Thermostat Settings

[Jeremy Harris]

37 minutes ago, Nickfromwales said:

@JSHarris

Do You know if there is a similar room thermostat to your wireless one, but hard wired? I'm thinking of recommending them to the customer I did the manifold and auto-balancing actuators for as they have the basic Honeywell rotary manual stats with a horrible 1.5oC hysteresis . The house has pretty poor floor insulation ( ~70mm pir over a ventilated block and beam ) and a thin liquid screed, so the overshoot is difficult to manage, plus, as the rooms are quite big, the undershoot takes some time to recover. The manifold temp is set a bit higher than would be ideal ( so further exaggerating the overshoot ) so I think an upgrade to the stats would pay dividends. 

 

 

The ones I fitted are Computherm Q3 RF models, but they do a cheaper wired version, the Q3, plus a smarter one, the Q7.  Here's the first link I could find to a seller with the Q3, you may find it cheaper by shopping around: https://www.amazon.co.uk/Computherm-Q3-Wired-Room-Thermostat/dp/B00SNHCWQC

 

The hysteresis is set by a link on the back, either +/- 0.2 deg or +/- 0.1 deg.

Edited November 17, 2017 by JSHarris

[SteamyTea]

Rather than worry about the hysteresis  (a word I still not spell or pronounce), could a thermostat be set to a temperature and then use a timer to delay the on/off period.

 

So say it takes 20 minutes for the room to cool by 0.5°C, onces the set point is set, there is no way the heat source can add to it for at least 20 minutes (times and temperatures may vary).

This would stop short cycling.

A similar idea could be used on the heating cycle but would need two set points, the upper and lower.  Once the lower set point is reach, the heating system stops.

That is really just a smarter hysteresis  idea, but with some lock outs.

 

I would think that it was a pretty simple thing to DIY.

[Lesgrandepotato]

Isn’t this stuff what the heatmiser smart start stats are about? Learning how long it takes to hit temp and starting appropriately? 

[Lesgrandepotato]

My plan is to use the fibaro HC2 to read my calendar for heating the office above setback temp. The music room / guest room will probably be monitored using a door switch and if the door is opened for a prolonged period of time - say 5mins then we’ll fire the temp to living standard for an hour or so. We could probably override with a guest calendar but or a switch on the Murphy bed. But that could be GUST (gratuitous use of superfluous technology...) 

[SteamyTea]

14 minutes ago, Lesgrandepotato said:

Isn’t this stuff what the heatmiser smart start stats are about? Learning how long it takes to hit temp and starting appropriately? 

I suppose it is, with a bit of automatic curve matching (which is not hard, though the coding my be tricky).

[Jeremy Harris]

When you have a house where the heat loss time constant is long, over ten hours or more for a 1 deg drop in temperature in winter with no heating, all these clever controls really become superfluous.  You just ensure that there's "just" enough heat put in to the house to maintain a comfortable temperature, as it becomes impractical to set back temperatures during the day, or over night, or whatever, because the cooling time is way too long.

 

The tricky bit is maintaining comfortable differential temperatures, so that bedrooms, for example, remain slightly cooler than the living area.  An upside down house might do this better than a conventional design, but I've found that just closing the bedroom doors tends to work OK, as that stops pretty much all the heat from downstairs getting to the bedrooms.  This only really works for us because the bedrooms both have ensuite bathrooms, so there is, in effect, a closed cycle bit of the MVHR where the fresh air fed to the bedrooms gets pulled out through each attached bathroom, with little air being pulled in under the  bedroom doors from the main house.

[SteamyTea]

I am going to start a new topic on heating control as a few of us are practical and like the subject (and got a box full of RPis and components).

https://forum.buildhub.org.uk/topic/4142-heating-control-systems/

 

Edited November 17, 2017 by SteamyTea

[Trw144]

12 minutes ago, JSHarris said:

When you have a house where the heat loss time constant is long, over ten hours or more for a 1 deg drop in temperature in winter with no heating, all these clever controls really become superfluous.  You just ensure that there's "just" enough heat put in to the house to maintain a comfortable temperature, as it becomes impractical to set back temperatures during the day, or over night, or whatever, because the cooling time is way too long.

 

The tricky bit is maintaining comfortable differential temperatures, so that bedrooms, for example, remain slightly cooler than the living area.  An upside down house might do this better than a conventional design, but I've found that just closing the bedroom doors tends to work OK, as that stops pretty much all the heat from downstairs getting to the bedrooms.  This only really works for us because the bedrooms both have ensuite bathrooms, so there is, in effect, a closed cycle bit of the MVHR where the fresh air fed to the bedrooms gets pulled out through each attached bathroom, with little air being pulled in under the  bedroom doors from the main house.

 

Yes, my thoughts are the same, hence not bothering with set backs (and a lot of you have better levels of insulation and airtightness than me).

[Lesgrandepotato]

1 hour ago, JSHarris said:

When you have a house where the heat loss time constant is long, over ten hours or more for a 1 deg drop in temperature in winter with no heating, all these clever controls really become superfluous.  You just ensure that there's "just" enough heat put in to the house to maintain a comfortable temperature, as it becomes impractical to set back temperatures during the day, or over night, or whatever, because the cooling time is way too long.

