@batalto 

Basically what I listed in the first post. 

@benseb Thanks. I shall get hold of one and give it a go then.

@benseb - where is the setting to change the auto adaption ramp up time please? I couldn't find that in the manual.

Posted by: @benseb

However the remote controller still lets you set the room temp in WC which I was expecting it to disable and “go dumb” or control The curve. 

 

What it does in practice is run off the WC curve but when the targetroom temp is exceeded by 1c it stops the heat pump completely which makes sense I guess as otherwise you’d have a completely useless expensive room controller unles you were in auto adapt mode  

Mine doesn't do that! When in weather compensation mode, the controller changes to +/- adjustment of the curve. The ecodan pump only stops when the flow and return temperatures get too close (which I think is within 3½ degrees) which usually means the room temperature is steady.

I didn't get on with auto adapt mode. It flip-flopped between running the heat pump at maximum and not at all, never reaching a steady state and so the COP was poor. "Completely useless" is a good summary of my feeling towards it! I put it back into curve mode and let another smart controller switch it off temporarily over melcloud on the rare occasions that the house gets too hot. I keep considering letting the other controller switch the heat pump off for an hour if it spots it cycling, but that happens rarely enough that I've not done it yet.

@mjr

As a Control Systems Engineer, I have often wondered how controllers utilising Auto Adaptation would cope with not only the varying operating conditions throughout the year, but also adapt to the varying conditions from home to home. Let me explain.

By adjusting the parameters for the Weather Compensation curve, one is actually defining the controller response to at least some of the external influences which cause the internal temperature to vary. When correctly adjusted, the energy produced by the heating system will reasonably match the heat loss at differing outside temperatures. But of course other effects such as solar gain, human activity, wind chill and rain are not being measured, and could cause the internal temperature to vary.

When functioning correctly, Auto Adaptation should help mitigate these additional effects, by varying the controller response to try to keep the internal temperature fairly constant.

Within an industrial controller there are three main parameters which are set to vary the response of the control system, these are Proportional Action, Integral Action and Derivative Action, hence the term often used being PID Controller. The Commissioning Engineer will often adjust these parameters to make the control system operate in the most efficient and effective manner. Once set these parameters are not adjusted unless the control system is no longer functioning correctly.

Having read quite a number of manuals for heat pump controllers, I do not remember seeing any auto adaptation parameters that can be adjusted by either the installer or user. I can therefore only conclude that the heat pump controller contains some software algorithm which carries out some form of tuning of the controller response. Because the response of home heating systems is very slow compared to industrial systems, I would expect such tuning to take days, if not weeks.

With regard to your system mjr, you may have a problem with the siting of the controller or external temperature sensor (if fitted). If the temperature sensor is too close to a heat emitter, then it could be reacting too quickly to the temperature increase and shutting down your heat pump. It would then cool fairly quickly and restart the heat pump, which of course leads to the type of operation that you described. Having a sensor within direct sunlight may have a similar effect.

If you have a suitable thermometer then place it near your temperature sensor and switch on auto adaptation. Then monitor how quickly the temperature varies as the system responds.

@derek-m 

have a look at page 38. Isn't that adjustments for auto adaptation? 

https://library.mitsubishielectric.co.uk/pdf/book/Ecodan_FTC6_PAC-IF071-3B-E_Installation_Manual_BH79D843H02#page-38-39  

I've been looking at this stuff again because I'm about to commission my remote thermistor, which should let me use auto adaptation. 

Posted by: @derek-m

With regard to your system mjr, you may have a problem with the siting of the controller or external temperature sensor (if fitted). If the temperature sensor is too close to a heat emitter, then it could be reacting too quickly to the temperature increase and shutting down your heat pump. It would then cool fairly quickly and restart the heat pump, which of course leads to the type of operation that you described. Having a sensor within direct sunlight may have a similar effect.

I'm not sure those are the problems: the controller and its temperature sensor are on the wall of the landing, not near a radiator, not in a draught, with no sunlight hitting it. The main drawback I think is that it's upstairs which tends to be slightly warmer than down because heat rises. The external sensor is in direct sunlight for the last few hours each spring/summer/autumn day (it's in a northwest-facing corner) but I think the worst that would do is lower a weather-compensated flow temperature or possibly make the controller think that the external temperature falls after dusk faster than it really does... and that wasn't a factor on the November mornings when we last tried auto-adapt mode and it went silly.

Posted by: @derek-m

If you have a suitable thermometer then place it near your temperature sensor and switch on auto adaptation. Then monitor how quickly the temperature varies as the system responds.

I'm really not keen on trying auto adapt again while energy prices are so high. Maybe it would learn, but there was no sign of it last time, so how long would it take, how much would it cost in the meantime and will it forget the learning again at the next firmware update?

@mjr

All your points taken. Without knowing exactly how the internal software is programmed and how it is supposed to function, it is difficult to give any definitive advice. I suppose that you could try contacting Mitsubishi and ask them to explain how an FTC Controller is supposed to function in auto adaptation mode.

If you have a multimeter that can measure resistance, you could try checking the sensor at various temperatures to see if it is functioning, obviously disconnect it from the controller first. Also check for loose connections. It could also be that the sensor or controller are faulty.

Kev M highlighted some settings on page 38 of the manual which may have some effect. Unfortunately there are no P, I or D parameters that could be used for fine tuning.

I set the mode to normal (which favours economy over fast warm up/comfort). I also changed the interval from 10mins to 60mins

my guess on that is it waits 60 mins before ramping the WC curve up/down if the room temp hasn’t met the target. As we have UFH and I wanted to avoid it ramping up too much I went for the longer interval. 

I actually switched to pure WC this week as a comparison. Tomorrow I plan on switching back to room adapt for a week to compare costs. 

very impressed since I’ve made the changes. Downstairs is a very steady 19c. Heat pump on about 40% of the time and COP about 4.7

i actually turned the WC curve down -3 today with no I’ll effects. That’s running at about 24c flow temp today. Which isn’t bad for a 300 year old house!

I did notice some of the upstairs rooms were a little cold so rather than ramp back up I turned the flow rate on the kitchen where the stat is, down by 0.5l/min so it would take longer to reach target temp and therefore give the other rooms a little longer to get warm. 

seems to have worked!

I think we’re benefiting from a lot of thermal mass and solar gain at the moment.

Yesterday was quite nippy outside, about 8c in the morning. We used 17kWh for heating (output 75kWh so COP around 4.4)

Posted by: @derek-m

Kev M highlighted some settings on page 38 of the manual which may have some effect. Unfortunately there are no P, I or D parameters that could be used for fine tuning.

That's still a big improvement on the FTC6, though. In the FTC5 manual, the similar settings on pages 41/42 are also labelled as "Heating operation" but footnote 8 says they only apply to room temperature mode, which I believe is what you get if the auto-adaptation isn't enabled by dip switch.

Oh and the sensor seems fine because I use it for decision-making in my overriding controller software. It seems lower-resolution (0.5°c steps) but more accurate than the one I have attached directly to the Raspberry Pi.