Evohome cylinder strap on sensor incompatible ?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • paulockenden
    Automated Home Legend
    • Apr 2015
    • 1719

    #16
    And more...

    CdJHtgD6.jpg

    Hope that helps.

    I'll stop now. ;-)
    Last edited by paulockenden; 18 October 2016, 06:20 PM.

    Comment

    • DBMandrake
      Automated Home Legend
      • Sep 2014
      • 2361

      #17
      Originally posted by paulockenden View Post
      I don't know why the advice is like that, but it's what people say (there are slight variations) whenever someone moans about siphoning.

      Your pics are interesting - the boiler appears to have either three or four connections. I thought most system boilers simply had a cold in and a hot out?
      There are four separate taps on the heat exchanger currently in use, exactly like the Honeywell C Plan diagram above. Two taps go to the header and vent pipes off which the cylinder indirect loop is tapped part way up the run with T joints.

      The pump and radiator returns come off the other two taps. There is no direct connection between the pump and the vent or the header, except via the boiler heat exchanger as shown in the diagram.

      Comment

      • paulockenden
        Automated Home Legend
        • Apr 2015
        • 1719

        #18
        Well I guess if you can vent directly from the boiler it solves the usual problem of sucking in air. So ignore everything I've just been banging on about!

        Comment

        • DBMandrake
          Automated Home Legend
          • Sep 2014
          • 2361

          #19
          Originally posted by paulockenden View Post
          Well I guess if you can vent directly from the boiler it solves the usual problem of sucking in air. So ignore everything I've just been banging on about!
          Don't be sorry - if there's anything at all that I hadn't considered I'd want to know about it. I'm certainly no expert in this area.

          The last one you posted with the inverted cold water feed is certainly food for thought - the current system piping doesn't comply with this, but neither does Honeywell's example S Plan diagram! So I'm not sure what to think about that. The vent pipe is effectively connected on the output side of the boiler before the pump (but through a different top tapping to the vent pipe) however the cold feed is not after the vent pipe, nor is it on the boiler output - it's effectively on the return (bottom) side of the heat exchanger, as shown in the Honeywell diagram. Hmm.

          I wonder what the technical reasoning behind the recommendations is - I'm someone that likes to know why something is recommended and not just take it on face value - once you understand the mechanism behind it it becomes common sense and becomes easier to adapt the concept to different layouts and system designs.

          I can certainly see why having the vent and header on opposite sides of the pump would be very bad - you would create suction on the input side and pressure on the output side that would cause the head of water to be different for the header feed and the vent pipe, potentially enough to cause the vent pipe to overflow and create a continuous flow between vent and header, which would be bad. But as long as both pipes are on the input side of the pump and have minimal pressure difference between them I can't see how there would be enough differential pressure to cause it to overflow.

          So whether the cold feed is attached to the bottom of the heat exchanger or the top of the heat exchanger (but in both cases before the pump) I can't see that it would matter much ? Unless there was a restriction in the heat exchanger perhaps ?

          Comment

          • g6ejd
            Automated Home Guru
            • Oct 2016
            • 153

            #20
            Originally posted by DBMandrake View Post
            Hi All,

            I thought a good test would be to turn on the immerser and see what the reading went up to - well, it went up to 34 degrees and didn't go any higher, even though after an hour turned on the hot water coming out of the tap was easily 50+ degrees.

            Anyone have any ideas ? Is it possible that the cylinder has a double skin, and therefore the outside copper layer is not in direct contact with the hot water, preventing me from getting a meaningful reading ? When I remove the sensor and touch the copper it feels cool (less than body temperature) to the touch, so 34 degrees seems plausible given how cool it feels.

            Or could it be that the location of the immerser element near the top prevents much heat transfer to the casing of the cylinder low down where the sensor is installed, and that heating via the indirect loop which I believe is nearer the bottom would heat the bottom of the cylinder much more ? (Grasping at straws here)

            Another question - is it normal for the hot water sensor reading on the controller to only update once every 8-10 minutes ? This seems very infrequent if you had a quick re-heat cylinder - it seems that overshooting the set temperature would be a problem if it could continue to heat for another 10 minutes after the set temperature had been reached ?