 

Thats the key differentiator, this is not a problem our house has! 

[dpmiller]

@Nickfromwales try looking inside the 'stat and see if you can move the accelerator resistor (heater) closer to the bi-metal...

[Grosey]

I have the heatmiser neostat system and I believe it can do a lot of what people are discussing here -

 

Optimum start - an allowed “pre heat” time so the stat can call for heat up to a set amount of hours before the time setting. 

 

Differential - amount temperature has to drop before stat will kick back in - 0.5 - 3.0c

 

Output delay - how long since heat was last called for before it can call again: 1 - 7 minutes. 

 

Up / down limit - how many degrees change from the current set temperature can be achieved using the wall stats 1-7c

 

Pin lock - locks the room stat entirely. 

 

 

All all of this can be done in the room stats but I access everything via the app which is much easier. 

[Jeremy Harris]

24 minutes ago, Grosey said:

Differential - amount temperature has to drop before stat will kick back in - 0.5 - 3.0c

 

This is way too large for a house with a long heat time constant.  I run our thermostat at the optional +/- 0.1 deg C hysteresis setting and if I have the flow temperature to the UFH set too high I will still get several hours of room temperature over-shoot of up to around 1 deg C, even with a thermostat that turns off 0.1 deg C above the set point.

 

There are several ways around control, depending on how low the heating requirement and heat loss is for a particular house, the effective thermal capacity of the interior of the house (really just the first 100mm in depth, at the very most, of all internal surfaces, with the inward facing 50mm having by far the greatest effect) and the local climate.  There's a big difference between being in a mild, sheltered, location and being in a severe, exposed location, big enough to radically change how a control system might best be optimised, I think,

Edited November 18, 2017 by JSHarris

[dpmiller]

need to force the hysteresis then of that accelerator resistor in the old mechanical stats. Could you put a small resistor across the OP to cheat some heat into the 'stat?

 

Another option for those that don't have great control of the flow temp might be one of the cheapo Chinese touchscreen stats w/  floor sensor- set the floor limit low and it'll override the room setpoint once the heat is into the slab.

[Nickfromwales]

16 hours ago, dpmiller said:

@Nickfromwales try looking inside the 'stat and see if you can move the accelerator resistor (heater) closer to the bi-metal...

The problem I have is I need to give my customers a regular, reliable, replaceable device that can be bought off the shelf by any other trade or heating installer. 

As there is a wired version of Jeremy's stat and the performance is tried and tested then I just can't really justify going and 'tweaking' anything else.  

If it were my house then it would be fair game but all customer solutions need to be off the shelf just in case, as I've said elsewhere, I get eaten by a prehistoric dinosaur brought back to life by an eccentric billionaire. ?

It could happen ?

[Nickfromwales]

11 hours ago, dpmiller said:

need to force the hysteresis then of that accelerator resistor in the old mechanical stats. Could you put a small resistor across the OP to cheat some heat into the 'stat?

 

Another option for those that don't have great control of the flow temp might be one of the cheapo Chinese touchscreen stats w/  floor sensor- set the floor limit low and it'll override the room setpoint once the heat is into the slab.

I've often pondered over using an electric undertile controller for running Ufh as the better ones, the Warmup 3ie for example, have a learning facility for the setback times. Basically if you want the room temp at 21oc at 07:30 and the heating comes on and takes 1 hour 30 mins to achieve the set temp, then the stat will sample this over ( iirc ) three 24 hr cycles and then adjust itself accordingly. So basically on the 4th day it'll switch on at 06:00 and then ( again iirc ) it'll do the same for the off setting as the heat will continue to emit after the heating goes off, just probably a lot quicker in some instances but each case is individual. Have a read of the bumf but I think it re-samples occasionally to adjust with seasonal / envoirmnental changes.

 

The 4ie has the wifi / smart phone connectivity so could negate a lot of other system components otherwise needed to control your heating from say a hand made Coracle, ( should, for eg, you have accidentally fallen out of one and into a cold river whilst trout fishing and need to heat the house up ready so when you get home you can run around naked to dry off ? ).

[Stones]

5 hours ago, JSHarris said:

There's a big difference between being in a mild, sheltered, location and being in a severe, exposed location, big enough to radically change how a control system might best be optimised, I think,

 

Absolutely! The higher average wind speeds that we experience, plus the multiplying effect on pressure differential caused when we get gale force winds, really does have an impact on our heating requirement when compared to the requirement (modeled) if our house was in a sheltered location.  We do however live in a relatively mild location, and the range of ambient temperature change is a lot lower than other parts of the country.  

 

We use a packaged system, with the master ASHP controller set in an 'adaptive' mode, which according to the blurb, learns how to heat your house.  I think in effect, all it does is work out the optimum weather compensation curve rather than having to manually set it.  I've yet to see the flow temp exceed 27C.  I let the heating run 24/7 as a single zone for simplicity with the master controller thermostat set to 20C.  Excluding solar gain (which we haven't had much of in the last week) internal temperature remains constant (+/- 0.2C) at 20.6C  When we do get solar gain, the living section of the house can increase in temperature, but this is something that we are happy with (very similar in many ways to having a fire on). The heating clicks in, as and when required, with a 30 minute UFH pump over-run of 30 minutes.  No other input from me and works really well.