New with Evohome. Is this a bug?

[arras2]

I have this system:

-Gas Boiler
-Evotouch Controller. In the option "Boiler Demand" of the Evotouch Controller I have configured the option "Boiler Relay", binded with a BDR91 relay which is wired directly to the boiler.
-RFG100 mobile Access Kit.
-8xHR92 valves.
-7 configured zones. 6 zones have a rad controlled by a HR92 valve. As temperature sensor I use the HR92. In the seventh area there are 2 rads controlled by two HR92 valves and as temperature sensor I use the internal sensor of the Evotouch controller.

Everything works perfectly. The boiler starts when any of the zones needs more heat. The HR92 opens and closes correctly at the assigned times, even the optimisations works correctly.

But there is something that doesn’t work and it makes no sense. I have configured that at 00:00, evotouch assigned all the setpoint temperatures to 5 degrees (the zones are now around 15-20 degrees). The problem is that the relay keeps starting and stopping during 30 or 40 minutes. This also happens if I use the quick action “Heating Off”.

Yesterday I checked and just before 00:00, every 10 minutes the boiler started during 5 minutes 20 seconds and stopped during 4 minutes and 40 seconds. At 00:00 approximately, evotouch assigned all the setpoint temperatures” to 5 degrees. The cycle of 5 minutes 20 seconds of the boiler working and 4 minutes 40 second of it stopped, happened 4 times. I checked all the HR92, and they were all completely closed (Option 10, "valve position” = 00).

¿Who demands that heat? ¿Why was my boiler wasting that gas during 21 minutes and 20 seconds if all the rads were completely closed?

I have restarted and rebindend everything four times.

[Rameses]

I have this system:

-Gas Boiler
-Evotouch Controller. In the option "Boiler Demand" of the Evotouch Controller I have configured the option "Boiler Relay", binded with a BDR91 relay which is wired directly to the boiler.
-RFG100 mobile Access Kit.
-8xHR92 valves.
-7 configured zones. 6 zones have a rad controlled by a HR92 valve. As temperature sensor I use the HR92. In the seventh area there are 2 rads controlled by two HR92 valves and as temperature sensor I use the internal sensor of the Evotouch controller.

Everything works perfectly. The boiler starts when any of the zones needs more heat. The HR92 opens and closes correctly at the assigned times, even the optimisations works correctly.

But there is something that doesn’t work and it makes no sense. I have configured that at 00:00, evotouch assigned all the setpoint temperatures to 5 degrees (the zones are now around 15-20 degrees). The problem is that the relay keeps starting and stopping during 30 or 40 minutes. This also happens if I use the quick action “Heating Off”.

Yesterday I checked and just before 00:00, every 10 minutes the boiler started during 5 minutes 20 seconds and stopped during 4 minutes and 40 seconds. At 00:00 approximately, evotouch assigned all the setpoint temperatures” to 5 degrees. The cycle of 5 minutes 20 seconds of the boiler working and 4 minutes 40 second of it stopped, happened 4 times. I checked all the HR92, and they were all completely closed (Option 10, "valve position” = 00).

¿Who demands that heat? ¿Why was my boiler wasting that gas during 21 minutes and 20 seconds if all the rads were completely closed?

I have restarted and rebindend everything four times.

Prior to the setpoint command of 5 degrees, the system had a 50% (appox) demand - which means a rad/area/zone was calling, as it should. When you initiated the command to close, (was this manual? not via the evohome controller), the HR92s closed, and all sent message back I am closed - but one/some didn't get through, either because of message clash or poor signal? evohome does 'catchup' with a 10 minute resend if it hasn't heard/or the Hr92 has not had a handshake. evohome attempt a co-ordinated and controlled communications process, where possible.

I would be more worried if this happens again and off a schedule. In which case we need to look at signal strengths in more detail.

[arras2]

Prior to the setpoint command of 5 degrees, the system had a 50% (appox) demand - which means a rad/area/zone was calling, as it should. When you initiated the command to close, (was this manual? not via the evohome controller), the HR92s closed, and all sent message back I am closed - but one/some didn't get through, either because of message clash or poor signal? evohome does 'catchup' with a 10 minute resend if it hasn't heard/or the Hr92 has not had a handshake. evohome attempt a co-ordinated and controlled communications process, where possible.