            This is not a quick re-heat cylinder but it seems strange to me that HR92's would report their temperature every 4 minutes but the hot water sensor only every 8 minutes.
            I have a standard indirect cylinder and have just fitted the Hot Water Kit and have noticed over the last few days the Controller display is not updated that frequently, tonight it was indicating 37-degrees, thinking there was a problem I grabbed my IR thermometer and went to the cylinder and the screen updated to 49-degrees by the time I got there, IR Gun reported 50-degrees at the top. I have also noted that with the differential set to 10-degrees (as expected) it will not demand heat until that hysteresis value has been exceeded, but also the display does not update either until there is a demand for heat.

            I have been doing some testing tonight and heat demand calls take between 2-4 mins to be acted on as expected, but cylinder heat changes are not happening either at all or are very slow, yet to be determined. My hot water has a turn-around time of about 20-mins and I was expecting to see a gradual rise in water temperature as I waited during the successive 4-mins changed demand testing periods that the demanded water temperature would start to increase and I could see the effect, but nothing was happening on the Controller display, so I think there is a process going on with the DHW display that I don't yet understand, it does not show gradual temperature increases, it appears that there is no reporting from the sensor to the BDR91 to the controller during temperature rises, but only when the hysteresis criteria has been met, if that's the path, either that or the Sensor range is limited and bring the controller nearer corrects the issue.

            My current settings are:
            DHW temp set-point: 55'C
            Differential Temp: 10'C

            I think there is no update or heat demand until water temp reaches 45'C and then heating begins and there is no Controller update until the DHW is 55'C again.

            Turning the demand temp down is responded to almost instantaneously, which surprised me, no 4-min delay to take effect.

            My equipment is about 1-week old and of the latest known firmware 1.03
            Last edited by g6ejd; 18 October 2016, 10:59 PM.

            Comment

            • DBMandrake
              Automated Home Legend
              • Sep 2014
              • 2361

              #21
              In the brief testing I've done so far the temperature sensor updates only seem to be sent at most once approximately every 8 minutes - and that's if there is a large temperature change. (Achieved by removing or refitting the strap on sensor then timing and waiting for a change to be displayed)

              I guess it's possible that if there is only a small temperature change the updates are sent even less frequently. I too was expecting a more frequent and incremental update of the temperature. I guess I'll see what happens once the conversion is done - hopefully this weekend.

              Changes to the hot water demand are sent to the BDR91 immediately so if you manually override hot water on or off and the temperature is below the target the relay does seem to switch immediately. It's just that the reception of temperature sensor readings seems quite infrequent, which would of course lead to a delay in the relay switching on or off as the temperature changes as it is not aware of the temperature change immediately.
              Last edited by DBMandrake; 18 October 2016, 11:09 PM.

              Comment

              • paulockenden
                Automated Home Legend
                • Apr 2015
                • 1719

                #22
                Odd - I haven't had the stopwatch out, but it seems to me that if I run the hot tap the boiler fires a minute or two later.

                Comment

                • g6ejd
                  Automated Home Guru
                  • Oct 2016
                  • 153

                  #23
                  I just tested my DHW setup and if I increase demand temperature from 55'C to 65'C it takes ~4-mins to actuate the relay, turning if back down is near instant off though.

                  Also adjusting the temperature up from 55'C to 57'C and turning down the hysteresis value to 1'C also takes the same 4-min to actuate, but all these actions will surely depend on the controllers (assuming it uses) PID control settings and triggered by the D element? but in other situations the I element may also make instant changes depending how long it has been running.

                  Comment

                  • DBMandrake
                    Automated Home Legend
                    • Sep 2014
                    • 2361

                    #24
                    Originally posted by paulockenden View Post
                    Odd - I haven't had the stopwatch out, but it seems to me that if I run the hot tap the boiler fires a minute or two later.
                    Perhaps the temperature updates are sent and processed more frequently but the screen update is "lazy" ? I was only going by the time between screen updates not hot water relay switching.

                    I could rearrange my test to cause the temperature change on the sensor to go above and below the desired water temperature instead and measure the time until the relay changes. (With a small differential and zero overrun time set)

                    Are you using the strap on sensor and if so how far up the side of the cylinder is it and how big is the cylinder ?

