Page 3 of 3 FirstFirst 123
Results 21 to 26 of 26

Thread: No heating this morning

  1. #21
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    1,928

    Default

    Quote Originally Posted by Dean View Post
    Yes, I had the issue with Gen 2 and was hoping it would go away with Gen 3, but seems to be a Software Issue. Appears to have a mind of it's own and hard to replicate (without any event logging). It happened twice in the past couple of days and multiple times previously (even completely reset, unbound and cleared everything (controller, HR92's, BDR91) to see if it would fix it). Yesterday I had one zone on until 18:00 and heard the HR92 close down, and all other zones were off, but the Boiler relay kicked in at around 18:12 for a few minutes, then off and then again around 18:32 (I have my controller set to 3 cycles with 2 mins min, used to be 6 cycles which also had the same issue)
    As far as I know the controller doesn't directly switch the relay on and off it simply sends an updated duty cycle / demand figure to the relay which handles the switching on and off in a TPI pattern autonomously. So even if you pulled the batteries out of the controller an already cycling BDR91 would continue to cycle by itself for some time until it eventually realised that the controller hadn't been heard for a while and switched to failsafe (or off) mode.

    My hunch is a heat demand message from the controller to BDR91 goes missing (collisions, interference etc) thus when you turn the heating off the the message to say "zero heat demand now" isn't received by the BDR91, which continues on its last known TPI duty cycle. Most likely the current (zero) heat demand is re-sent and on a subsequent transmission the message gets through, then the BDR91 turns off for real. The lost messages also might be between an HR92 and the controller.

    Without more information it's hard to say how much of the wireless protocol is acknowledged and how much is unacknowledged transmissions - I get the impression that some system messages are unacknowledged in the design of the protocol and thus can go missing if there is interference or collisions. For example adjusting the set point directly on an HR92 updating the set point on the controller (local override display) is a message that apparently can go missing - make several adjustments in a row and you'll find one doesn't go through at random - at least I find that. And then the set point on the two devices are out of sync until you next adjust it.

    Another possibility with the BDR91 is if failsafe mode is enabled and it missed some messages from the controller it might drop into failsafe mode (20% duty cycle) until it next hears from the controller then go back off again. I'm pretty sure I've seen the odd behaviour with failsafe mode disabled though so it can't explain all examples of unexplained boiler firing.

    Good to know, just thinking it would affect the learning algorithm with heating on, with constantly dropping 0.5c, but guessing as the display is rounded it's just for display purposes and not the temperature it uses in heating algorithms. Personally I would prefer to see the temperature in 0.1c increments
    As the temperature is rounded to 0.5 degrees the fluctuation that results in the figure jumping back and forth 0.5 degrees is probably in reality only 0.1 degrees or so, but its right on the margin of rounding.

    You can actually get temperatures to two decimal places now using the old V1 Web API - in another thread I posted a small change to the domoticz plugin to take advantage of this, so my domoticz dashboard shows all zone temperatures to two decimal places and without the bogus rounding towards the set point that you see on the controller. After comparisons with another digital thermometer I've come to the conclusion that whilst both decimal places change, true resolution/granularity is somewhere on the order of 0.1 degrees, but that is still a lot better than 0.5 degrees and rounding towards the set point. I sometimes look at my domoticz dashboard from my phone to see what the "real" temperatures are, rather than the sanitised figures that the controller and iPhone app shows...

    I too would love to see 0.1 degree resolution without rounding towards the set point on the controller itself, but somehow I don't think this will happen! It's only a cosmetic issue though, the underlying control algorithms work with the full measurement resolution.
    Last edited by DBMandrake; 4th October 2016 at 04:02 PM.

  2. #22
    Automated Home Jr Member
    Join Date
    Feb 2016
    Posts
    11

    Default

    Quote Originally Posted by DBMandrake View Post
    As far as I know the controller doesn't directly switch the relay on and off it simply sends an updated duty cycle / demand figure to the relay which handles the switching on and off in a TPI pattern autonomously. So even if you pulled the batteries out of the controller an already cycling BDR91 would continue to cycle by itself for some time until it eventually realised that the controller hadn't been heard for a while and switched to failsafe (or off) mode.
    Many thanks, didn't realise that, this certainly clears things up.

    Quote Originally Posted by DBMandrake View Post
    Another possibility with the BDR91 is if failsafe mode is enabled and it missed some messages from the controller it might drop into failsafe mode (20% duty cycle) until it next hears from the controller then go back off again. I'm pretty sure I've seen the odd behaviour with failsafe mode disabled though so it can't explain all examples of unexplained boiler firing.
    There is no Red LED showing when this occurs, if BDR91 has gone into failsafe mode, would it show the Red LED?

  3. #23
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    1,928

    Default

    Quote Originally Posted by Dean View Post
    There is no Red LED showing when this occurs, if BDR91 has gone into failsafe mode, would it show the Red LED?
    I don't think so.

    A couple of times I have deliberately let the batteries run right down on the controller by leaving it off the dock, and I don't recall seeing the red LED - the only time I have seen the red LED is flashing when unbound or during binding. But keep in mind I only enable failsafe mode in winter.

    I might enable failsafe mode and take the batteries out of the controller for an hour to see what happens though - you've got me curious now!

  4. #24
    Automated Home Jr Member
    Join Date
    Jul 2016
    Posts
    10

    Default

    Hi all,
    I'm here to report that my Gen3 did exactly as OP's post. Seemed to turn on my zones this morning but never switched them off.

    I thought it was lagging 1 hour behind but when that time had passed I then spied the time, it was out (but updated itself when I went through the options)...

    I have rebooted it now using the "off the wall mount, batteries out" method and allowing it to settle down (although as above, some zones are taking their time to re-sync with the controller!)

  5. #25
    Automated Home Legend
    Join Date
    May 2014
    Posts
    998

    Default

    Quote Originally Posted by SplanK View Post

    I have rebooted it now using the "off the wall mount, batteries out" method and allowing it to settle down (although as above, some zones are taking their time to re-sync with the controller!)
    You can speed up the re-synch process with a push/release on the menu button on an HR92 in each of the zones.

  6. #26
    Automated Home Jr Member
    Join Date
    Jul 2016
    Posts
    10

    Default

    Ta, spotted that a few seconds ago..... to lazy to get out of my office chair now though

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •