Evohome controller bug during Heating Off, but great job Dan and Domoticz

[paulockenden]

Turns out that the radiator that had been turned up coincidentally also suffered a comms failure with the controller. But was able to turn the boiler on.

I don't see how that's possible. The HR92 talks to the controller, which in turn talks to the BDRs.

Of course (with HGI80s or equivalent) it IS possible for the HR92 to send a message which the controller see but Domoticz misses.

P.

[bruce_miranda]

I too am scratching my head about this one. All I can tell you is what I saw. On Domoticz there was 0 demand from any zone, yet the CH Valve Relay was showing 100%. So something was requesting heat but Domoticz couldn't see it and ofcourse with the controller being in Heating Off mode, it wasn't visible at all. I am going back today to study the graphs to see if I can explain this behaviour.

[bruce_miranda]

Capture.JPG

OK here is what seems to have happened. The Snug is the room that caused the boiler to come on. It is the only room in the house that had a Heat Demand at the time. What is strange is that the reported Set Point seems to stay at the 5C (Controller Heating Off) but clearly it wasn't. You can see the room temperature rising in the time period that the boiler has been fired. The 5C set point is reported both by the controller back to the Honeywell servers as well as to the HGI80. It could just be a bug with the HR80 where a local over-ridden set point is not registered by the Controller when Off, but the Heat Demand message can fire the boiler, as though it's acting independently.

What was even more coincidental was this Snug zone had a comms failure in the time period before and after this Heating period. But not during, as I previously incorrectly reported. I confirmed this via the Fault Log. @Dand we really need to work out how to pull the Controller fault log into Domoticz, so that we immediately be notified of a problem. I know this is possible, but will Honeywell tell us the command to use - probably not!

[DBMandrake]

[ATTACH=CONFIG]1100[/ATTACH]

OK here is what seems to have happened. The Snug is the room that caused the boiler to come on. It is the only room in the house that had a Heat Demand at the time. What is strange is that the reported Set Point seems to stay at the 5C (Controller Heating Off) but clearly it wasn't. You can see the room temperature rising in the time period that the boiler has been fired. The 5C set point is reported both by the controller back to the Honeywell servers as well as to the HGI80. It could just be a bug with the HR80 where a local over-ridden set point is not registered by the Controller when Off, but the Heat Demand message can fire the boiler, as though it's acting independently.

Set point and heat demand are completely independent.

I think you'll also find that when the controller is in Heating Off mode and you manually turn up an HR92's set point that the controller despite showing an off symbol on the display reports that as 5 degrees back to the Honeywell servers regardless of what the real set point is on the HR92. (It ignores the set point but still accepts the heat demand)

As for the HGI80 - that depends whether it is checking the set point from the controller or from the HR92 directly, (Dan would know the answer to that...) the HR92 would report the true set point but again I think you'll find in Heating Off mode the controller will only report 5 degrees for the set point regardless of what the HR92 is set to, hence your graph.

What was even more coincidental was this Snug zone had a comms failure in the time period before and after this Heating period. But not during, as I previously incorrectly reported. I confirmed this via the Fault Log. @Dand we really need to work out how to pull the Controller fault log into Domoticz, so that we immediately be notified of a problem. I know this is possible, but will Honeywell tell us the command to use - probably not!

In my testing (by pulling the batteries) if an HR92 goes unexpectedly offline while it is currently demanding heat, the controller maintains the same heat demand for that zone for approximately 30-60 minutes before assuming the HR92 went AWOL and setting the implied heat demand for that zone back to zero.

As for your problem with that particular zone - I'd wager that the problem was a low battery or intermittent battery connection that resulted in the HR92 rebooting and/or going offline for a while, rather than an RF problem.

If you reboot the HR92 it goes back to 20 degrees by default and will stay there until the next schedule change or update from the controller - which could be quite a while. During this time it may call for heat especially if it was night time and it was cold, but because you were in heating off mode you would not be aware it's set point was not really 5 degrees.

It's too much of a coincidence to have both lost comms and an unexplained call for heat around the same time - it's got to be the HR92 losing power and/or rebooting and reverting to the default 20 degree set point for a while.

I'd try re-tensioning the bottom battery contacts and maybe fit some new batteries - I had trouble with my bathroom HR92 once where even though the batteries were good it was sometimes spontaneously dying (screen going completely blank as well) when it tried to turn the motor, and it was just a poor battery contact.

[bruce_miranda]

Let me check the batteries again. But the not reporting the HR80s set point is just plain dumb. It's like the controller saying, Don't talk to me, I ain't listening but check if the boiler will do what you are asking!

[bruce_miranda]

I just realised that a TR87 setup as a temperature sensor in a room cannot perform a similar override while in Heating Off mode. We have an UFH zone that uses a TR87RF as the temperature sensor and a BDR91 as the actuator. Even if I push the button on the BDR91, the light comes on, but it won't fire the boiler via OT.

[bruce_miranda]

The situation gets more interesting in a zone with a TR87 as temp sensor and a normal radiator valve as the actuator. Again overrides on the TR87 don't work. But if you over ride on the HR80 itself, it will fire the boiler. But now if you check the setpoints, the radiator valve shows the over-ridden value but the TR87 shows the Heating Off value of 5C.