Thermostat has a own life

Tips, tricks and frequently asked questions
minkemans
Posts: 2
Joined: 23 Jan 2017, 08:54

Thermostat has a own life

Post by minkemans » 23 Jan 2017, 09:05

2017-01-23 (1).png
2017-01-23 (1).png (155.11 KiB) Viewed 879 times
Since five weeks i have a Netatmo thermostat. But the last two weeks something wrong. The asked temperature is reached but after that the temperature fall back to a lower temperature. Who have a solution?

runemann
Posts: 59
Joined: 16 Oct 2014, 14:58

Re: Thermostat has a own life

Post by runemann » 24 Jan 2017, 12:50

I figure it could depend on the algorithm you are using, and your schedule. If say (referring to the chart you attached) you want to have the temperature at 21 degress until 09 o'clock and then you want 17 degrees after that (say because you go out for work or something), the Netatmo could set the boiler to "off" at a time prior to 09, eventhough the degrees are less than 21, in order to adapt to the new demperature period coming up. So in effect the Netatmo could set the boiler to ON eventhough the temperature is above the current period target (e.g. when preparing for a higher temperature at dinner time) and could set the boiler to "OFF" eventhough the degree is lower than the current target (as mentioned above).

I find that the PID algorithm works fine. Typically late evenings I can see that Netatmo stops the heating before the eventing-end-time so that the temperature smoothly lowers into the cooler night period. For work days I use i) a night period 23 - 06,. breakfast period ca. 06-08, work period 08 - 16, and evening 16 - 23. The heating takes some time (in my house 1 to 2 degrees an hour, depending on the outside temperature), and having many small period could mess up things I guess.

I hope this makes sense!

minkemans
Posts: 2
Joined: 23 Jan 2017, 08:54

Re: Thermostat has a own life

Post by minkemans » 25 Jan 2017, 11:54

Bestand_000.png
The problem is thatthe temperature fall down after it has reach the 20 degrees. The problem is that the temperature is often belows the asked temperature. You can see it in the attachment and I don't think the algoritm is the problem. The first two weeks I was satisfied with the Netatmo-thermostat. After an updata o f the firmware it was wrong, The firmware now is thermostat 40 relaistation 71
(For a Dutchmen it's difficult to write English)

PhilHornby
Posts: 87
Joined: 26 Oct 2016, 14:51
Location: Devon, UK
Contact:

Re: Thermostat has own life

Post by PhilHornby » 25 Jan 2017, 23:46

Your title actually says it all...

The issue is that the Netatmo suffers from huge thermal lag. See this thread: viewtopic.php?f=20&t=7373

When I compared the actual room temperature against what the Netatmo reports, I found huge discrepancies. This is especially the case when the temperature is changing rapidly. The Netatmo software guys have actually done a really good job, getting this useless hardware to function at all!

The Netatmo produces beautiful graphs - but they're largely meaningless :(

Michael0610
Posts: 41
Joined: 01 Aug 2016, 15:02

Re: Thermostat has a own life

Post by Michael0610 » 26 Jan 2017, 05:33

PhilHornby: But I have 2 Netatmo stats working perfect, compared up against a logger, AND a thermometer.

One in my house with fast changing in themperature, and one where themperature changes is very slow.

I have both on 21, and the logger and thermo show between 20,9 and 21,1, all time. So my Netatmo's are working perfect...............

runemann
Posts: 59
Joined: 16 Oct 2014, 14:58

Re: Thermostat has a own life

Post by runemann » 26 Jan 2017, 11:17

Ref. PhilHornby and Michael0610 my thermostat is also working perfectly, I have it fixed on the wall ca. 1.75 meters high, and I have tested i.e. checked the temperature vs. two other temp. measurement sources, and all is ok - they show the same degree +/- some few decimal parts of a degree. The Netatmo has been doing an excellent job since late 2014 I believe it is.

Ref. the charts, I don't see the meaningless argument - they work fine and show what they are supposed to show: the bars indicate the amount of time (percentage) the boiler/Netatmo is in ON mode (relative to the zooming you choose), and lines show the target vs. actual temperature.

