Problem with Optimum Start

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • GAndrea
    Automated Home Lurker
    • Jan 2016
    • 3

    Problem with Optimum Start

    This is my first post and, first of all, I would like to thank you for writing and maintaining this forum. I learned a lot on the use of the EvoHome system here.
    Unfortunately, I found the official documentation really poor for such a kind of expensive system.

    I have installed the evohome wifi (a controller, 10 HR92 valves, a gateway and a BDR91 relay box) since October but I noticed a strange behavior.
    It seems that the Optimum Start feature is not working when the difference between the current temperature and the next one is less than two degrees.

    For example, I have 17 degrees for the night and 18 from 7:00. In this case the Optimum Start not works (no icon on the controller before 7:00) and the Evohome is like a normal thermostat or even worst. Because it turns on the boiler at 7:00 but not at the maximum power, it continues to turn on and off every like 10 minutes more o less and the valves are open at 50% more o less (lukewarm radiator). I would expect always on with valves opened at 100 until the new temperature is reached.
    This is a problem because after one hour, let say at 8:00, the temperature in that room is still 17,5 degrees.
    To solve it I have to leave the temperature constantly at 18, but is not so smart...
    This behavior is the same for all the rooms/areas.
    As soon as the difference is higher that 2 degrees the optimum start starts (icon appears) and works fine.

    I have already tried to reset the full system.

    Any of you have noticed this or can make a test?
    Thanks
  • paulockenden
    Automated Home Legend
    • Apr 2015
    • 1719

    #2
    can you really tell the difference between 17 and 18 degrees? I'm not sure I could.

    The two temperatures are so close that Evohome will be operating in the TPI band, which is why you are seeing the boiler cycling. It's very hard to accurately get from 17 to 18, and if everything came on full-blast you'd almost certainly overshoot. As you'll have seen from other posts here, people don't like temperature overshoots.

    I suspect it's quite unusual to schedule such very subtle changes.

    P.

    Comment

    • GAndrea
      Automated Home Lurker
      • Jan 2016
      • 3

      #3
      I think you are right...
      What I have is a big room with two radiators/valves linked to the central controller and I am trying to understand how to best manage the temperature in term of gas consumption.
      I would like to have low temperature at night (16), warm in the morning (18), then it could go down and up again in the evening (20).
      Doing like that I noticed that the boiler was always running at full blast during the changes. And, even with optimization on (with start set at 3 hours before), it was never able to reach the desired temperature on time. That is why I was trying to change the temperature of one degree only.
      But at the end I had to set always the same temperature (19). It seems that maintaining the same temperature is less expensive that turning on and off, but this could be another topic to open :-)
      Do you have any suggestion? Do you think that increasing the start time at more than 3 hours it could help?

      A.

      Comment

      • paulockenden
        Automated Home Legend
        • Apr 2015
        • 1719

        #4
        When your rooms aren't being used (at night) I'd be temped to sent them much lower than 16. What's the point of heating spaces that aren't occupied?

        P.

        Comment

        • Cchris
          Automated Home Sr Member
          • Nov 2015
          • 95

          #5
          Ive just set mine to optimum start, and I have set to allow up to 4 hours to warm up. Specifically because I have underfloor heating zones which may take longer to heat in the winter (as they would have to warm the whole slab and tiles etc). I would never anticipate that a single room would take that long to heat using radiators. (though other than holiday, I would probably not let those zones go overly cold overnight due to the energy required to heat the slab back up)

          If a room is taking that long to heat, then Id be looking at a list of things - boiler size, radiator size, pump size, is everything working as it should - are the locksheild valves closed too much etc... If its a domestic boiler putting out 30kw, then there is no reason that shouldnt be able to heat a very big room in well under 3 hours, and from a lot lower starting point than 16.

          I have an issue with a large room at the opposite end of the house taking a while to heat up, and a bedroom at the far end also taking a long time to heat. My issue is down to the pumps in the boilers not having enough puff to get the heat to that end of the house when other rooms are also calling for heat. If they are on on their own, and the only route for flow, then they are fine. Ive a low loss header and larger pump going in tomorrow, and will rebalance the system once the big pump is on.

