Central Heating randomly turning on when it should be off

[Jim]

Is it possible there are guests or children in the house who may be fiddling with their HR92's without your knowledge ? (It is possible to lock the controls on specific HR92's so they can only be controlled via the controller or phone apps if you have any suspected fiddle fingers in the house )

Also you say it was turning on in 5 minute cycles - the default cycle time is actually 10 minutes (6 firings per hour, set in the installer menu) which means it should only fire once per 10 minute period. If you can't find any other cause can you time how long it fires in each 10 minute cycle ?

Also in response to a comment in your original post where you said pushing the button on the BRD91 "reset" it, that is actually only a temporary override, not a reset. It will turn the relay off if it is on, or on if it is off, but at the next scheduled TPI switch time it will resume where it left off. So if it is modulating the boiler on and off in TPI mode it will come back on again a few minutes later.

Hi again and thanks for your suggestions.
I have double checked the HR92s and can confirm they haven't been fiddled with - pity because would have been an easy resolution.
It's definitely a 5 minute cycle when it happens - on 5 then off 5.
And good point about the "reset" - I hadn't thought about that.
The mystery continues.

[Jim]

When you rebound the BDR91 what process did you take?

I suggest the following which is how I rebind them:

Hi and thanks for your detailed instructions on rebinding. I omitted setting the boiler control to none so will now retry exactly as you have outlined. Hopefully that will solve the problem.

[Jim]

Just a quick update - after a weeks testing following rebinding the system is still working the way it should. Fingers crossed it stays that way now.

Thanks to The Evohome Shop for your solution.

[Captainham]

I've had the exact same problem as Jim and have run out of ideas, have tried everything mentioned in this thread.

My BDR91 is configured via Sundial Valve menu (I have 2 zone valves, 1 for HW and 1 for CH), so Boiler Relay is set to None.

Have rebound the Heating Valve following detailed instructions above yet still it keeps on cycling.

When I do a comms check, I get 5 red LED flashes on BDR91 (not green)...is this normal? Same on the HW BDR but this is working fine.

I do get a success message on the evohome (for both valves, both with excellent signal).

[paulockenden]

Check your zone vales. Sometimes the microwswitches go faulty, and so they fire the boiler even when closed. And sometimes they do the opposite!

The good news is if that is the fault you can just replace the head - no plumbing required!

P.

[Captainham]

Check your zone vales. Sometimes the microwswitches go faulty, and so they fire the boiler even when closed. And sometimes they do the opposite!
P.

Ok, dumb question here then...are you saying that if the zone valve (incorrectly) calls for heat, which fires the boiler, then it will also cause the LED on the BDR91 to come on?

I would've only expected that the BDR would light up if it was itself calling for heat, whereas anything 'upstream'/non-evohome-related (such as the zone valve) would not 'talk back down the line' and light the BDR up...if what I'm saying makes sense!

[paulockenden]

No, sorry, it won't light the LED on the BDR91 - I missed that bit!!!!

P.

[Captainham]

No, sorry, it won't light the LED on the BDR91 - I missed that bit!!!!

No probs - is there any mileage in me replacing the BDR91? Or I guess I could do a factory reset on the evohome, but then I'll have to set-up/re-sync everything, which I was hoping to avoid but perhaps I just need to bite the bullet.

[DBMandrake]

I've had the exact same problem as Jim and have run out of ideas, have tried everything mentioned in this thread.

Without me going back and reading the entire thread, have you tried removing the batteries from each HR92 and re-fitting them ? And I assume the controller has been rebooted ?

Have rebound the Heating Valve following detailed instructions above yet still it keeps on cycling.

Can you time how long it cycles on for in each 10 minute period when all HR92's are turned down ? Failsafe mode is 20% duty cycle so if its something other than 2 minutes out of 10 it won't be failsafe related.

If you unclick the latch of every HR92 in the house (which you would normally do to lift it off the radiator) then wait about 30 minutes does it continue to keep cycling the boiler on ? In theory an HR92 should not call for heat when the latch is unclicked.

If that doesn't work, try removing the batteries from every HR92 and wait 30 minutes - after 30 minutes of no transmissions from an HR92 the heat demand that HR92 was responsible for should cease. If that DOES work, try adding the batteries back into a couple of HR92's at a time and test again to see if the issue has returned.

My money is on a problem with one specific HR92 sending out a bogus heat demand.

Another thing that might help is start with all HR92's set to 5 degrees and confirm it is still cycling on intermittently when it shouldn't, then one at a time try turning each HR92 right up to 30 degrees (and turn down the previous one) to make sure that the heating relay comes on and stays on - if you find one where it doesn't and it just continues to cycle intermittently you may have found your culprit.

It's possible you may be seeing an issue related to the one dty had in this thread, where one HR92 is sending a low, bogus heat demand.

http://www.wordpress-1219309-4387497.cloudwaysapps.com/vbulletin/showthread.php?5073-HR92-not-responding-correctly

When I do a comms check, I get 5 red LED flashes on BDR91 (not green)...is this normal? Same on the HW BDR but this is working fine.

I do get a success message on the evohome (for both valves, both with excellent signal).

Sounds normal - but unfortunately I've not found the signal test to be a good indication of whether there are difficulties with comms or not. It's not uncommon to get a 5 out of 5 signal strength reading but still have problems.

No probs - is there any mileage in me replacing the BDR91? Or I guess I could do a factory reset on the evohome, but then I'll have to set-up/re-sync everything, which I was hoping to avoid but perhaps I just need to bite the bullet.

My guess is that your comms problem is from an HR92 to the controller and probably not to the BDR91.