The Dutch guy: my point is that the PID alogithm sort of things for itself sometimes, so you could be in for some (albeit small) surprises depending on yuor schedule. If you use the hysteresis mode* on the other hand, and say a 0.3 degree threshold, then if the temperature drops to below 19.7 degrees (20 - 0.3) then the thermostat/boiler should go to on mode - does it, or does it not? So you could simply try to set the mode to hysteresis and check if the thermosta fires up the boiles when the temp drops to 19.7 degrees. It should fire up and continue heating (red bars in chart) until the temp reaches 20.3 degrees (upper threshold band).

On one of your charts there it seems like the boiler is ON but the actual temperature is not increasing, either there is the usual lag (hard to see the time frame / zoom) or maybe it is very cold and the boiler cannot get the temperature up, or maybe boiler and/or the Netatmo doesn't work. To check the Netatmo is simple: look for the blue or white light on the relee box, or maybe check that the circuit is closed or not (the signal sent from the Netatmo relee). I use a small electrical bulb to check that.

* https://phc.parts/userfiles/Product_dat ... 50-FAQ.pdf

almoko
Posts: 1
Joined: 26 Jan 2017, 11:18

Re: Thermostat has a own life

Post by almoko » 26 Jan 2017, 11:23

I've just installed my Netatmo thermostat connected with the Valliant gas boiler.\

I find that it is the boiler that has an own life. Meaning, if the boiler is "off" and Netatmo sends command to turn on, the boiler does turn on, but stops heating prior to achieving the desired room temperature, while Netatmo is still showing "Boiler On" status.

Could the boiler stop heating based on achieving the *water* temperature (it's own internal setting) vs. Netatmos' command to keep heating?

runemann
Posts: 59
Joined: 16 Oct 2014, 14:58

Re: Thermostat has a own life

Post by runemann » 26 Jan 2017, 11:49

Could be, in the beginning for maybe a week I was struggling a bit because I had to "intelligent systems" trying to do the same thing: the Netatmo, and a Danfoss controlling unit ("ECL", Danfoss ECL 110). The Danfoss gives you the option to program daily heating schedules and could do the whole job, but it's a last century thing (no app nor wfi connection, i.e old school and insufficient).

Luckily I found some settings in the ECL so that it became "100% slave" to whatever closed-or-open circuit signal that was being fed into it. So now the ECL is just controlling a valve (in effect telling the value to open or close mechanically), and all the thinking is going on in the Netatmo unit.

I use an electrical water heater combi unit though, and I am not sure if there's security issues with your gas boiler or something, some internal settings that tells it not to burn gas forever or something, could you check that

PhilHornby
Posts: 87
Joined: 26 Oct 2016, 14:51
Location: Devon, UK
Contact:

Re: Thermostat has a own life

Post by PhilHornby » 26 Jan 2017, 15:23

Michael0610 wrote:PhilHornby: But I have 2 Netatmo stats working perfect, compared up against a logger, AND a thermometer.
Great! ... please publish your data and 'logger' details.

PhilHornby
Posts: 87
Joined: 26 Oct 2016, 14:51
Location: Devon, UK
Contact:

Re: Thermostat has a own life

Post by PhilHornby » 26 Jan 2017, 15:36

runemann wrote:Ref. the charts, I don't see the meaningless argument - they work fine and show what they are supposed to show: the bars indicate the amount of time (percentage) the boiler/Netatmo is in ON mode (relative to the zooming you choose), and lines show the target vs. actual temperature.
I wasn't questioning the way it presents the data ...I was questioning the actual data itself... its plot of temperature vs time does correlate with reality (as I measured it).
He also wrote:If you use the hysteresis mode* on the other hand, and say a 0.3 degree threshold, then if the temperature drops to below 19.7 degrees (20 - 0.3) then the thermostat/boiler should go to on mode - does it, or does it not? So you could simply try to set the mode to hysteresis and check if the thermostat fires up the boiler when the temp drops to 19.7 degrees. It should fire up and continue heating (red bars in chart) until the temp reaches 20.3 degrees (upper threshold band).
From the measurements I took, the Netatmo isn't capable of implementing a 0.3 degree threshold. I set mine to 0.1°threshold and it actually delivered ±0.6°C. (Needless to say, it thought it had managed better than that...)

Post Reply

Return to “General questions”