          Im not a plumber, but Ive managed to learn a fair bit while having our house ripped apart and rebuilt including all new plumbing and elecs.

          What sort of size system are you running? Where is the room located in relation to the boiler? Are the radiators actually getting hot in that room when you expect them to be heating the room? Are they hot all over or are there cold areas on the radiators etc..

          My personal thought is that its not evohome that is the root of your problem here.
          Last edited by Cchris; 20 January 2016, 05:45 PM.

          Comment

          • DBMandrake
            Automated Home Legend
            • Sep 2014
            • 2361

            #6
            Originally posted by paulockenden View Post
            can you really tell the difference between 17 and 18 degrees? I'm not sure I could.

            The two temperatures are so close that Evohome will be operating in the TPI band, which is why you are seeing the boiler cycling. It's very hard to accurately get from 17 to 18, and if everything came on full-blast you'd almost certainly overshoot. As you'll have seen from other posts here, people don't like temperature overshoots.

            I suspect it's quite unusual to schedule such very subtle changes.

            P.
            I have to agree with this. If the radiator (and boiler duty cycle) was to come on full blast to try to increase from 17 to 18 degrees then it would almost certainly overshoot by at least 3 degrees. An easy way to demonstrate this is to have the room set at 17 degrees, (and up to temperature) and have at least one other room turned up high for a while so that the boiler is operating at 100% duty cycle and the system was hot.

            Then turn the HR92 in question up to 25 degrees for say 3 minutes to allow the valve to fully open and the radiator to fully heat, then set it back down to 10 degrees. In our bedroom in particular the thermal mass of the radiator is such that it will overshoot the current temperature by about 3 degrees performing a test like that.

            The only way the system can make an incremental change from 17 to 18 degrees is to open the HR92 valve slightly and run the boiler at a low duty cycle to cautiously increase the temperature of the room. My guess is that optimum start doesn't operate on 1 degree differentials on purpose. I do have some set point changes that step by 2 degrees, such as the bedroom, which comes on at 18 degrees at waking up time and then goes to 20 degrees after we are up but before we come back to dress for work, and optimum start does operate with this differential, and typically steps up the set point about half an hour before 20 degrees is scheduled, and does make it in time.

            Originally posted by paulockenden View Post
            When your rooms aren't being used (at night) I'd be temped to sent them much lower than 16. What's the point of heating spaces that aren't occupied?

            P.
            16 does seem quite high. One of the advantages of optimum start is that (in theory) you can set the overnight temperatures of unoccupied rooms as low as you want and it will be smart enough to turn the system back on in time to get it back up to the set point.

            I've settled on 12 degrees for unused rooms at night, partly as a poor mans frost protect as the boiler does not have frost protection (must do something about that...) and I've found that on moderate nights (down to about 0) the rooms do not get that low and thus the heating does not come on for those rooms, but if we get a really cold night like -5 the rooms get cold enough to trigger the heating to come on to maintain 12 degrees in some rooms - but they only require a low boiler duty cycle to do so, but enough to give some measure of frost protection...

            Comment

            • GAndrea
              Automated Home Lurker
              • Jan 2016
              • 3

              #7
              Originally posted by paulockenden View Post
              When your rooms aren't being used (at night) I'd be temped to sent them much lower than 16. What's the point of heating spaces that aren't occupied?

              P.
              There are several discussions regarding if it is better to maintain a constant temperature and let work the system at minimum or turn off and on and have the system working at the max power for a certain period. Unfortunately, I am not able to say exactly which strategy is the best one.

              Originally posted by Cchris View Post
              I would never anticipate that a single room would take that long to heat using radiators.
              The room is a large room which communicate with other areas without a door to close. However, the problem I have is that the system during the start period is not running at maximum and I don't understand the reason. As I wrote, the system not open the valves at 100% and it is not a full blast, which means lukewarm radiators for at least the first hour. That's why is taking that long.

