Results 1 to 15 of 15

Thread: Are mercury thermometers more acurate than digital thermometers?

  1. #1
    Senior Member
    Join Date
    Sep 2015
    Location
    Turkey
    Posts
    285
    Post Thanks / Like

    Default Are mercury thermometers more acurate than digital thermometers?

    I had 2 thermometers. One is digital and one is mercury.

    When it was above 0°C or sligtly below zero, they both showed the same temperature but when it was well below 0°C, the mercury thermometer showed colder than the digital one.

  2. #2
    Senior Member Larkenfield's Avatar
    Join Date
    Jun 2017
    Location
    Sedona
    Posts
    3,179
    Post Thanks / Like
    Blog Entries
    4

    Default

    Well, I would consider that the digital could be affected by extreme temperatures and register a different reading than the mercury which has no electronics. It’s another example of the difference between digital and analog and that they can be close but aren’t necessarily equivalent.
    Last edited by Larkenfield; Apr-25-2019 at 06:51.
    "That's all Folks!"

  3. Likes EddieRUKiddingVarese liked this post
  4. #3
    Senior Member CnC Bartok's Avatar
    Join Date
    Jun 2017
    Location
    Oxford-ish
    Posts
    809
    Post Thanks / Like

    Default

    Depends on how its calibrated. If a thermometer is standardised to a fixed value (eg the triple point of water - if you can be bothered!) then both a mercury and a digital will be both as accurate as each other. However, mercury does not expand with temperature in a perfectly linear fashion. Thus it will be perfectly accurate at all its calibrated points, not necessarily in between. This is particularly true at low temperatures, when mercury is getting close to its melting point (-40 or so off the top of my head), or indeed close to its boiling point (350C?). Then again the thermistor in a digital thermometer won't correspond in a perfect linear fashion either, so it will vary in accuracy as well, depending on where it is calibrated! Neither is "better", unless you like sticking fragile glass containing a highly toxic liquid metal in your mouth!!

    This is accuracy. Being a pedant for a change, precision is not the same thing, it's how many decimal points you can measure to! Similarly resolution is to do with how precisely (!) one can actually read an instrument! All these words seem very interchangeable these days...
    Last edited by CnC Bartok; Apr-25-2019 at 14:42.

  5. Likes Becca, Pat Fairlea liked this post
  6. #4
    Senior Member senza sordino's Avatar
    Join Date
    Oct 2013
    Posts
    3,471
    Post Thanks / Like

    Default

    I'm no expert on the construction of thermometers. What I do know is that digital thermometers measure the temperature using some sort of an analog device: a thermocouple / thermistor. It probably measures how the electrical resistance changes with temperature. There is some sort of analog to digital converter to give a digital LCD display. Our physical world is analog.

    A digital thermometer is converting an analog signal to a digital signal, which is easier to read than an analog scale.

    Everything CnC Bartok wrote is correct.
    Last edited by senza sordino; Apr-26-2019 at 06:45. Reason: Proof reading helps me

  7. Likes CnC Bartok liked this post
  8. #5
    Senior Member philoctetes's Avatar
    Join Date
    Jun 2017
    Posts
    778
    Post Thanks / Like

    Default

    One can think of many error causes with both methods. For example, there must be an A2D component that has to convert a thermocouple voltage to a digital reading... basically a digital voltmeter... and linearity is an issue with both methods

    Temperature is an odd duck. as a physical quantity, we feel it easily but defining it is complicated, much of thermodynamics is pretty strange all over...

  9. #6
    Senior Member KenOC's Avatar
    Join Date
    Mar 2011
    Location
    SoCal, USA
    Posts
    18,658
    Post Thanks / Like

    Default

    I believe thermometers specify a usable range and a possible error bar within that range. For instance, -20F to +140-F, +/- 1 degree F. In theory at least, two thermometers of any type should be accurate within their specifications.

    I'm looking at two cheap digital thermometers on my desk right now. One reads 75.6F, the other 76.1F. I believe the specs on each are +/- 2 degrees F, so I'm comfortable that they're within specified accuracy. Just grabbed a third, a cheap Taylor digital dipstick, that reads 75.2F.


  10. #7
    Senior Member
    Join Date
    Sep 2015
    Location
    Turkey
    Posts
    285
    Post Thanks / Like

    Default

    Quote Originally Posted by KenOC View Post
    I believe thermometers specify a usable range and a possible error bar within that range. For instance, -20F to +140-F, +/- 1 degree F. In theory at least, two thermometers of any type should be accurate within their specifications.

    I'm looking at two cheap digital thermometers on my desk right now. One reads 75.6F, the other 76.1F. I believe the specs on each are +/- 2 degrees F, so I'm comfortable that they're within specified accuracy. Just grabbed a third, a cheap Taylor digital dipstick, that reads 75.2F.
    Yeah but the point is, the digital one dropped less compared to the mercury one at well below temps (°C).

    When it was above zero or sligtly below zero, the difference was no higher than 0.5°C between them. But when it was well below zero, the difference got as high as 2°C ( when the mercury thermometer was showing as cold as -30°C) and official weather station was showing even colder ( 2°C colder than the mercury one).

    Readings are -32°C, -30°C and -28°C.

    Official weather station: -32°C
    Mercury termometer: -30°C
    Digital thermometer: -28°C
    Last edited by atsizat; Apr-26-2019 at 02:57.

  11. #8
    Senior Member KenOC's Avatar
    Join Date
    Mar 2011
    Location
    SoCal, USA
    Posts
    18,658
    Post Thanks / Like

    Default

    If two thermometers are each accurate to +/-2 degrees, then they could (theoretically) read 4 degrees apart and still be in spec! And they might agree (rightly or wrongly) at one temperature and differ, maybe significantly, at another. I wouldn't ascribe any difference to the type of thermometer.

    Of course you can spend some extra for, say, ±0.4°F guaranteed accuracy.
    Last edited by KenOC; Apr-26-2019 at 05:34.


  12. #9
    Senior Member Fritz Kobus's Avatar
    Join Date
    Oct 2011
    Location
    Next to Detroit, Michigan
    Posts
    12,008
    Post Thanks / Like
    Blog Entries
    1

    Default

    Quote Originally Posted by atsizat View Post
    Yeah but the point is, the digital one dropped less compared to the mercury one at well below temps (°C).
    Can't really determine anything from two thermometers like that. You need ten digital of the same model and ten mercury of the same model. Then record all of them over some temperature variations. After that, unless it is very obvious that the vast majority of one kind of thermometer varied vs the vast majority of the other kind, you will need to do some statistical analysis. Of course variation within a kind will be interesting, but as long as it is within spec (as Ken notes) then it should not throw off your comparison of the two different types of thermometers. This could get pretty complicated.
    Last edited by Fritz Kobus; Apr-26-2019 at 06:25.
    "All of Italian opera can be heard in [Bellini's] "Ah! non creda [mirarti]."
    --Renata Scotto in "Scotto, More Than a DIva."

  13. #10
    Senior Member DaveM's Avatar
    Join Date
    Jun 2015
    Location
    So. California, USA
    Posts
    2,392
    Post Thanks / Like

    Default

    Speaking of inaccurate mercury thermometers: A man called Wunderlich in the 19th century determined that the normal body temperature was 98.6 F, but he was using a flawed mercury thermometer. More recent evidence indicates that it is closer to 98.2 F.

  14. Likes Fritz Kobus liked this post
  15. #11
    Senior Member
    Join Date
    Sep 2015
    Location
    Turkey
    Posts
    285
    Post Thanks / Like

    Default

    0.2°C difference is not important at all. His mercury temperature was way so accurate then. I would want one.
    Last edited by atsizat; Apr-27-2019 at 02:35.

  16. #12
    Senior Member Fritz Kobus's Avatar
    Join Date
    Oct 2011
    Location
    Next to Detroit, Michigan
    Posts
    12,008
    Post Thanks / Like
    Blog Entries
    1

    Default

    Best thermometer: You freeze your bottom off. Then you know it is TOO COLD!
    "All of Italian opera can be heard in [Bellini's] "Ah! non creda [mirarti]."
    --Renata Scotto in "Scotto, More Than a DIva."

  17. #13
    Senior Member CnC Bartok's Avatar
    Join Date
    Jun 2017
    Location
    Oxford-ish
    Posts
    809
    Post Thanks / Like

    Default

    "Is that an aynal thermometer in your top pocket, Doctor?"

    "Yes. Some a***hole has got my pencil"

    Ahhh! The old ones are the best! Deliberate bad spelling circumvents the censorship, btw!
    Last edited by CnC Bartok; Apr-27-2019 at 12:13.

  18. #14
    Senior Member CnC Bartok's Avatar
    Join Date
    Jun 2017
    Location
    Oxford-ish
    Posts
    809
    Post Thanks / Like

    Default

    Quote Originally Posted by Fritz Kobus View Post
    Can't really determine anything from two thermometers like that. You need ten digital of the same model and ten mercury of the same model. Then record all of them over some temperature variations. After that, unless it is very obvious that the vast majority of one kind of thermometer varied vs the vast majority of the other kind, you will need to do some statistical analysis. Of course variation within a kind will be interesting, but as long as it is within spec (as Ken notes) then it should not throw off your comparison of the two different types of thermometers. This could get pretty complicated.
    I am glad to see some people can hold onto the principles of "how to design a scientific experiment". Nice one, Fritz!

  19. Likes Fritz Kobus liked this post
  20. #15
    Senior Member Larkenfield's Avatar
    Join Date
    Jun 2017
    Location
    Sedona
    Posts
    3,179
    Post Thanks / Like
    Blog Entries
    4

    Default

    ...............
    Last edited by Larkenfield; Apr-27-2019 at 22:46.
    "That's all Folks!"

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •