Results 1 to 6 of 6

Thread: Long delays from Evohome until BRD91 activates

  1. #1
    Automated Home Jr Member
    Join Date
    Sep 2018
    Posts
    29

    Default Long delays from Evohome until BRD91 activates

    Hello all,

    Anyone had an experience with long delays from Evohome sending the 'on' signal and the BRD91 actually turning on.?

    Mine started this morning with the Heating not coming on. Investigation showed the rad valves open but the BRD91 not on, having seemingly ignored its on instruction for the morning heating run, oddly the hot water had fired . So testing followed and there is a 3-4 mine delay in the on time until the BRD91 actually turn on and the heating fires.

    Any one have experience with this ?

    Thanks,

    FLY

  2. #2
    Automated Home Guru
    Join Date
    Oct 2020
    Posts
    160

    Default

    Quote Originally Posted by Fly100 View Post
    Hello all,

    Anyone had an experience with long delays from Evohome sending the 'on' signal and the BRD91 actually turning on.?

    Mine started this morning with the Heating not coming on. Investigation showed the rad valves open but the BRD91 not on, having seemingly ignored its on instruction for the morning heating run, oddly the hot water had fired . So testing followed and there is a 3-4 mine delay in the on time until the BRD91 actually turn on and the heating fires.

    Any one have experience with this ?

    Thanks,

    FLY
    Interesting. Since installing 19.33 and turning on the scaling feature, I've noticed delays in morning startup. Cold house, you would expect boiler to be at max immediately, but that is not the case. In the graph the purple represents the percentage demand, and the green the boiler relay.
    Capture.JPG

  3. #3
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    2,361

    Default

    Quote Originally Posted by Fly100 View Post
    Hello all,

    Anyone had an experience with long delays from Evohome sending the 'on' signal and the BRD91 actually turning on.?

    Mine started this morning with the Heating not coming on. Investigation showed the rad valves open but the BRD91 not on, having seemingly ignored its on instruction for the morning heating run, oddly the hot water had fired . So testing followed and there is a 3-4 mine delay in the on time until the BRD91 actually turn on and the heating fires.

    Any one have experience with this ?
    What firmware version are you on ? And if you are on 19.31 or 19.33 do you have load scaling enabled in full or partial mode ?

    Knowing the answer to those questions will decide whether what you're seeing is normal or a fault condition...

    However if you're talking about a 3-4 minute delay between when the schedule starts switching zones on until the boiler fires, that is normal as set point changes are only sent out to HR92's every 3 1/2 minutes, and an HR92 won't call for heat until it has received the set point change and turned its motor. Hot water turning on has no delay by comparison.

    Load scaling can cause an additional "delay" because prior to load scaling a single zone more than 1.5C below the set point would cause a 100% heat demand to the boiler causing the relay to switch on immediately. (After the up 3 1/2 delay sending the set points to the HR92's that is)

    This is because 100% demand from a zone can easily produce a much lower heat demand sent to the boiler (say 40%) and it may take a few minutes into the TPI cycle before the relay is due to turn on with a 40% heat demand. (Up to 6 minutes for 40% and 6 cycles per hour)
    Last edited by DBMandrake; 1st December 2020 at 10:00 AM.

  4. #4
    Automated Home Jr Member
    Join Date
    Sep 2018
    Posts
    29

    Default

    Im on the 19.33 Firmware, Im unsure about the scaling as its running an Oil boiler, so im guessing I dont have that option as oil boilers dont modulate.I have to say it great system but when it plays up its a pain in the ass!!

  5. #5
    Automated Home Ninja
    Join Date
    Dec 2016
    Posts
    273

    Default

    Quote Originally Posted by DBMandrake View Post
    ..., that is normal as set point changes are only sent out to HR92's every 3 1/2 minutes,...
    Isn't that the other way around? As was explained to me that as part of an energy saving scheme it is the HR92 that wakes up every 3.5 minutes to request input from the controller. The energy saving thus follows from the HR92 not having to consider radio transmissions meant for other devices.

    And there are other energy preserving measures as well, one of which the controller actually allows to be altered within a rather limited range: the maximum number of boiler valve switches per hour. Which is interesting because according to my info the HR92 does not send an `open valve` request to the controller but directly to the BDR91, so again this indicates some type of time slotting that can be loaded into the BDR91 (which can be always on listen mode because it's not battery powered) and is in fact rather pointless in my case because I actually have no boiler at all.