              I am now making a new experiment letting go down the temperature to 14 degrees during the night and turning on at 18 increasing the warm time up to 4 hours to see what's happen.

              But, I have now a doubt. All the other rooms work well but they are all rooms with a single sensor. The one having problems has two valves using the controller as thermometer. Is there any particular setting to do in this case?

              A.

              Comment

              • DBMandrake
                Automated Home Legend
                • Sep 2014
                • 2361

                #8
                Originally posted by GAndrea View Post
                There are several discussions regarding if it is better to maintain a constant temperature and let work the system at minimum or turn off and on and have the system working at the max power for a certain period. Unfortunately, I am not able to say exactly which strategy is the best one.
                The answer is pretty clear - heat loss is proportional to the temperature differential between the inside of the room and the outside of the room.

                If it's zero degrees outside and you maintain the room at 10 degrees you use half the energy to maintain this 10 degree differential than you would to maintain the 20 degree differential of the room being 20 degrees and outdoors being zero degrees.

                Letting a room cool down over night and then having to heat it up again the next morning definitely uses less energy than keeping it warm all night.

                Condensing boilers do complicate things slightly because they tend not to run in condensing mode during the initial heatup phase if your flow temperature is set high, but the gain from operating in condensing mode is no more than about 20%, and to really gain the full benefit of a lower flow temperature when the zones are in "proportional" control mode would require an OpenTherm gateway which can actually specify a low flow temperature when demand is low.

                With a BRD91 it will use TPI control to attempt to approximate flow temperature modulation but it is not very accurate compared to using an OpenTherm bridge.
                The room is a large room which communicate with other areas without a door to close. However, the problem I have is that the system during the start period is not running at maximum and I don't understand the reason. As I wrote, the system not open the valves at 100% and it is not a full blast, which means lukewarm radiators for at least the first hour. That's why is taking that long.
                How are you determining that the valves are not fully open ? If you set the HR92 to more than 2 degrees above the current temperature then unlock it and remove it from the radiator after the motor has finished turning, can you manually open the valve significantly further by turning the black gear and hear the flow increase significantly in the radiator ?

                If so you need to use "Full stroke" mode, as you apparently have valves that require a greater pin travel for the valve to fully open. To do this hold down the button for 10 seconds until the configuration menu comes up, choose option 6, stroke, changing it's value to 1, then turn to the exit option and press the button again.

                After doing this I would recommend unlocking the HR92 and lifting it off the radiator, turn the black wheel fully anti-clockwise to fully open the radiator valve then lock the HR92 back on. After about a minute it will perform a calibration where it re-learns the full travel of the pin in the valve.

                From now on it will operate the valve pin over its maximum possible travel rather than about 70% of its travel that it uses by default, so when it opens it will open as much as possible and it will also close down a bit harder as well in the closing direction.

                It may take the HR92 a couple of days of normal use for it to calibrate itself to the new operating point for the valve - during this time it might overshoot or undershoot the target temperature a bit but it will adjust after a few days.

                Despite all my radiator valves being the same make and model, and only a year old, I have one valve where I have had to put the HR92 into full stroke mode because it doesn't always open sufficiently to allow enough flow to get up to temperature, like you describe. In full stroke mode it works fine, and none of my other valves need full stroke mode. (It's possible the pin is a little sticky)

                I am now making a new experiment letting go down the temperature to 14 degrees during the night and turning on at 18 increasing the warm time up to 4 hours to see what's happen.

                But, I have now a doubt. All the other rooms work well but they are all rooms with a single sensor. The one having problems has two valves using the controller as thermometer. Is there any particular setting to do in this case?
                Be careful that you aren't being fooled by that room having a remote sensor when all your other rooms use the HR92 for the temperature sensor.

