Page 2 of 4 FirstFirst 1234 LastLast
Results 11 to 20 of 31

Thread: HR92's: A rant

  1. #11
    Automated Home Guru
    Join Date
    May 2017
    Posts
    165

    Default

    From only a few days use, the temperature sensors on the HR92 are wildly wrong in every room... They are frequently reading 20-21c when the room is 15c.

    My radiators do run pretty hot; putting a sensor 1" away from 80c metal it's hard to see how it can work accurately.

    Actually this might be a reason for plastic parts, to restrict heat conduction from the valve body?

  2. #12
    Automated Home Jr Member
    Join Date
    Dec 2017
    Posts
    13

    Default

    Quote Originally Posted by MrBoy View Post
    From only a few days use, the temperature sensors on the HR92 are wildly wrong in every room... They are frequently reading 20-21c when the room is 15c.
    There's a good discussion about this here, in particular see @DBMandrake's post #3.

    When the room is at a steady state, I normally find the HR92s measure about 1C higher than a known-good thermometer in the middle of the room. This offset can be corrected on each HR92 individually using option 8 in the settings menu.

  3. #13
    Automated Home Ninja Dan_Robinson's Avatar
    Join Date
    Jun 2012
    Posts
    347

    Default

    Quote Originally Posted by MrBoy View Post
    My radiators do run pretty hot; putting a sensor 1" away from 80c metal it's hard to see how it can work accurately.
    Physics..... Convection.
    Kind Regards - Dan Robinson (Jennings Heating Ltd)

  4. #14
    Automated Home Guru
    Join Date
    May 2017
    Posts
    165

    Default

    Quote Originally Posted by Dan_Robinson View Post
    Physics..... Convection.
    you might've been selling these things a decade but I've a degree in physics and that's not how it works. Radiators largely work by convection as we all know but they do radiate some heat and at close range this is significant... Hold your hand an inch from a hot radiator and you can feel it.
    The TRV being at the side should minimise the effect, and at the bottom it might be partially mitigated by the cold air from the convection current being sucked back in but not fully.
    Then of course your radiator transmits heat a third way... Physics... Conduction... But only to things touching it. Your valve body is likely to be roasting hot which will to some degree be conducted into the head. When I removed my old TRVs with metal fixings, some were so hot I could barely do it.

    I guess a well ventilated radiator suffers less but sadly many are stuck in corners or next to things so the TRV is likely to end up in a pocket of unusually warm air. No physics can fix this... Move your rad, put the TRV in a better position, or get a room thermostat

  5. #15
    Automated Home Guru
    Join Date
    Apr 2016
    Posts
    189

    Default

    Initially the radiator will be hot and sensor will likely read high. As room gets nearer to the set temp radiator will be cooler and have less affect on sensor. Eventually a steady state will be reached with radiator only providing enough heat to maintain set temp and having less influence on sensor. Having the head set at 90 degrees to the radiator would be better than vertical orientation but looks unusual and is more vulnerable to being used as a foot hold for adventurous toddlers.

  6. #16
    Automated Home Legend paulockenden's Avatar
    Join Date
    Apr 2015
    Location
    South Coast
    Posts
    1,719

    Default

    Like MrBoy I also have a Physics degree. A couple of points to note: 1) the white casing of the HR92 will reduce the absorption of radiated heat, 2) TRV mounted sensors do the best they can, but will always be a compromise. That's why Honeywell has at least four different room stats to choose from for use in zones where the HR92 sensing doesn't work well.

  7. #17
    Site Sponsor The EVOHOME Shop's Avatar
    Join Date
    Dec 2014
    Location
    Powys, Mid Wales
    Posts
    483

    Default

    I love Physics! I don't have a degree I can boast about but I do have a HNC in Mechanical Engineering and 20 years heating experience (well 26 if you classify helping my Dad strip commercial boilers down when I was a kid).

