UFH - Am I expecting too much?

[lloyda]

Recently had a bathroom refit with electric under floor heating fitted. This is controlled with a Warmup Tempo thermostat, using a floor sensor. It works, but tends to overheat the floor by about 2 degrees. e.g. stat set to 22, it continues to heat to about 24, shuts off, and then overshoots by perhaps another two degrees. This is unit is supposed to have a proportional adaptive function, which I would expect to have better control than it does. It is enabled, as the floor comes on prior to the set time. I understand that due to the nature of the floor the response is going to be slow, but should I really expect it to overshoot by so much?

I've tried warmup support, and their response is a) it is only accurate to +/- 2 degrees (!), and b), it will eventually learn.

Anyone have a positive experience of this thermostat?

 

[flyingsparks]

If the technical spec says accurate to +/- 2 degrees then I don't think you can expect much better to be honest, can you?

 

[ericmark]

Hysteresis is always a problem does not matter if heating fermenting beer or a room at home. When doing stress relief and pre-heating when welding pipes we had some very clever equipment but it would still over shoot and the skill of the operator was required. The absolute maximum (BS EN 1264) is 55°C, but for bathrooms normally 33°C is considered as maximum certainly for a wet room. The Telegraph did a write up and pointed out if you want the room at 22°C and max is 33°C with extractor fans required under Part F building regulations then UFH will not heat the room. There has to be also some other form of heating. This is where the whole idea of a learning system falls down as it does not know how much heat is coming from else where. So I would say within 4°C is doing quite well.

 

[lloyda]

But why would you have a control system that continues to supply heat, two degrees after you have reached the desired temperature? (And all the temperatures that I mention are as read by the floor probe - I am not comparing air temp and floor temp)

 

[ban-all-sheds]

Well - think about how much hysteresis you think there should be.

Imagine that you had a probe & thermostat system which was accurate to 0.1%, and you set the temperature to 22°. Would it be realistic to have it switch off at 22.022° and back on again at 21.978°?

Also, is floor temperature the best way to control the heating anyway? I've never heard of anybody controlling conventional CH by measuring the surface temperature of their radiators. Surely what matters is the room temperature, so how constant is your system keeping that?

 

[Diyisfun]

You learn what it does & set accordingly.
I think that is what most people would do.
It took me a year to get to grips with my heat/timer settings.

 

[lloyda]

Well - think about how much hysteresis you think there should be.

Imagine that you had a probe & thermostat system which was accurate to 0.1%, and you set the temperature to 22°. Would it be realistic to have it switch off at 22.022° and back on again at 21.978°?

Of course not. But for a system that claims to offer proportional control, I would expect it to turn off at say 20.5°, and then gradually approach 22° by switching on/off as required. Yes it will go a bit over (and a bit under), but it should learn what the hysteresis is, and minimise the swing. To repeat myself, why would it still heat when it is 2° over? What is it trying to achieve?

Also, is floor temperature the best way to control the heating anyway? I've never heard of anybody controlling conventional CH by measuring the surface temperature of their radiators. Surely what matters is the room temperature, so how constant is your system keeping that?

I'm not saying it is - I just want to understand that this thermostat is doing what it is supposed to do, which I don't believe it is.

 

[davelx]

surely a system that has proportional control cannot work with a thermostat (an all or nothing, on or off, device) - you need a temperature sensor so that the system knows the delta between the actual temperature and the set point, and varies the heat output accordingly, slowly reducing it as the set point is approached?

 

[stillp]

It might just learn the time it needs to be on in order to reach the set temperature at the set time.