I would be more worried if this happens again and off a schedule. In which case we need to look at signal strengths in more detail.

The command to close was via the evohome (Evohome is programmed to set 5 degrees in all zones at 00:00).

My house is small. I have a very good signal strengths (5) with all HR92.

If I can, today, I will make a video.

Thank you.

[Rameses]

The command to close was via the evohome (Evohome is programmed to set 5 degrees in all zones at 00:00).

My house is small. I have a very good signal strengths (5) with all HR92.

If I can, today, I will make a video.

Thank you.

Do you have stored hot water? What sort of valve system do you have?

[arras2]

I don't have hot water stored. It's a simple monotube installation. My installation is like this:

esquema-monotubular.jpg

Yesterday the system worked fine! The optimum stop activated itself at 23:45. At that time, the temperature of all zones were above the setpoint temperature and the boiler did not start again.

Today I'll watch the system again.

Thanks.

[Rameses]

How long has your system been in?

[arras2]

How long has your system been in?

I installed it a month and a half ago. But I haven't used it a lot because it wasn't very cold.

One question. For example, if after 3 hours demanding heat, I activate the quick action "Heating Off", the system shouldn't demand more heat right?

I'm going to try something. Today I'll let the system put the setpoint temperatures of all the zones to 5 degrees. And if the system goes well, another day I'll try to activate the heating off manually.

[Cchris]

Was the boiler actually firing - or was it just circulating? i.e. pump overrun.

[arras2]

Was the boiler actually firing - or was it just circulating? i.e. pump overrun.

firing and circulating.




Yesterday I went to sleep early and I used the quick action "heating off" and it also worked. From the moment I activated it, the boiler didn't work again.

Until now, I have been doing the same thing I explained before and the system kept starting. And now, in the last two days that hasn't happened again.

I will keep watching to see if it happens again.

Thanks.

[arras2]

Hello,

I have been using the system since last year and the same thing keeps happening. It must be a bug, because I have tried everything and the system works perfectly except for what I told you before.

I have tried to find the pattern that is causing this, but I haven't found it. It happens if I stop it manually before its time, if I schedule all the areas automatically to 5 degrees, with the optimum stop activated, deactivated, if I do it from EvoHome, from the mobile app...

Everything is closed and approximately in 20% or 30% of the days there are between 2 and 4 cycles more than there should be, with all the valves closed.

This season I have written down in an excel document the days and hours when the system failed, in case something could be veryfied in the servers about why this is happening (my system is connected to the internet).

My system is turned on approximately 200 days a year. If this bug occurs 20% of the days, with 3 extra cycles on average, and with a heating request of 50%, this causes the boiler to be firing for 10 hours (600 minutes), which equates to 20m^3 of gas. It is not too much, but I would like to know why it does that and if it can be avoided.

[mylesm]

Hello,

I have been using the system since last year and the same thing keeps happening. It must be a bug, because I have tried everything and the system works perfectly except for what I told you before.

I have tried to find the pattern that is causing this, but I haven't found it. It happens if I stop it manually before its time, if I schedule all the areas automatically to 5 degrees, with the optimum stop activated, deactivated, if I do it from EvoHome, from the mobile app...

Everything is closed and approximately in 20% or 30% of the days there are between 2 and 4 cycles more than there should be, with all the valves closed.

This season I have written down in an excel document the days and hours when the system failed, in case something could be veryfied in the servers about why this is happening (my system is connected to the internet).

My system is turned on approximately 200 days a year. If this bug occurs 20% of the days, with 3 extra cycles on average, and with a heating request of 50%, this causes the boiler to be firing for 10 hours (600 minutes), which equates to 20m^3 of gas. It is not too much, but I would like to know why it does that and if it can be avoided.