    So radiators are white to promote convection and from your physics days you will probably remember the cube... dull dark surfaces radiate much more heat than light shiny surfaces. The very reason the HR92 is white is for the same reason and the sensor is positioned as best as possible away from the radiator. If it doesn't work because you have black radiators or multiple rads in one room, just use a well positioned external sensor like the DTS92E or T87RF.

    I love when you guys argue with evohome installers about the one evohome installation you have done in your home... To underestimate Dan's evohome and heating experience would be silly.

  8. #18
    Automated Home Ninja Dan_Robinson's Avatar
    Join Date
    Jun 2012
    Posts
    347

    Default

    I'll have to find the one I pulled apart recently. Iirc there is a sensor either side of the PCB and the controller uses the average of the two.
    Kind Regards - Dan Robinson (Jennings Heating Ltd)

  9. #19
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    2,361

    Default

    Quote Originally Posted by Dan_Robinson View Post
    I'll have to find the one I pulled apart recently. Iirc there is a sensor either side of the PCB and the controller uses the average of the two.
    That's what I was wondering. Somewhere in a really old post of mine I theorised that it had a sensor on either side of the HR92 and used the lower of the two readings.

    And in fact by comparing the two readings it could estimate how much influence there is from the direct heat of the side of the radiator. If one side reads quite a bit hotter than the other it's safe to assume the radiator is currently hot and that the lower of the two readings is closer to room temperature! And if the radiator is cold they'll probably both read identically.

    You could even go a step further - you could try to estimate the temperature gradient beside the radiator from the difference in the two readings to extrapolate a "true" reading from further away. So say the sensor on the radiator side says 22 and the one on the opposite side of the HR92 away from the radiator says 20, the true reading is probably below 20, perhaps 18.

    Whether it does that and how accurate that could be I'm not sure. It may just take the lower of the two readings.

    In any case I wonder if any HR92's mounted "sideways" will work less accurately, since the two sensors would both be the same distance from the radiator then ? I have a couple of radiators like my hall where I have to point the HR92 with screen extended along the wall instead of out from the wall, otherwise it will just get kicked/knocked.

  10. #20
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    2,361

    Default

    Quote Originally Posted by MrBoy View Post
    From only a few days use, the temperature sensors on the HR92 are wildly wrong in every room... They are frequently reading 20-21c when the room is 15c.

    My radiators do run pretty hot; putting a sensor 1" away from 80c metal it's hard to see how it can work accurately.

    Actually this might be a reason for plastic parts, to restrict heat conduction from the valve body?
    Quote Originally Posted by MrBoy View Post
    you might've been selling these things a decade but I've a degree in physics and that's not how it works. Radiators largely work by convection as we all know but they do radiate some heat and at close range this is significant... Hold your hand an inch from a hot radiator and you can feel it.
    The TRV being at the side should minimise the effect, and at the bottom it might be partially mitigated by the cold air from the convection current being sucked back in but not fully.
    Then of course your radiator transmits heat a third way... Physics... Conduction... But only to things touching it. Your valve body is likely to be roasting hot which will to some degree be conducted into the head. When I removed my old TRVs with metal fixings, some were so hot I could barely do it.

    I guess a well ventilated radiator suffers less but sadly many are stuck in corners or next to things so the TRV is likely to end up in a pocket of unusually warm air. No physics can fix this... Move your rad, put the TRV in a better position, or get a room thermostat
    Not wanting to repeat too much of what I said in that post Ian_W linked to but they're not as inaccurate as you're making out, and it depends highly on the individual radiator design, placement, room conditions etc as to whether you can get a reasonable estimate of room temperature or not - in many rooms you can, in some you can't.

    You'll get a more accurate reading with a modern convector vs an older panel without convecting fins (where a lot higher proportion of the total heat into the room is via direct IR radiation, which tends to lead to a greater heat gradient across the room) and it also depends on whether your radiator has "side skirts" or not. (don't know their proper name sorry) The new radiators I've installed all have removable side skirts - these are right beside where the HR92 sits and they barely get warm when the radiator panel is completely hot. This will block the majority of the IR radiation from the panel directly into the HR92 and lead to a more accurate reading.

