New evohome FW - update WiFi 02.00.17.00 and Application 02.00.17.03

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • DBMandrake
    Automated Home Legend
    • Sep 2014
    • 2361

    Now at 1:50pm the controller reading has changed from 5 to 6 degrees, while the iOS app is still reporting 8 degrees.

    At least for my location, it doesn't even appear that the controller and iOS app are using the same weather data.... as if it was just delayed it should have caught up to 8 degrees now instead of changing from 5 to 6.

    Comment

    • Dan_Robinson
      Automated Home Ninja
      • Jun 2012
      • 347

      I wonder if it can pick up the weather data over Opentherm because mine is saying 11 degrees when the boiler is reporting 11.4 degrees? Assorted other services are saying 10 degrees which is more accurate than normal.

      Also though. I have zero % demand on the zones, but the heat demand is asking for 27%. The OT monitor is saying there is a control setpoint of 39 being asked of the heating. Although SWMBO is having a shower at the moment so the boiler is in hot water mode.

      Bizarre.
      Kind Regards - Dan Robinson (Jennings Heating Ltd)

      Comment

      • Dan_Robinson
        Automated Home Ninja
        • Jun 2012
        • 347

        Now decided that I think I'll keep mine on the summary screen. Can we have night mode on all the other screens now?


        Getting this idea in first before RB nicks the idea .
        Kind Regards - Dan Robinson (Jennings Heating Ltd)

        Comment

        • mtmcgavock
          Automated Home Legend
          • Mar 2017
          • 507

          Is anyone else experiencing overshoots? I know the weather is milder than usual, however i'm showing 2.5oc overshoots in most zones. To the point where i'm feeling rather warm

          Can someone confirm whether the learnings have been reset during the update?

          Comment

          • DBMandrake
            Automated Home Legend
            • Sep 2014
            • 2361

            Not seeing any overshoots here that I wouldn't associate with my flow temperature still being set high for sub zero conditions while actual weather is suddenly ~10 degrees in the space of only a day or so.

            Something to keep in mind is that the PID controller with self learning heuristics which adapts to the thermal response of your room to avoid overshoots is done entirely in the individual HR92's - the controller just commands the set point changes in accordance with your schedule, (and sends the set points early as part of optimal start if enabled) but it's the HR92 itself that decides what to do with that set point and measured temperature data and how much to open the valve and how much heat to demand from the boiler.

            So as far as I can see a controller firmware update can't/wouldn't reset the heuristics in the HR92.

            On the other hand in a zone that doesn't use an HR92 like a zone valve zone or the new electric only zones, the PID controller with self learning heuristics would be implemented in the controller itself, so those might have been reset by a firmware update. (I only have HR92 zones so I can't comment on those)

            Comment

            • mtmcgavock
              Automated Home Legend
              • Mar 2017
              • 507

              Originally posted by DBMandrake View Post
              Not seeing any overshoots here that I wouldn't associate with my flow temperature still being set high for sub zero conditions while actual weather is suddenly ~10 degrees in the space of only a day or so.

              Something to keep in mind is that the PID controller with self learning heuristics which adapts to the thermal response of your room to avoid overshoots is done entirely in the individual HR92's - the controller just commands the set point changes in accordance with your schedule, (and sends the set points early as part of optimal start if enabled) but it's the HR92 itself that decides what to do with that set point and measured temperature data and how much to open the valve and how much heat to demand from the boiler.

              So as far as I can see a controller firmware update can't/wouldn't reset the heuristics in the HR92.

              On the other hand in a zone that doesn't use an HR92 like a zone valve zone or the new electric only zones, the PID controller with self learning heuristics would be implemented in the controller itself, so those might have been reset by a firmware update. (I only have HR92 zones so I can't comment on those)
              That's interesting to know. But surely if you've got the optimisation feature turned on (Mines set at one hour) the controller must tell the HR92 what time to come on (As shown on each zone). So I would assume some learning must be done also by the controller, correct me if i'm wrong.

              I have an additional towel rail zone controlled by the T87RF, however it's difficult to tell if this has been affected as it also comes on with the hot water in winter so overshoots in this zone occur all the time.

              Comment

              • paulockenden
                Automated Home Legend
                • Apr 2015
                • 1719

                Originally posted by DBMandrake View Post
                On the other hand in a zone that doesn't use an HR92 like a zone valve zone or the new electric only zones, the PID controller with self learning heuristics would be implemented in the controller itself...
                Not sure about that. The wall stats all do their own TPI (when paired to a BDR91 to directly control a boiler), but - I think - without learning. Electric zones would, I'd imagine, work the same way.

                P.