                    It seems that the higher up the cylinder it is the more delay in responding to water use there would be due to stratification as cold water enters the bottom while it is not being heated.

                    A lower location would detect that entering cold water sooner and start heating well before the output temperature at the top had started to drop ?
                    Last edited by DBMandrake; 19 October 2016, 09:48 AM.

                    Comment

                    • DBMandrake
                      Automated Home Legend
                      • Sep 2014
                      • 2361

                      #25
                      Originally posted by g6ejd View Post
                      I just tested my DHW setup and if I increase demand temperature from 55'C to 65'C it takes ~4-mins to actuate the relay, turning if back down is near instant off though.
                      Do you have a non zero hot water overrun time set ? I found that if you do once the hot water overrun period is triggered it won't come back on again until that has completed even if a demand returns in the meantime.

                      Set your hot water overrun temporarily to zero during your testing and see if that makes a difference to response time.

                      Comment

                      • g6ejd
                        Automated Home Guru
                        • Oct 2016
                        • 153

                        #26
                        I'm used the strap-on sensor about 1/3 up from the base of a 1050x400 (newly fitted: http://www.screwfix.com/p/rm-indirec...-x-450mm/40082) tank. Took me a while to clear all the insulation back to nice clean copper.

                        OK, so setting the overrun period to 0 results in instant response to changes in heat demand, it was on 10-min. Now I'm wondering why Overrun would affect demand response times, maybe it has a dual purpose. If the dynamics of a system was that overruns were required because of latent heat capacity, then why would that affect response times? Presumably based on an assumption that the system is slow as overrun increases as part of the PID system controls.

                        Comment

                        • paulockenden
                          Automated Home Legend
                          • Apr 2015
                          • 1719

                          #27
                          Maybe there's no logic behind it. Perhaps Over-run just kicks off an internal timer, and nothing happens until that expires?

                          Comment

                          • DBMandrake
                            Automated Home Legend
                            • Sep 2014
                            • 2361

                            #28
                            Originally posted by g6ejd View Post
                            I'm used the strap-on sensor about 1/3 up from the base of a 1050x400 (newly fitted: http://www.screwfix.com/p/rm-indirec...-x-450mm/40082) tank. Took me a while to clear all the insulation back to nice clean copper.

                            OK, so setting the overrun period to 0 results in instant response to changes in heat demand, it was on 10-min. Now I'm wondering why Overrun would affect demand response times, maybe it has a dual purpose. If the dynamics of a system was that overruns were required because of latent heat capacity, then why would that affect response times? Presumably based on an assumption that the system is slow as overrun increases as part of the PID system controls.
                            10 minutes for hot water overrun sounds excessive to me.

                            As far as I can see there are two reasons to use hot water overrun:

                            1) After the boiler relay turns off when hot water demand is satisfied if you close the zone valve immediately the latent heat in the boiler heat exchanger will end up dissipated uselessly and very slowly by the automatic bypass loop, unless the heating zone valve is also currently open due to heating demand. Leaving the zone valve open for a few minutes lets that latent heat pass into the cylinder.

                            2) If you fire the boiler up to say 80 degrees to heat hot water quickly then close the zone valve at the same time you stop the burner you are suddenly taking away the heatsink of the cylinder (assuming there is currently no heating demand) and circulating only through the automatic bypass loop - this will almost certainly cause the heat exchanger/bypass loop to overshoot the boilers target flow temperature perhaps to the point of kettling and steam generation in the boiler, which particularly on an old boiler like mine could lead to steam generation and vapour lock. Leaving the valve open for long enough for the latent heat to pass to the cylinder and equalise the temperature between the heat exchanger and cylinder would prevent this happening.

                            This should only take about 2-3 minutes at most though - if you left the zone valve open and the pump running a full 10 minutes its very likely that the heat exchanger in the boiler would cool down enough to start sapping heat from your hot water cylinder rather than providing it. The optimal amount of overrun would be the length of time it takes the flow temperature to stabilise at cylinder temperature and no more. Anything further just removes heat from the cylinder, dissipating it in the pipe work and boiler itself. My guess is this will be around 2-3 minutes, but will depend on the mass of your heat exchanger etc.