                Measuring the temperature at the radiator gives a very different impression of how quickly the room warms up compared to measuring the temperature remotely. The remote sensor gives a "truer" indication of the temperature of the room and its warm up time, whereas measuring the temperature at the radiator will make the room appear to warm up much quicker than it really does, and seem warmer than it really is during that warm up time.. See the following thread for some discussion on this:



                I bet that if you go back to using an HR92 as the temperature sensor in the room you currently use the remote sensor will make it seem like the room is warming up much quicker, but in reality it is just measuring the warm side of the room, like all your other rooms.

                If you have heavy, solid internal walls, it can take a long time for the entire room to truly warm up, and the remote sensor accurately indicates this.
                Last edited by DBMandrake; 21 January 2016, 01:13 AM.

                Comment

                • The EVOHOME Shop
                  Site Sponsor
                  • Dec 2014
                  • 483

                  #9
                  Originally posted by DBMandrake View Post
                  If so you need to use "Full stroke" mode, as you apparently have valves that require a greater pin travel for the valve to fully open. To do this hold down the button for 10 seconds until the configuration menu comes up, choose option 6, stroke, changing it's value to 1, then turn to the exit option and press the button again.
                  Be careful doing and advising this. I have never had to use or advise using 'Full-stroke mode'. It is more than likely when the HR92 radiator controller was put on the valve baseplate initially, the black hand wheel was not all the way wound anti-clockwise - so when the radiator controller was put back onto the valve baseplate and has not calibrated correctly. I would suggest doing this first.

                  The 'stroke' on the HR92 is set to 4mm governed by the stops on the valve baseplate and changing to 'full-stroke' just increases the torque settings of the motor and decreases battery life of the HR92 - nothing else. It is more than likely the radiator valve is faulty if you have to do this, so I suggest replacing the valve.

                  Comment

                  • DBMandrake
                    Automated Home Legend
                    • Sep 2014
                    • 2361

                    #10
                    Not wanting to contradict you (I don't do this for a living after all! ) but I don't think things are quite as clear cut as you are implying.

                    Originally posted by The EVOHOME Shop View Post
                    Be careful doing and advising this. I have never had to use or advise using 'Full-stroke mode'. It is more than likely when the HR92 radiator controller was put on the valve baseplate initially, the black hand wheel was not all the way wound anti-clockwise - so when the radiator controller was put back onto the valve baseplate and has not calibrated correctly. I would suggest doing this first.
                    I have been through the calibration procedure many times and I can assure you that I do turn the black hand wheel fully anti-clockwise before fitting the HR92, so this is not the issue.

                    The 'stroke' on the HR92 is set to 4mm governed by the stops on the valve baseplate and changing to 'full-stroke' just increases the torque settings of the motor and decreases battery life of the HR92 - nothing else. It is more than likely the radiator valve is faulty if you have to do this, so I suggest replacing the valve.
                    Again, my testing suggests this is not the case.

                    It's true that enabling full-stroke mode does increase the motor torque a small amount - you can hear it from the speed increase. However it does also increase the number of revolutions it will rotate from fully closed to fully open, in both directions. Eg it will close down harder when closing the valve (probably due to the increased motor torque) but it will also open further when opening the valve. It actually says this in the installation manual as well:

                    Parameter 6 – Valve stroke
                    The radiator controller operates with the optimum valve
                    stroke set in the factory.
                    If the entire valve stroke is to be used or if the valve does
                    not open completely
                    , activate the full-stroke mode.
                    I have tested this back and forth multiple times now - in normal stroke mode it does not rotate the wheel fully anti-clockwise when opening, I can unlock it and lift it off and I still have another revolution or more available to turn the wheel. However in full stroke mode it does definitely open the valve fully until it hits the limit stop built into the HR92 base, as the installation manual implies. You actually hear it stop suddenly (and see a slight jump) when it gets to the limit stop in full stroke mode that you don't see in normal mode.

                    As far as I can see there is nothing wrong with the valve, and it is only a year old anyway. Different makes of valves have different operating points, it stands to reason that on some valves that the fully open operating point might be close to the limit of travel of the HR92 base, just due to the precise pin dimensions on the valve... in fact that is the case on my valves when listening to the flow and manually turning the wheel. I'm sure on other models of valves with slightly different pins the fully open point is probably two or three turns away from fully extended and thus doesn't need full stroke.