Well if everything is closed then there would be very little gas used

[DBMandrake]

Some thoughts -

1) Keep in mind that when you use the off action on the controller or schedule all zones to 5 degrees at the same time there can be a delay of up to 5 minutes before the heating demand ceases. This is because set point updates are only sent from the controller to HR92's about every 4 minutes, and it takes over half a minute for an HR92 to fully close and send a new heat demand. During this period of time the boiler relay may remain on or cycle back on. This is perfectly normal and expected behaviour, however if the relay is still cycling on after 10 minutes this is not normal and indicates a problem.

2) You have the older non-Wifi controller (since you have the RFG gateway) and this has older firmware that does not support local override display.

What this means is that if you directly adjust an HR92 this set point is not reflected on the controller. This can cause an issue if you have multiple set points in a row in your schedule that are the same.

For example my bathroom is scheduled to go from 16 to 5 at 8pm and scheduled for 5 again at midnight. The reason for this is I want it to go off at 8pm but if someone manually turned it up between 8pm and midnight (for example a late shower) I want to ensure it is off again at midnight if someone forgot to turn it back down.

This works OK with the wifi controller because the controller is aware of any manual overrides made at the HR92, however on the older controller like yours it is not. So if you turned it up manually at 10pm not only would this not be displayed on the controller, the midnight set point change to 5 degrees would NOT be sent by the controller because it believes the HR92 is already set to 5 degrees. This means the HR92 will still be at whatever manual set point it was adjusted to, calling for heat, despite the controller saying all zones are set to 5 degrees or the system is "off". (HR92's can still call for heat from the boiler in controller "off" mode if they are turned up manually!)

Call it a bug, call it a limitation, either way this can cause there to be an actual heat demand even if you have used the off action on the controller.

To confirm if this was happening you'd have to physically go around to every HR92 and verify on the display that they are in fact set to 5 degrees. A workaround is to make sure you never schedule two identical set points in a row. For example if you had multiple 5 degree set points in a row as I described intended to turn a zone off, alternate them between 5 and 5.5 degrees so that there is always a change for the controller to be sent. This doesn't help for the off action though as the off action is always the same set point. (By default 5.0)

3) If the above has been ruled out then unfortunately it does sound like lost heat demand messages between devices. Either between individual HR92's and the controller, or between the controller and your BDR91 boiler control relay.

I have a relatively small bungelow with the controller only about 6 metres from the boiler closet with three BDR91's (including hot water control) and everything always reports 5/5 excellent signal. However I occasionally see boiler "run on" like you describe, where there is no heat demand but the boiler relay (or sometimes the heating zone valve relay) cycles on and off for a further 20 minutes beyond what I would expect.

The cause of this is a lost heat demand message from controller to BDR91. Normally a heat demand message is sent from the controller to the relay either when there is a change in heat demand, or every 20 minutes if the heat demand stays the same.

So you turn the heating off and then a few minutes later when a zero heat demand is sent to the BDR91 it is for some reason not received, so the BDR91 continues on whatever duty cycle it was previously following, say 5 minutes on 5 minutes off. With no communication from the controller the relay will continue following the last know duty cycle for 45-60 minutes before lighting the red fault LED.

After 20 minutes another heat demand will be sent and if this gets through the relay will at that point go off. So if you see the boiler running on for an additional 20 minutes past the normal up to 5 minute delay then this is probably what is happening. If the second message also didn't get through then it would run on for 40 minutes. If the third message also didn't get through you'd see the red error LED light on the BDR91 and probably get a fault log entry on the controller.

If this is your problem then about the only thing you can do is either try to identify any sources of 868Mhz interference, (wireless alarms, Loop, other similar wireless devices) or relocate the BDR91 - even moving it a couple of feet or re-orientating it (on a different wall, if its in a boiler closet) might solve the issue.

Unfortunately the signal test function seems all but useless at diagnosing these type of problems. I had a communication problem with my CS92 hot water sensor where the signal test was fine but I kept getting hot water overshoot due to temperature readings not getting through reliably. In the end I moved it to another wall in the boiler closet (90 degree orientation change as well as moving a couple of feet) and even though the signal is now 4/5 instead of 5/5 it has largely fixed the problem. I still get an overshoot but only very rarely maybe only a couple of times a month instead of a few times a week.

If it's practical I would try moving your BDR91. Can you post a picture of where it is in relation to other objects ?

