Heatmiser Neostat v2 temperature sensor problem

[Ultima357]

I have a Samsung ASHP linked via Heatmiser controls to a 14 zone, 21 loops UFH system. We commissioned this back in March 2020 just before lock down and after 1 week, it shut down as the property is passive insulated and triple glazed. Airtight too at just 0.17achr. Due to lock down delaying our finishing and selling the old place we finally moved in in September. Since the cold weather in early December I've been trying to get the heating system to work to our requirements. Initially sitting in the lounge and not being warm enough despite the Neostat showing 23deg, I invested in a calibrated thermometer and set about calibrating the Neostats to this. Property has MVHR too, so adds to the overall picture. This has now been nearly two months of adjusting but recently discovered that when the MVHR was switched off, the stats jump up in temperature value and records the house warming up overnight when the heating is off. So further probing found that the stats are recording their own heat. Front panel runs at 27 degs, the thermistor sensor on the bottom edge has no natural vents in the rest of the housing so with no air movement, the central heat of the stat just warms this up. Taking the stat off the wall I found the rear fixed plate running at 29 degrees. Top this all off with the fact that the bathrooms with remote sensors are fine and record the actual temperature and are not affected by MVHR airflow leads me to conclude that the Neostats are not fit for purpose. They generate too much internal heat to read correctly. To prove this I took one apart, drilled four holes in each side of the housing at the top and behold, it reads more accurately, now only 0.6 deg high with still air as opposed to 2 deg plus beforehand. The photos show the temperature with the stat given a gentle hand fan movement of air, still air without fan and the temperature of the face. The final one shows the drop in recorded temperature when. the MVHR cuts in. Note this MVHR is a Brink Flair 400 running at just 50m3/hr in this shot, the dining room being over 120m3, it is not the slightly cooler air hitting the stat directly as the dining room outlet will be serving up only about 6m3/hr but merely the stirring of the air stopping or reducing the stat reading its own temperature. So thanks for reading this far and just wondering if anyone has had head scratching moments with these stats? Going to try buying a decent looking remote thermistor and hooking it up just below the stat to see if this cures the problem and then seek redress through Heatmiser who have already suggested using remote sensing. Their suggestion rather reinforces my view of not fit for purpose. 

 

 

[ProDave]

My thought is send the thermostats back as not fit for purpose and replace with something else. You should not have to bodge them.

[PeterW]

Sadly this confirms most of my opinions about the quality of the products under that brand. 
 

Have you used the Heatmiser wiring centre and their actuators ..?

[Temp]

Ive had Heatmiser stats for 13 years without any issues but they are an older model.

 

I'm wondering if you have a wiring issue? Are they the12V version powered by the wiring centre? Is the wiring centre putting out a higher voltage than it should?

 

If you call Heatmiser don't bother going into all the detail about the MVRH that just confuses the picture. Just tell them the stats appear to be over reading and its not a calibration issue. 

 

 

Edited January 17, 2021 by Temp

[Ultima357]

In regards to the wiring centres, they are Heatmiser, the actuators are third party and the stats are the v2 model, 240v powered. I think the fact that the back plate contains the 240 to 12/5v psu (whatever the front panel runs at) is where all the false heat is generated. I'm frankly amazed that such a poor design gets to market and so widely distributed /badged by others too. 

31 minutes ago, PeterW said:

Sadly this confirms most of my opinions about the quality of the products under that brand. 
 

Have you used the Heatmiser wiring centre and their actuators ..?

 

[Temp]

Mine are 230V but most are in surface mount back boxes. The stats don't have massive vents but never seen this issue. They seem to read pretty accurately.

 

If anything the one we have with the remote sensor is the least accurate and I made the remote sensor myself.

[PeterW]

That stat in the pic looks quite close to a work surface ..? Have you got a further back view ..? That could be creating a heat pocket or lack of airflow 

 

 

[Onoff]

12 minutes ago, PeterW said:

That stat in the pic looks quite close to a work surface ..? Have you got a further back view ..? That could be creating a heat pocket or lack of airflow 

 

 

 

The instructions say to mount at eye level and in a 35mm back box.