                    The operating point of some of the other valves in the house (all the same make, model and age) are very slightly different, but we're probably only talking half a millimetre difference in the tolerance of the pin extension.

                    Draining down the system in the winter to replace a valve that isn't really faulty, (but might have marginal dimensional compatibility with the HR92 base, granted) is just not going to happen. I'm happy to just set it to full stroke mode. If it wasn't occasionally needed to cope with some 3rd party valves they wouldn't have added it as a user configurable setting...
                    Last edited by DBMandrake; 21 January 2016, 01:56 PM.

                    Comment

                    • SteveP
                      Automated Home Guru
                      • Dec 2012
                      • 190

                      #11
                      I also have one rad out of 14 that doesn't quite open fully. Swapped around the different heads and no difference. Eventually tried the full stroke and it opened fully. However, did notice a big difference in battery life (I've had the evohome system in for 3 years now). As there is a second rad in the room then between them they do the job. So set it back to normal stroke and just notice it always closes much earlier than the other rad even though the room is not anywhere near its temp setting. So I have concluded that the rad valve must have a shorted control pin than normal. I actually found I could get it to open fully just by loosening the head - but then it doesn't always close properly! As DBMandrake - I also am not contemplating draining the system just for one valve that has eccentric tendencies

                      Comment

                      • The EVOHOME Shop
                        Site Sponsor
                        • Dec 2014
                        • 483

                        #12
                        Everyone is entitled to an opinion and I am here to be factual. So I suggest in the name of facts you attach the HR92 valve baseplate to a radiator valve and then see what difference 'Full-stroke mode' makes as far as dimensions of the pin it operates go.

                        In my previous testing, the standard M30 x 1.5mm valve baseplate has an internal dimension of 14mm from the pin in the fully open position to where the base plate will bottom out on the valve (at hand tightness). A Honeywell Valencia V120-15A co-incidentally has a dimension of 14mm too. When the pin is fully depressed on a V120-15A the dimension is 12mm. Therefore only 50% valve baseplate travel/stroke will fully close the valve from fully open. After the HR92 had calibrated the valve, the position at 0% was at the closed dimension (12mm) and the fully open position at 100% was 14mm. I have just re-tested this in 'Full-stroke mode' and it is exactly the same. The only difference now is that the torque has been increased and the battery life reduced.

                        I agree there is a variation between radiator valves, but based on my personal experience 'Standard valve stroke' is sufficient for the majority of installations with industry standard BS-EN radiator valves. Using 'Full-stroke mode' without complete justification will cause people to be replacing HR92 batteries more often.

                        Comment

                        • DBMandrake
                          Automated Home Legend
                          • Sep 2014
                          • 2361

                          #13
                          Originally posted by SteveP View Post
                          I also have one rad out of 14 that doesn't quite open fully. Swapped around the different heads and no difference. Eventually tried the full stroke and it opened fully. However, did notice a big difference in battery life (I've had the evohome system in for 3 years now). As there is a second rad in the room then between them they do the job. So set it back to normal stroke and just notice it always closes much earlier than the other rad even though the room is not anywhere near its temp setting. So I have concluded that the rad valve must have a shorted control pin than normal. I actually found I could get it to open fully just by loosening the head - but then it doesn't always close properly! As DBMandrake - I also am not contemplating draining the system just for one valve that has eccentric tendencies
                          Is that HR92's you're using or the older HR80's ? I didn't think that the HR92 had been out for 3 years... If its the HR80 then I'm not sure that the results would be directly transferable to the HR92, although it does echo what I see with the HR92 as well.

                          Personally I would rather keep warm and have my living room get up to temperature properly and in a timely fashion even if it means the battery only lasts for a year instead of two, in that one affected HR92! I think it's fairly clear that the reduced stroke setting (by default) is a battery saving measure and that with 95% of valves it works fine without full stroke.