                Comment

                • DBMandrake
                  Automated Home Legend
                  • Sep 2014
                  • 2361

                  Originally posted by mtmcgavock View Post
                  That's interesting to know. But surely if you've got the optimisation feature turned on (Mines set at one hour) the controller must tell the HR92 what time to come on (As shown on each zone). So I would assume some learning must be done also by the controller, correct me if i'm wrong.
                  Optimal start is separate and is handled by the controller. It monitors the warm up rate of a zone and learns how quickly the room warms up - and uses that information together with starting and target temperature to work out how early the radiator needs to come on. So it's only contribution is to send the set point change to the HR92 earlier than the scheduled time. This has nothing to do with whether a zone overshoots the target or not - that depends on (among other things) the tuning of the PID controller in the HR92 - which is self tuning using its own heuristics.
                  I have an additional towel rail zone controlled by the T87RF, however it's difficult to tell if this has been affected as it also comes on with the hot water in winter so overshoots in this zone occur all the time.
                  I have 3 zones with DTS92's - can't say I've noticed any difference in overshoot on them. But even if you have a wall thermostat the PID controller is still in the HR92 if it is a radiator controller zone. All the wall stat does is sends periodic temperature readings to the controller which then forwards them on to all HR92's in the zone.

                  I was referring zones which don't feature an HR92 at all as being different. In a zone valve zone where there is no HR92 the logic must be in the controller.

                  Comment

                  • DBMandrake
                    Automated Home Legend
                    • Sep 2014
                    • 2361

                    Originally posted by paulockenden View Post
                    Not sure about that. The wall stats all do their own TPI (when paired to a BDR91 to directly control a boiler), but - I think - without learning. Electric zones would, I'd imagine, work the same way.

                    P.
                    I think we've had this conversation before. The wall stats like the DTS92 have more than one mode of operation. When you pair it directly with a BDR91 it acts as a full controller in its own right and you can access the TPI settings in its configuration menu, and it will even show a flame icon when it is calling for heat from the boiler.

                    When you pair it to an Evohome it acts as a "dumb" temperature sensor and set point controller to send set point changes when you press the buttons. It will not let you adjust TPI settings or show a flame icon, with most of the menu options disabled.

                    Now it's only contribution to controlling the zone aside from set point changes is to send a periodic temperature measurement to the controller which then forwards it to the HR92's, which in turn make decisions about how to adjust the valve (PID controller) and the heat demand to send back to the controller.

                    Control wise a zone that is DTS92+HR92 is no different than one that is HR92 only, the only difference is where the temperature measurement is taken.

                    You won't find any heat demand messages being sent from the DTS92 to the Evohome controller - they are still coming from the HR92's.
                    Last edited by DBMandrake; 22 December 2017, 07:30 PM.

                    Comment

                    • mtmcgavock
                      Automated Home Legend
                      • Mar 2017
                      • 507

                      Originally posted by DBMandrake View Post
                      I have 3 zones with DTS92's - can't say I've noticed any difference in overshoot on them. But even if you have a wall thermostat the PID controller is still in the HR92 if it is a radiator controller zone. All the wall stat does is sends periodic temperature readings to the controller which then forwards them on to all HR92's in the zone.

                      I was referring zones which don't feature an HR92 at all as being different. In a zone valve zone where there is no HR92 the logic must be in the controller.
                      I should of stated, this zone is controlled by a zone valve (Separate towel rail circuit). No HR92s involved at all.

                      Maybe it's just a coincidence with the warmer weather then, will continue to monitor over the next few days. Was just interested to see if anyone else had noticed the same since the update.

                      Comment

                      • DBMandrake
                        Automated Home Legend
                        • Sep 2014
                        • 2361

                        Originally posted by mtmcgavock View Post
                        I should of stated, this zone is controlled by a zone valve (Separate towel rail circuit). No HR92s involved at all.
                        Ah ok. In that case yes, it is possible the firmware update reset some learnt heuristics for the zone as they would be implemented in the controller for that type of zone.

                        Comment

                        • HenGus
                          Automated Home Legend
                          • May 2014
                          • 1001

                          I am having a feeling of deja vu. This was how it was 4 years ago as early Evohome adopters tried to work out what was normal and abnormal operation. Some Release Notes on the upgrade would be helpful.
                          Last edited by HenGus; 23 December 2017, 12:25 PM.

                          Comment

                          • StephenC
                            Automated Home Guru
                            • Feb 2017
                            • 102

                            DBMandrake is right - the explanation of the HR92 adaptation over time (when to shut down) vs Evotouch controller optimisation (on/off timing) is accurate.

                            This is the same explanation of the workings I was provided by Honeywell when troubleshooting large overshoots earlier in the year due to OpenTherm compatibility issues (very high and permanent flow temps).

                            As an example, that the HR92’s don’t sample the temp at a very high rate (every 4 mins I believe to conserve battery), coupled with a very high flow temp make it very difficult to avoid overshoots in all but very undersized radiators and or very high heat loss room characteristics. It’s all about balance.
                            Last edited by StephenC; 22 December 2017, 07:52 PM.

                            Comment

                            • Dan_Robinson
                              Automated Home Ninja
                              • Jun 2012
                              • 347

                              Another interesting thing to note.

                              My mixing valve zone is running (pump etc) but the demand screen is still saying 0%.
                              Kind Regards - Dan Robinson (Jennings Heating Ltd)

                              Comment

                              • paulockenden
                                Automated Home Legend
                                • Apr 2015
                                • 1719

                                Again, that could be a missed comms message. That's the trouble with RF that doesn't ACK messages.

                                Comment

                                Working...
                                X