Another possibility is poor communication from an HR92 back to controller - which could cause the same sort of issue of lost heat demand messages but only when a specific zone is being scheduled off. I don't know what the retry schedule of an HR92 is and whether it is the same 20 minutes as controller to BDR91 but plan to test it soon.

[Rameses]

One more thing that could possibly cause this. If the boiler BDR is double bound (BDRs can have multiple sources of control) (or that the one that came with the Kit you used elsewhere in the home and then bound again) . You probably already did this - but I assume you did a full wipe of the BDR when you got it (as they come pre-bound and this is a common error as people rebind again)

[arras2]

Some thoughts -

1) Keep in mind that when you use the off action on the controller or schedule all zones to 5 degrees at the same time there can be a delay of up to 5 minutes before the heating demand ceases. This is because set point updates are only sent from the controller to HR92's about every 4 minutes, and it takes over half a minute for an HR92 to fully close and send a new heat demand. During this period of time the boiler relay may remain on or cycle back on. This is perfectly normal and expected behaviour, however if the relay is still cycling on after 10 minutes this is not normal and indicates a problem.

2) You have the older non-Wifi controller (since you have the RFG gateway) and this has older firmware that does not support local override display.

What this means is that if you directly adjust an HR92 this set point is not reflected on the controller. This can cause an issue if you have multiple set points in a row in your schedule that are the same.

For example my bathroom is scheduled to go from 16 to 5 at 8pm and scheduled for 5 again at midnight. The reason for this is I want it to go off at 8pm but if someone manually turned it up between 8pm and midnight (for example a late shower) I want to ensure it is off again at midnight if someone forgot to turn it back down.

This works OK with the wifi controller because the controller is aware of any manual overrides made at the HR92, however on the older controller like yours it is not. So if you turned it up manually at 10pm not only would this not be displayed on the controller, the midnight set point change to 5 degrees would NOT be sent by the controller because it believes the HR92 is already set to 5 degrees. This means the HR92 will still be at whatever manual set point it was adjusted to, calling for heat, despite the controller saying all zones are set to 5 degrees or the system is "off". (HR92's can still call for heat from the boiler in controller "off" mode if they are turned up manually!)

Call it a bug, call it a limitation, either way this can cause there to be an actual heat demand even if you have used the off action on the controller.

To confirm if this was happening you'd have to physically go around to every HR92 and verify on the display that they are in fact set to 5 degrees. A workaround is to make sure you never schedule two identical set points in a row. For example if you had multiple 5 degree set points in a row as I described intended to turn a zone off, alternate them between 5 and 5.5 degrees so that there is always a change for the controller to be sent. This doesn't help for the off action though as the off action is always the same set point. (By default 5.0)

3) If the above has been ruled out then unfortunately it does sound like lost heat demand messages between devices. Either between individual HR92's and the controller, or between the controller and your BDR91 boiler control relay.

I have a relatively small bungelow with the controller only about 6 metres from the boiler closet with three BDR91's (including hot water control) and everything always reports 5/5 excellent signal. However I occasionally see boiler "run on" like you describe, where there is no heat demand but the boiler relay (or sometimes the heating zone valve relay) cycles on and off for a further 20 minutes beyond what I would expect.

The cause of this is a lost heat demand message from controller to BDR91. Normally a heat demand message is sent from the controller to the relay either when there is a change in heat demand, or every 20 minutes if the heat demand stays the same.

So you turn the heating off and then a few minutes later when a zero heat demand is sent to the BDR91 it is for some reason not received, so the BDR91 continues on whatever duty cycle it was previously following, say 5 minutes on 5 minutes off. With no communication from the controller the relay will continue following the last know duty cycle for 45-60 minutes before lighting the red fault LED.

After 20 minutes another heat demand will be sent and if this gets through the relay will at that point go off. So if you see the boiler running on for an additional 20 minutes past the normal up to 5 minute delay then this is probably what is happening. If the second message also didn't get through then it would run on for 40 minutes. If the third message also didn't get through you'd see the red error LED light on the BDR91 and probably get a fault log entry on the controller.