                          Comment

                          • SteveP
                            Automated Home Guru
                            • Dec 2012
                            • 190

                            #14
                            Yep it is the HR80 that was the problem and I have some HR92s in the mix. You are correct and maybe the results aren't transferable although the actual actuators appear the same in both the HR80 and HR92 as I am able to swap the HR92s I have onto HR80 bases, but the behaviour does appear to be similar to yours. Just sharing In case anyone wondered, I move the HR92s around as they have a backlight and so great for rads that get the temperature locally changed a lot.

                            Comment

                            • DBMandrake
                              Automated Home Legend
                              • Sep 2014
                              • 2361

                              #15
                              Originally posted by The EVOHOME Shop View Post
                              Everyone is entitled to an opinion and I am here to be factual. So I suggest in the name of facts you attach the HR92 valve baseplate to a radiator valve and then see what difference 'Full-stroke mode' makes as far as dimensions of the pin it operates go.
                              No real need to do this - what's invalid in test I already described ? After being calibrated turn the temperature down low so the valve "fully" closes, then turn it up high again so that it fully opens. Then unlock the HR92, lift it off, and check how many turns away from fully open the wheel is. On this particular valve it is approximately 1-2 turns away from hitting the built in limit stop in the HR92 adaptor base. (I will test it again and get a more accurate figure) Turning the wheel the remaining revolutions does make a very obvious difference to the radiator flow and temperature (measured with an IR gun) on this particular valve/radiator.

                              Change to full stroke and let it recalibrate then follow the same procedure of allowing it to fully close then fully open - in this case you can actually hear it hit the limit stop and when I remove the HR92 I find that the black wheel is fully unwound to the limit stop. On my particular valve there is no doubt whatsoever that full stroke fully opens the valve and standard stroke does not. No opinion required, just a simple, sound test. I'm not saying that it's a common occurrence in general by any means.

                              In my previous testing, the standard M30 x 1.5mm valve baseplate has an internal dimension of 14mm from the pin in the fully open position to where the base plate will bottom out on the valve (at hand tightness). A Honeywell Valencia V120-15A co-incidentally has a dimension of 14mm too. When the pin is fully depressed on a V120-15A the dimension is 12mm. Therefore only 50% valve baseplate travel/stroke will fully close the valve from fully open. After the HR92 had calibrated the valve, the position at 0% was at the closed dimension (12mm) and the fully open position at 100% was 14mm. I have just re-tested this in 'Full-stroke mode' and it is exactly the same. The only difference now is that the torque has been increased and the battery life reduced.
                              It sounds like your valve has a fairly limited travel of only 2mm, which would explain why you see no difference between the modes. I will have to check now as you have got me thinking about it, but I think my valves have significantly more travel than that. Will have a play over the weekend and measure the exact pin travel to see how it compares with your measurements.

                              The calibration process begins by closing the valve as much as possible until the motor reaches a certain load, this then counts as 0% and the next time the valve tries to fully open it calibrates that as 100%. (The calibration process can be observed by watching the valve pos figure while it is calibrating) It seems from my testing though that in reduced stroke mode it will only open by a maximum of X number of turns from the calibrated closed position, rather than opening until it hits the limit stop. So if you have a valve that fully opens within that number of turns you won't see a difference in performance, however if you have a valve that opens over wider range of movement it may stop before the valve is fully open.
                              I agree there is a variation between radiator valves, but based on my personal experience 'Standard valve stroke' is sufficient for the majority of installations with industry standard BS-EN radiator valves. Using 'Full-stroke mode' without complete justification will cause people to be replacing HR92 batteries more often.
                              I was certainly not advocating using it all the time - I am only using it on the one valve where I seem to need it and not the other valves in the house, and have only suggested it to two people who were both potentially suffering from the same issue as me, and only suggested that they try it before they go to the trouble of replacing a radiator just in case the valve may not be fully opening like mine.
                              Last edited by DBMandrake; 21 January 2016, 04:58 PM.

                              Comment

                              Working...
                              X