[dpmiller]

perhaps calibration is all that is required as the self-heating offset is going to be a pretty stable thing I'd guess...

 

/ I'd opine that a good old-fashioned mechaninical stat with accellerator could very well read higher than ambient for the same reason.

[SteamyTea]

33 minutes ago, dpmiller said:

perhaps calibration is all that is required as the self-heating offset is going to be a pretty stable thing I'd guess

Probably not.

Heat loss and gain is not linear.

[Ultima357]

The one in the photo is close to a worktop in the kitchen but they all exhibit the same issue. Take note though of the calibrated thermometer, the difference this reads between static and fanned air is just 0.2deg whilst the stat records 1.4deg. The self heat issue is a problem though as if you walk past a stat, the change in temperature easily triggers a call for heat and hence we do not get good hysterisis on the room temperature. Eg, sitting in our family/dining room which is 8.5 x 4m can feel chilly whilst the stat is sitting there reading around 23 deg. Place a thermometer by where your sitting or anywhere in the room and you find it's just around the 21 deg mark. Give a brief fan of air at the stat and it falls below 22 deg, so simply not achieving the right temperature. They are set to 0.5deg differential but are only really doing something like 2 degrees. And yes, we like the temperature at 23 deg, that's why we built passive. It only costs around £2 a day to keep it warm, all 253m2 of living area and warm attic space too ?. 

[Ultima357]

10 hours ago, Temp said:

Mine are 230V but most are in surface mount back boxes. The stats don't have massive vents but never seen this issue. They seem to read pretty accurately.

 

If anything the one we have with the remote sensor is the least accurate and I made the remote sensor myself.

These, being a new build are all flush as you'd expect. They unfortunately have to be at 1200mm from floor level maximum these days to meet disability requirements in building regs, same as wall plugs have to be at least 450mm up from the floor. My remote ones use the Heatmiser housing and work fine. 

[ProDave]

14 minutes ago, Ultima357 said:

These, being a new build are all flush as you'd expect. They unfortunately have to be at 1200mm from floor level maximum these days to meet disability requirements in building regs, same as wall plugs have to be at least 450mm up from the floor. My remote ones use the Heatmiser housing and work fine. 

All my thermostats are more like 1.5M above floor level and BC have not (yet) raised any issues with this. 

 

But I still think your best bet is change for a different thermostat.  A thermostat that is using enough energy to self heat itself a few degrees is both wasteful and a flawed design.

[AliG]

I have over 20 of these thermostats and don't have any issues as described, but I have a few questions and they do have their quirks.

 

Did you just urn off the MVHR to see what happens with the thermostats? They do generate a little heat, but there would be an assumption that air movement within the house would dissipate this. With your exceptionally low ACHR figure this would not happen and I could see why they would warm up. If however, the MVHR is never turned off normally then what happens in this situation is less relevant.

 

I am not sure if the thermostats will be impacted by being as close to a surface as shown in the picture? Also is there some kind of phone on the surface below the thermostat or was that just something you were using whilst testing them? That might be raising the temperature in a relatively sheltered area.

 

The big drop you show in the dining room temperature - is the stat near the door? The stat in our bedroom often drops a couple of degrees in the morning as it is somewhat cooler in the hall outside the bedroom and so when you open the door the temperature it reads changes changes.

 

I can see why you would want to calibrate the thermostats to the correct temperature, but I just accept that different thermostats in different rooms read slightly differently depending on where they are in the room. Different rooms also feels comfortable at different air temperatures depending on the solar gain and the flooring. So I adjust the setting on the thermostat relative to how the room feels to me. For some rooms I have them set at 20.5, for some rooms I have them set at 22C. I don't really care if the temperature is exactly that I just want them running the heating at a comfortable temperature. I have the same issue moving between different cars - 22C may be comfortable in one car, too hot in another and too cold in another so I change the climate control setting to the comfortable setting irrespective of what the actual number is.

[Ultima357]

12 hours ago, PeterW said:

Sadly this confirms most of my opinions about the quality of the products under that brand. 
 