If a BDR91 doesn't hear from the controller in at least 45 minutes or so the red error LED will come on. If this LED does NOT come on at all AND the bogus relay cycling continues for at least an hour with all HR92's turned down I think that points the finger of blame at an HR92 or the communication from HR92 to controller.

If none of the testing above pin points the issue then the only way that I'm aware of to dig deeper into the problem is set up Domoticz with an HGI80 to monitor the wireless comms and heat demands directly.

[Captainham]

Awesome and great info

Huge thanks for that post, I'll start working through it and see if it pinpoints the issue. Not heard of those types of issues with the HR92 before - being a semi-dunce, I've tended to point the finger of blame at the BDR91 (it does occasionally flip out like this but usually sorts itself out quickly without intervention).

I love my evohome and I tell others the same, but it's times like 01:30 this morning when it was driving me potty (I could hear the boiler firing up plus our bedroom has a regular TRV, not HR92) that I do have serious doubts about retaining it. Problems like this would definitely keep it to a hardcore/niche audience in my view.

[Captainham]

I hereby retract any previous slights aimed at the evohome system, and am also escalating myself to fully-fledged Dunce status.

I neglected the most basic of checks, which was walking to each rad to see if they were hot, even though the controller said no heat was required in any room. Turns out one room (which my nephew plays in...) had been turned up to 21 manually, however the app and evohome were not reporting this (they do usually), they were only reporting the regular schedule (which, without override, would have meant no heat was being called for).

So technically still a bug in the system (as it should have highlighted there was an override in place), but not one that should have irritated/disturbed me so much last night when I should've known better.

Thanks for the assistance guys, I shall dunk my head down the toilet as penance for my ignorance!

[DBMandrake]

I neglected the most basic of checks, which was walking to each rad to see if they were hot, even though the controller said no heat was required in any room. Turns out one room (which my nephew plays in...) had been turned up to 21 manually, however the app and evohome were not reporting this (they do usually), they were only reporting the regular schedule (which, without override, would have meant no heat was being called for).

So technically still a bug in the system (as it should have highlighted there was an override in place), but not one that should have irritated/disturbed me so much last night when I should've known better.

Aha! Glad the problem is solved.

Local overrides made on an HR92 not being reported back to the controller 100% of the time seems to be a "known" bug in the system. Honeywell haven't acknowledged it to my knowledge but I see it on a semi-regular basis and I've seen it discussed on the forum as well. Happens for me approximately one in ten times that I make a manual override from the HR92 (which I don't do often) and tends to happen more often if I make multiple adjustments in a row - for example if I overshot my desired setting, paused a moment and turned it back.

It can cause some confusion when there is a discrepancy between what the controller believes the set point is and what the HR92 believes it is. Normally they would get back into sync at the next scheduled set point change however if you don't have any set point changes scheduled (say its a spare room that is scheduled to 5 at multiple times of day) then it won't send out a message to the HR92 to change back to 5 if it already believes it is at 5.

[Captainham]

Normally they would get back into sync at the next scheduled set point change however if you don't have any set point changes scheduled (say its a spare room that is scheduled to 5 at multiple times of day) then it won't send out a message to the HR92 to change back to 5 if it already believes it is at 5.

Therein lies the other oddity/bug (I was "double-bugged" on this occasion which is what caught me out, as never had a double before!). When I first set up the system, I intentionally did exactly as you said and put in checkpoints for every day, to avoid just such an incident. Ok, the temp in both setpoints was the same, but there was a setpoint which (I think, although now doubting myself) has worked in the past.

But in this case, the set points did not remove the override. I've now changed one of the temps by half a degree in case that was the issue, but I do have a dim recollection of same temps across set points not being an issue in the past.

Hey-ho, sorted anyway so I can glue back in the hair I pulled out when bleary-eyed early this morning!

[DBMandrake]

Therein lies the other oddity/bug (I was "double-bugged" on this occasion which is what caught me out, as never had a double before!). When I first set up the system, I intentionally did exactly as you said and put in checkpoints for every day, to avoid just such an incident. Ok, the temp in both setpoints was the same, but there was a setpoint which (I think, although now doubting myself) has worked in the past.

But in this case, the set points did not remove the override. I've now changed one of the temps by half a degree in case that was the issue, but I do have a dim recollection of same temps across set points not being an issue in the past.

Yes, the issue is that if you have a string of set points in a row that are all the same, say 5.0 degrees, if the controller already believes that the HR92 is set to 5.0 degrees it will not send a "redundant" set point update to the HR92, thus it would remain on 21...

I would argue that not always sending the set point update even if it seems "redundant" is a bug. Normally when you adjust the HR92 the controller will be informed of the change and know that it is really set to 21 so when the 5.0 degree set point comes around it will send a message to set it back to 5. However if the message from the HR92 at the time the dial was turned telling the controller it is now set to 21 is not received and the controller continues to indicate 5.0 degrees, the next 5.0 degree set point change won't be sent and the HR92 will remain at 21... their set points would stay out of sync until there was some other set point change other than 5.0 scheduled or manually applied.

The workaround to avoid this happening again is to alternate your set points slightly so you don't have multiple set points of the exact same value in a row - for example alternate between 5.0 and 5.5. That way the controller always believes it has a set point change that needs sending and doesn't skip any. Hopefully this bug will be fixed one day...

[Captainham]

So I stand corrected on my recollection - you live and learn! Yeah I'd agree, a set point is a set point, so if there's one there, it should be sent out, regardless of any similarity to the prior set point.

Looking forward to a less sweaty night's sleep tonight (I removed the HR92 in the bedroom as I'm a light sleeper, and the whirring of the valve opening up early morning is enough to disturb me, so I've had to go back to an old-school TRV in that room).