    So if you upgraded to newer firmware, you might want to have a look at these parameters and see if the upgrade caused any changes on them.

  6. #6
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    2,361

    Default

    Quote Originally Posted by gordonb3 View Post
    Isn't that the other way around? As was explained to me that as part of an energy saving scheme it is the HR92 that wakes up every 3.5 minutes to request input from the controller. The energy saving thus follows from the HR92 not having to consider radio transmissions meant for other devices.
    Someone like zxdavb or dand are more expert on the intricacies of the protocol but this is my understanding:

    In each cycle the controller sends a wake up interval in seconds to the HR92's, (approx 210 seconds) and they use this to synchronise their wakeups. The HR92's go to sleep for this specified sleep time and unless they wake up for some other reason (user turning the knob for example) they sleep that amount of time then wake up and listen for the controller for a short time.

    The controller then just sends out the set points for the zones and also the measured temperature for each zone during this wake up interval and I think it also sends another wake up interval message as well so the HR92's know how long to wait for their next wake up period.

    It wouldn't make sense for every HR92 to send a request for information just for itself when they wake up as the chance of collisions among HR92's would be very high and it would be a lot of redundant traffic to ask individually for information that the controller already knows it needs to send. It makes far more sense for them to just wake up and listen at the allotted time window so that the controller can rapidly send out the information for all zones to all HR92's with no "back chat" from the HR92's to cause collisions.
    And there are other energy preserving measures as well, one of which the controller actually allows to be altered within a rather limited range: the maximum number of boiler valve switches per hour.
    Not really. You might be thinking of the TPI cycle which you can set to 3, 6 or 12 cycles per hour which is how many times the relay would switch for a constant partial heat demand. The controller doesn't send a message every time the relay switches, it just sends heat demands which translate as a duty cycle and the relay takes care of the switching on and off for itself.

    The number of heat demand transmissions from controller to BDR91 adjusting the duty cycle doesn't seem to be particularly limited - if you keep turning an HR92 up and down every few seconds the controller will keep sending new heat demands to the relay. Remember that a change in heat demand part way through a TPI cycle often won't result in an immediate relay switch - more often than not it will just cause the next switch time to come sooner or later. Only if you're within a period of time when the relay is about to switch anyway would a heat demand change cause it to switch then and there.
    Which is interesting because according to my info the HR92 does not send an `open valve` request to the controller but directly to the BDR91, so again this indicates some type of time slotting that can be loaded into the BDR91 (which can be always on listen mode because it's not battery powered) and is in fact rather pointless in my case because I actually have no boiler at all.
    I'm afraid you've drifted way off the reservation here... HR92's aren't bound to BDR91's and don't send their heat demands directly to BDR91's. They are only bound to the controller and only send their heat demand to the controller.

    The "heat demand" sent by an HR92 is just the valve pin position, and is identical to the value reported by option 10 in the HR92 menu. This value once received by the controller goes through a non-linear function (which I've posted a graph of a few times) before being displayed as a heat demand on the controller system info page. From there the values from the different zones are combined and sent to the boiler relay.

    Also of note - the method of combining the heat demands of different zones to send to the BDR91 is totally different with load scaling enabled or disabled.
    So if you upgraded to newer firmware, you might want to have a look at these parameters and see if the upgrade caused any changes on them.
    One thing that could potentially change (but hasn't as far as I know) is the wake up interval sent from the controller to HR92's - the protocol seems to be designed so that the controller can specify custom arbitrary wake up periods that differ depending on circumstances, but as far as we know the same time interval is always sent.

    One possible use for this might be to reduce the wake up interval when the system is in a state where zones are heating rapidly - by receiving more frequent updates chances of overshoots are reduced compared to HR92's only responding every 3 1/2 minutes. Then when the system is stable (rooms at equilibrium) the controller could increase the sleep time again to preserve battery life.

    I'm not sure why they haven't done something like this yet, because the long delay between updates to HR92's is no doubt partly responsible for the overshoots that can happen, as nearly 4 minutes is a long time between valve adjustments compared to a wax pellet TRV which is using continuous real time analogue adjustment in response to temperature changes.
    Last edited by DBMandrake; 5th December 2020 at 07:52 PM.

Posting Permissions

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