Have you used the Heatmiser wiring centre and their actuators ..?

Peter. Just wondering if you have any recommendations of a better solution? Bearing in mind that it's 14 zones and Internet connected. 

[Onoff]

53 minutes ago, Ultima357 said:

Peter. Just wondering if you have any recommendations of a better solution? Bearing in mind that it's 14 zones and Internet connected. 

 

How cheap do you want, any good?

 

[£23.28]MK72GA Smart Water Heating Thermostat WIFI LCD Touch Screen Temperature Control Regulator for Water Heating System Measurement & Analysis Instruments from Tools on banggood
https://banggood.app.link/m5kUTOr88cb

[Ultima357]

In reply to AliG, the issue with the MVHR is just that it demonstrated that the stats are guilty of recording their own temperature and do not have adequate airflow though the housing to be reliable. I get what you say about ignoring what they say and just get it comfortable but thermistor based digital sensing should not be this inaccurate. The dining one is near enough in the middle of the house, well away from the door to the hall. Our master bedroom one and all the others show similar results. When you look at the design of these stats, the thermistor is on the bottom edge, just below the tick symbol where the vents are. But there are no other vents, so any build up of heat has to come out of these vents or be disappated through the plastic body. The latter isn't much use of course. The remote housings are very well vented and do not exhibit this problem. QED. So therefore my findings hold true, they sense their own temperature and are very poor in sensing actual room temperature due to the enclosure and a brief puff of air just makes them erratic. 

[Ultima357]

3 minutes ago, Onoff said:

 

How cheap do you want, any good?

 

[£23.28]MK72GA Smart Water Heating Thermostat WIFI LCD Touch Screen Temperature Control Regulator for Water Heating System Measurement & Analysis Instruments from Tools on banggood
https://banggood.app.link/m5kUTOr88cb

They look like they'd have the same issue, poorly vented housing. 

[mvincentd]

I have 6 of these v2's 240v spread between 2 heatmiser wiring centres.   They're set at the default 1 degree drop to trigger so right now i'm setting 23 degrees (uncalibrated).  The only measurement of temperature aside from the v2's I have is from the mvhr....they generally only disagree by around 0.5 degree and like @AliG i'm not concerned with the figure itself just my warmth perception relative to it.

I don't get the issues discussed here but agree the bloody things aren't fit for purpose;  The V2 auto backlight comes on of its own accord randomly in the middle of the night.  I've caught 4 out of 6 doing it so far.  The light then stays on for an inconsistent period that doesn't accord with the auto-backlight function.  There is no viable trigger in the house for this, i've consulted with Heatmiser technical, head scratched it with the sparky, and now removed the two which keep waking me up in the night.  Heatmiser aren't very interested and being only 6 i've decided to get on with my life rather than have a fight.  So i've been searching for V1's which only differ by the lack of auto backlight and the backlight's colour.  They can snap onto the same backplate so a fuss free transition.......if I can find some....not yet!

I will however be going to war with Heatmiser if I have to fit different backplates and stats from a different brand to the wiring centres.

My other issue is that with a power cut they fail to retain programmed settings and time, which they should do.

[Ultima357]

Hi mvincentd. Yes we too have had a couple of instances where they self illuminate. How often they do it I don't know, it just depends if you happen to be facing them and only half asleep at the time. I've taken one fully apart, it's just the 4 little screws holding the pcb in so quite easy. I suppose you can put some nail varnish or tape over the leds if you want to. 

Having just come off the phone to Heatmiser and speaking with the guy who designed them, we went round in circles a lot, but basically he agrees that they a) do generate heat and b) this affects the sensor. This apparently is taken into account in the software algorithm and once calibrated, should be fine. But we couldn't really agree that this shouldn't be happening and that there wasn't enough air flowing through the housing to properly register room temperature. He claimed that my insulated stud walls were the problem but I pointed out that in all new builds, stud walls have to be acoustically insulated according to building regulations and therefore the stat has to work with these.