                            Originally posted by paulockenden View Post
                            Odd - I haven't had the stopwatch out, but it seems to me that if I run the hot tap the boiler fires a minute or two later.
                            I did a bit of further testing and I've figured it out - there seem to be both infrequent scheduled temperature readings, and also unscheduled readings.

                            Every time the temperature crosses the hot water set point (in the increasing direction) and hot water set point minus differential (in the decreasing direction) the wireless sensor sends an unscheduled temperature reading immediately.

                            I tested this by setting the hot water to 40 degrees, one degree differential and zero minutes overrun, then pressed the sensor against a boiling kettle and then a cold package from the freezer alternatively to heat and cool it.

                            It turns the relay off about 5 seconds after pressing it to the kettle, and that includes the time it takes for the sensor probe to actually heat up.

                            Likewise pressing it against the cold freezer package causes the relay to switch back on in roughly 10 seconds, which is probably the time it takes to cool again. I could do this backwards and forwards multiple times and just like clockwork it would turn the relay on and off within seconds. The temperature reading on the display would update about 5-10 seconds after the relay activated so there is some delay between acting on the temperature reading and updating the display.

                            For it to be possible for the sensor to do this, it looks like the Evohome must inform the CS92 of the hot water set point and differential so that the CS92 knows when the set point has been crossed and that an immediate transmission must be sent.

                            What I also discovered is that after the temperature crosses the set point and it sends a temperature reading it sends two (or more ?) further temperature readings spaced approximately 10 seconds apart - I could notice this if the kettle was in the early process of boiling when I applied the sensor - it would give one reading that was high enough to turn the relay off, but then about 10 seconds later I'd get a higher reading (the kettle was still getting hotter) then a 3rd even higher reading another 10 seconds or so later, but after that the reading would stop updating even though the kettle was still heating up.

                            Almost certainly the reason for sending two additional readings spaced only 10 seconds apart after the set point is crossed is to deal with the situation of a lost transmission - since there are no acknowledgements if only one was sent and there was a collision, the relay would not switch off and the cylinder would continue to heat for another 8 minutes until the next periodic transmission, causing an overshoot in the temperature.

                            Two additional transmissions makes the chance of all three transmissions being lost extremely remote, so at worst the relay turning off will be delayed about 20 seconds.

                            When the temperature is not crossing the set point it sends a periodic update which seems to be approximately every 8 minutes. Which explains why the reading updates so infrequently when the set point is not being crossed - which will be done for battery conservation reasons.

                            So as far as controlling the temperature goes the system seems to work well and should be very responsive and accurate, however the displayed temperature on the screen may lag a long way behind actual temperature if it has not recently crossed the set point.
                            Last edited by DBMandrake; 19 October 2016, 07:08 PM.

                            Comment

                            • paulockenden
                              Automated Home Legend
                              • Apr 2015
                              • 1719

                              #29
                              Great research, and glad it mirrors my experience so I'm not going mad!

                              Comment

                              • g6ejd
                                Automated Home Guru
                                • Oct 2016
                                • 153

                                #30
                                A good set of experiments and now we all know a lot more about the control algorithms. It would be good if Honeywell published the method of operation and in doing so would not give a lot away to competitors but it would help customers understand how the system works. Maybe we are just meant to install it and not be inquisitive as to its operation.

                                Last night I too concluded the CS92 must receive the system parameters for DHW to know when to send control signals; makes sense, but there is a little bit of uncertainty, because it is just possible the Controller is controlling things...

                                My system has been running for about 4-days now and I have to say the increase in comfort and control over a Nest is amazing, and in a little bit of wonderment, I keep wandering into rooms to check if radiators are off as programmed and of course they are, so now I've started recording gas consumption to see if there are any changes, normalised to outside temperature. I have 10-years of weekly gas consumption and can get temperature from almanacs, so I should be able to see if there is any obvious energy saving, not that I spent £1400 to save energy, for me it's about solving rooms that were always colder than I preferred, problems that are solved through this Evohome system.

                                I am getting communication errors on screen and in the log from the two BDR91 relays and but that aside everything is working as expected. I have other equipment running on 868MHz so I'm going to switch them off and see if that's the cause of the communication errors.

                                I'm pleased I discovered this forum, as the source of knowledge is invaluable.

                                Comment

                                Working...
                                X