks from our piping expanding - barely noticeable. We have a remote sensor near the bed (one of the round ones, T87RF).

        The TPI temperature control is not as good as the HR92 - there is a distinct tendency to overshoot by up to a degree above the setpoint. I suspect that the accuracy of the control depends a lot on the characteristics of the coupling between the MT4 and the valve - i.e. how 'open' the MT4 needs to be in order for water to start flowing. I've recently started to experiment by adding some cardboard shims between the actuator and valve body, to delay the valve opening - it's too early to say if this is helping or not.

        A proper BDR91-like box that provides proportional control of an MT4 would be nice in this situation, but I believe that the basic MT4/BRD91 combination is good-enough anyway.

        Comment

        Working...
        X