So conclusion. All modern 230v stats with digital displays must generate heat because they have an inbuilt psu. Therefore any internal thermistor sensor is compromised unless there is very good air flow and a larger housing is therefore required. The modern sleek look just doesn't work and only solution is a remote sensor. We did agree on the fact that a remote sensor would be better at least. 

[AliG]

1 hour ago, Ultima357 said:

He claimed that my insulated stud walls were the problem but I pointed out that in all new builds, stud walls have to be acoustically insulated according to building regulations and therefore the stat has to work with these.

 

This probably explains it. Almost all of mine are in dot and dab plasterboard with a small cavity behind it which would allow heat to dissipate. If the back box is surrounded by insulation then they would probably get somewhat warmer. I have one which I have noticed consistently reads 1C higher than I reckon it should, funnily enough also in the dining room.

 

The question is, does the extra heat actually stop them reading the temperature at all or just raise it relative to the room? As long as the temperature reading still rises and falls with the room temperature then you should be able to recalibrate it or set a higher target temperature. If not and they are just heated by the internal heat generation then you do have a problem.

 

Could you create a little space around the backboxes inside the wall?

 

I find mine sensitive enough that I have changed the sensitivity from 1C to 0.5C as I don't like the temperature falling 1C before the heating kicks in.

[Temp]

 

10 minutes ago, AliG said:

The question is, does the extra heat actually stop them reading the temperature at all or just raise it relative to the room? As long as the temperature reading still rises and falls with the room temperature then you should be able to recalibrate it or set a higher target temperature.

 

You would expect them to track room temperature but only if the power they dissipate is constant (probably true) and the thermal resistance between inside of stat and the room is constant. Sounds like the latter it isn't true due to changing air flow from the MHVR.  In other words they can only be calibrated with MVHR either on or off not both.

 

[Ultima357]

13 minutes ago, AliG said:

 

This probably explains it. Almost all of mine are in dot and dab plasterboard with a small cavity behind it which would allow heat to dissipate. If the back box is surrounded by insulation then they would probably get somewhat warmer. I have one which I have noticed consistently reads 1C higher than I reckon it should, funnily enough also in the dining room.

 

The question is, does the extra heat actually stop them reading the temperature at all or just raise it relative to the room? As long as the temperature reading still rises and falls with the room temperature then you should be able to recalibrate it or set a higher target temperature. If not and they are just heated by the internal heat generation then you do have a problem.

 

Could you create a little space around the backboxes inside the wall?

 

I find mine sensitive enough that I have changed the sensitivity from 1C to 0.5C as I don't like the temperature falling 1C before the heating kicks in.

Well it sort of explains it and yes, you can calibrate the stats BUT.... my issue is that with a fully sealed passive house, merely stirring the air as you walk past can change the reading by 0.5 deg because in static air they are reading their own temperature. This little bit of air flow then gets it to read the actual room temperature. Rather than bodging the temperature reading by effectively calibrating the stats incorrectly, they should have a better designed housing that allows sufficient air flow up through the device to measure correctly. I could try to create a space behind them but in reality the back boxes are steel, screwed to noggins. Moving the insulation away a bit would just create a slightly larger space to heat. This probably would change the offset a little but wouldn't make them work as they should. I have drilled some extra holes in the top side edge of one and this reduces the self read offset to about 0.6deg. Probably needs another 6 across the top to be right! 

[Temp]

Here is a photo of the remote sensor I built.  Just a small project box from Maplin with three holes drilled top and bottom and an NTC inside. Cant remember if its a 10k or 20k Ohm but it was in the manual for the stat somewhere.

 

 

and this is the old Heatmiser stat we have showing the vent holes. Same both sides and similar on the bottom. None on the top - probably to keep dust out?

 

 

[Ultima357]

9 minutes ago, Temp said:

 

 

You would expect them to track room temperature but only if the power they dissipate is constant (probably true) and the thermal resistance between inside of stat and the room is constant. Sounds like the latter it isn't true due to changing air flow from the MHVR.  In other words they can only be calibrated with MVHR either on or off not both.

 

MVHR can run at 4 different levels too. Comes back to the fact that the MVHR has no effect on the ones with remote sensors, so the ones without are reading/being affected by their own heat source.