    Another factor is whether you have the HR92 on the flow or return side of the radiator. Even though modern valves are all bidirectional you'll get more accurate temperature sensing on the colder return side...

    The big thing you're overlooking though when you say the HR92 is reading 21 when the middle of the room is 15 degrees is that when the room has not yet reached equilibrium there is a large temperature gradient across the room. Physics.

    Out of that 6 degree "error" only 1 or 2 degrees is due to the HR92 being beside the radiator and being influenced by direct heat, the rest is genuinely that the further parts of the room are still cold and heat is still flowing from the radiator side of the room to the cold side of the room. The HR92 is correct about the temperature of its side of the room, within a degree or so.

    What you should find is that in a closed room over time the room will eventually reach equilibrium - but it can take many hours. Once that happens the HR92 reading will only be about 1 degree higher than the "true" room temperature away from the radiator, so a calibration of -1C on the HR92 is usually about right.

    But it's important to realise that as the room warms up the HR92 will report the room is up to temperature long before it really is because there is still a large temperature gradient across the room. It will start throttling the radiator back because it thinks the room is up to temperature when it isn't, this slows down the process of reaching equilibrium by quite a bit. As the radiator cools and the room warms the room will start to come much closer into agreement with the reported temperature.

    This only applies to a closed room. If you have a room that is open at the far end from the radiator into a colder space then there will always be a gradient across the room with the HR92's reported temperature being "optimistic" because its end of the room is warmer than the other end.

    The same is not true with a remote sensor some distance from the radiator. As well as having no direct influence from the radiator, it is sampling the room further along the temperature gradient in the room. This will actually lead to the radiator side of the room "overshooting" the target temperature for a while (although the system won't indicate this) as the room reaches equilibrium, but it will allow the room to reach an equilibrium much faster as it won't prematurely throttle back the radiator.

    When we used the HR92 as the sensor in our living room with a -1C calibration, average winter warm up time from say 14 degrees to 21 was only about 1 1/2 hours until the system thought the room was at temperature and was throttling back the radiator. However measured out in the room it would only be about 17 degrees when the system first said 21, and it would take a further 3 hours or so for it to gradually creep up to 21 in the room and reach a steady equilibrium, with the HR92 reporting 21 the whole time, as it gradually throttled the radiator further back - and that was assuming the door was kept closed or the hallway was also warm. Total warmup time to get within a degree of the target over 4 hours...

    When we changed to a DTS 92 in the Living room as sensor half a room away the reported warmup time "increased" from 1 1/2 to about 2 1/2 hours in the same conditions, however this is now "true" warmup - by that time the room really was 21 degrees away from the radiator. So even though the system makes it seem as if the room is warming up slower with the remote sensor this is a true indication and the real warmup time had reduced from over 4 hours to 2 1/2 hours.

    So yes, a remote sensor is always better - more accurate, not influenced by direct heating from the radiator, and it actually improves the warm up time due to not throttling the radiator too soon and provides various other benefits like not needing to adjust your set points between summer and winter. But in appropriate circumstances sensing at the HR92 is not nearly as inaccurate as you suggest - you just need to keep in mind the gradient across the room and the fact that it takes a long time for a room (especially with solid walls like ours) to reach an equilibrium and for that gradient to minimise.

    Be thankful that Honeywell do offer wall sensors as an option BTW - some like Tado do not. Well, not unless you want to pay over £100 each for the full main thermostat box with boiler switching circuitry that you won't use for every zone. My Tado V3 owning colleague was just bemoaning a couple of days ago that Tado don't make a reasonably priced standalone wall sensor, as he was looking to get some for some of his rooms, but not that that price!
    Last edited by DBMandrake; 18th January 2018 at 04:22 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
  •