If this is your problem then about the only thing you can do is either try to identify any sources of 868Mhz interference, (wireless alarms, Loop, other similar wireless devices) or relocate the BDR91 - even moving it a couple of feet or re-orientating it (on a different wall, if its in a boiler closet) might solve the issue.

Unfortunately the signal test function seems all but useless at diagnosing these type of problems. I had a communication problem with my CS92 hot water sensor where the signal test was fine but I kept getting hot water overshoot due to temperature readings not getting through reliably. In the end I moved it to another wall in the boiler closet (90 degree orientation change as well as moving a couple of feet) and even though the signal is now 4/5 instead of 5/5 it has largely fixed the problem. I still get an overshoot but only very rarely maybe only a couple of times a month instead of a few times a week.

If it's practical I would try moving your BDR91. Can you post a picture of where it is in relation to other objects ?

Another possibility is poor communication from an HR92 back to controller - which could cause the same sort of issue of lost heat demand messages but only when a specific zone is being scheduled off. I don't know what the retry schedule of an HR92 is and whether it is the same 20 minutes as controller to BDR91 but plan to test it soon.

1. I understand how it works and I also understand the delays when the devices communicate with each other, but it’s not that. I’ll explain with an example that happened yesterday.

I have 7 areas. At 00:00 all the areas should change to 5 degrees automatically (I programmed it like this, I don’t change it manually). At 23:54 a cycle starts and it ends at 23:55 (a15% of heat was requested in each cycle). At 00:00 all the areas should be assigned to 0. At 00:02 I see in the controller that all the areas turn blue and a few seconds later I hear how all the valves close. I go to check them myself and all the valves are at 5 degrees and completely closed. At 00:04 another boiler cycle starts and finishes at 00:05. Till then everything is normal. But what isn’t normal is that there is another cycle at 00:14-00:15 and another at 00:24-00:25.

2. I never touch the temperatures in the HR92. I do it through the mobile app or directly through the controller.

3. With this point, you have given me a big clue. It seems that, indeed, it is in the 20 minute cycles when it stops working. But the red error LED light on the BDR91 has never turned on (3 consecutive errors). I have the controller right next to the loudspeaker of the home cinema, the router, the TV, etc… I’ll try to move it, just in case I am losing some packages that the controller sends to the BDR91 because of interferences.

One more thing that could possibly cause this. If the boiler BDR is double bound (BDRs can have multiple sources of control) (or that the one that came with the Kit you used elsewhere in the home and then bound again) . You probably already did this - but I assume you did a full wipe of the BDR when you got it (as they come pre-bound and this is a common error as people rebind again)

I think that I did a full-wipe and a rebound, but I will try it again just in case, maybe it is that. With this, I remembered another question, who makes the heating request to the BDR91? The controller, the HR92 or both? If all the areas are programmed to 5 degrees and a HR92 manually requests heat, who is communicating with the BDR91? The controller is old and so it doesn’t show the changes applied to the HR92, but the HR92 requests heat to the controller and the controller to the BDR91, or the HR92 requests it directly to the controller?

Thank you for the ideas.


Two days later...

I tried to do a full-wipe to the BDR91. Also, I reduced the number of cycles per hour to 3, but nothing has changed. One day, at 23:55 the boiler stops. At 00:02 all the valves close and at 00:12 a cycle starts till 00:15. It didn’t start again. The next day, the same but with an extra cycle between 00:30 till 00:35, which shouldn’t have happened.

When I do the RF test of the BDR91 it gives an excellent signal and the led of the BDR91 flickers 5 times. All the valves have an excellent signal. I have a doubt, when I test the HR92 from the controller, the led of the BDR91 also flickers 5 times (as if it was testing itself). Is this normal?

I’m going to try to change the location of the controller (I have done the signal test maybe 100 times and it has an excellent signal).

[g6ejd]

Try what I do, that is set some zones to go off at different times. Without fail if I set all zones to go off together at the same time I get this cyclng effect you describe when no heat is required, but when by trial and error I discovered either using Optimum off or different zones set to different off times - I only set one differently from the rest, then the problem goes away. I'd call it a bug.