Hr92 v other temperature source

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • DeeJitsu
    Automated Home Jr Member
    • Oct 2018
    • 10

    Hr92 v other temperature source

    I think the picture says it all. No doubt this has been covered before but i cannot find an exact match.

    any ideas? i appreciate there could be a difference but nearly 3 degrees seems extreme.

    temp.jpg
  • anniesboy
    Automated Home Sr Member
    • Aug 2019
    • 71

    #2
    Temperature showing on Hr92 is set temperature not room temperature.
    Hence I have my heating off all Hr92 show 10c.

    Comment

    • DBMandrake
      Automated Home Legend
      • Sep 2014
      • 2361

      #3
      HR92's show set point by default but you can change it to display measured temperature in the HR92 menu.

      Comment

      • paulockenden
        Automated Home Legend
        • Apr 2015
        • 1719

        #4
        And even if you change it to measured temp it's a) to the nearest .5 degrees, and b) biased towards the setpoint rather than obeying normal rounding rules.

        P.

        Comment

        • DeeJitsu
          Automated Home Jr Member
          • Oct 2018
          • 10

          #5
          nope, it is 100% room temperature as i have changed the setting on the hr92. The calibration was at -1, so changed that, does it reflect on current temperature? After 2 hours, it did change slightly, so the difference was down to 1.5 which is acceptable.

          thanks for replies

          Comment

          • benwillcox
            Automated Home Jr Member
            • Apr 2020
            • 32

            #6
            Originally posted by DeeJitsu View Post
            so the difference was down to 1.5 which is acceptable.

            thanks for replies
            That depends, maybe the HR92 is now more wrong?! I think you need a third device to decide which one is measuring incorrectly :-)

            Ben

            Comment

            • G4RHL
              Automated Home Legend
              • Jan 2015
              • 1580

              #7
              I can demonstrate the same result with two different temperature sensors side by side in a room. At one stage I had three lined up and all showed slightly different temperatures. Same with humidity readings. I think also the HR92 has an inbuilt weighting to compensate for its usual location. Not sure on that but am sure I read it somewhere.

              Comment

              • DBMandrake
                Automated Home Legend
                • Sep 2014
                • 2361

                #8
                Originally posted by DeeJitsu View Post
                nope, it is 100% room temperature as i have changed the setting on the hr92. The calibration was at -1, so changed that, does it reflect on current temperature? After 2 hours, it did change slightly, so the difference was down to 1.5 which is acceptable.

                thanks for replies
                The calibration will affect the reading, yes, that's the point of it. In operation on an actual radiator you'll probably want to set calibration to -1 anyway to partially compensate for the direct heat from the radiator.

                If the temperature was still out by 1.5C after 2 hours then I'd say the device you're comparing it to is not very accurate. We have one of those gro egg thermometers as well and they are not very accurate, certainly not accurate enough to be checking the calibration of something else against. The problem is the lights inside them generate a lot of heat and raise the temperature reading artificially. I long ago gave up on using it for measuring temperature - it was only used as a night light. The HR92's built in sensor is much more accurate than the gro egg partly because there are no significant heat sources inside it.

                When you're using the sensor in a TRV the thing that really matters is the room temperature achieved. So install it on the radiator, place an accurate thermometer (I have the receiver unit of my outdoor weather station) on a table in the room, set the room to a constant temperature and close the door for a few hours to see what temperature is finally reached. If it's lower than the set point (which it usually will be) then adjust the calibrate setting in the negative direction. So if you have the set point at 20C and the room stabalises at 19C set the calibrate to -1C.

                If you're REALLY picky about having the room temperauture exactly as you set it (as I am) then I'm afraid a remote wall sensor is the only way to go as measuring the temperature beside the radiator is always only an approximation to the actual room temperature even after calibrated due to the influence of direct heat from the radiator.
                Last edited by DBMandrake; 17 April 2020, 09:14 AM.

                Comment

                • DBMandrake
                  Automated Home Legend
                  • Sep 2014
                  • 2361

                  #9
                  Originally posted by G4RHL View Post
                  I can demonstrate the same result with two different temperature sensors side by side in a room. At one stage I had three lined up and all showed slightly different temperatures. Same with humidity readings. I think also the HR92 has an inbuilt weighting to compensate for its usual location. Not sure on that but am sure I read it somewhere.
                  Nope, no inbuilt weighting.

                  If you set calibrate to 0 the internal sensor is very accurate (down to about 0.2C) in what it sends to the controller. One issue is that the temperature shown on the display is rounded 0.5C towards the set point and also to the nearest 0.5C. So the reading shown may be up to 0.5C in error.

                  However if you get the exact (non-biased, non-rounded) temperature from the Honeywell API or an HGI80 you can get a very precise reading that is not affected by this rounding.

                  I did a test once of an HR92 with calibrate 0 sitting disconnected on a table top, beside it a DTS92E and my Bresser weather station. Left them for a few hours in a closed room to fully equalise and the readings between the weather station and two Honeywell devices (as measured through the API) was extremely close - I don't remember the exact figures but all three were within +/- 0.2C of each other.

                  One large source of errors when doing these kind of comparisons is the thermal constant of different thermometers can be drastically different - some are fast to respond to changes in temperature, some are slow. So if the room temperature is changing (or the thermometers have only just been put in place for the test - same effect) there can be a large difference in the reading while they settle. That's why it's important to leave them a few hours in a constant temperature room when doing a precise comparison.

                  So in my opinon the sensor in the HR92's is very accurate on the order of about 0.2C or better. However all the cavets of having a sensor beside a radiator apply - it's no good having a sensor that is accurate to better than 0.2C if the direct heat from the radiator is going to increase the reading by 1-2C. Calibrate helps but doesn't deal very well with changes in conditions like outdoor temperature. All TRV based temperature measurements will suffer from this effect so for those who really care about this stuff a remote sensor is the way to go.

                  Comment

                  • DeeJitsu
                    Automated Home Jr Member
                    • Oct 2018
                    • 10

                    #10
                    Originally posted by DBMandrake View Post
                    The calibration will affect the reading, yes, that's the point of it. In operation on an actual radiator you'll probably want to set calibration to -1 anyway to partially compensate for the direct heat from the radiator.

                    If the temperature was still out by 1.5C after 2 hours then I'd say the device you're comparing it to is not very accurate. We have one of those gro egg thermometers as well and they are not very accurate, certainly not accurate enough to be checking the calibration of something else against. The problem is the lights inside them generate a lot of heat and raise the temperature reading artificially. I long ago gave up on using it for measuring temperature - it was only used as a night light. The HR92's built in sensor is much more accurate than the gro egg partly because there are no significant heat sources inside it.

                    When you're using the sensor in a TRV the thing that really matters is the room temperature achieved. So install it on the radiator, place an accurate thermometer (I have the receiver unit of my outdoor weather station) on a table in the room, set the room to a constant temperature and close the door for a few hours to see what temperature is finally reached. If it's lower than the set point (which it usually will be) then adjust the calibrate setting in the negative direction. So if you have the set point at 20C and the room stabalises at 19C set the calibrate to -1C.

                    If you're REALLY picky about having the room temperauture exactly as you set it (as I am) then I'm afraid a remote wall sensor is the only way to go as measuring the temperature beside the radiator is always only an approximation to the actual room temperature even after calibrated due to the influence of direct heat from the radiator.
                    thanks for the detailed response man, great help. Had a good idea the Egg wasnt the best device to use, was just a starting place. To be honest i am not that picky about exact readings, i would tolerate anything around 2 degress difference. Especially as there can be a variation in readings from various sources.

                    Comment

                    Working...
                    X