Thermometer Accuracy

[Robbo3]

So as not to encroach on NikNak's weather station thread - How the heck do you know if the temperature reading is accurate?

I have one conventional & three digital outdoor thermometers which when placed side by side, vary by over 10 degrees.

 

[Jacob]

Put them in water with ice. The ones that settle at around 0º are OK (at 0º at least!).
Then ditto with boiling water and 100º.
But in any case it should be bloody obvious which ones are wrong if you compare them with the weather forecast.

 

[bugbear]

So as not to encroach on NikNak's weather station thread - How the heck do you know if the temperature reading is accurate?

I have one conventional & three digital outdoor thermometers which when placed side by side, vary by over 10 degrees.

If they're water proof, just put them all in the same container of water. They should read the same.

If they're in wooden cases or something, they'd be damaged, of course.

BugBear (proud owner of a NPL calibrated thermometer, cost £3.00 )

 

[mind_the_goat]

The conventional on is probably more accurate, assuming the scale is printed correctly. The colour, shape and mass could all cause them to react differently to external inputs. What you need is a Stevenson Screen to put them in.
Wouldn't recommend putting the digital ones in water, either boiling or melting, but you can calibrate the glass one then use that to compare against the others.

 

[Jacob]

.....
If they're water proof, just put them all in the same container of water. They should read the same....

It wouldn't tell you if any of them was accurate.

 

[CHJ]

Just take heed of the temperature range if conventional mercury patterns when comparing.
Not all cover the whole range between 0C and 100C. I have seen burst mercury items that have been casually placed in hot liquid without due regard to working range.

One of my Paterson Certified Thermometers only covers 13C to 30C over a 112mm scale for instance. (quick check source for metrology checks)

 

[mjcann]

There may be facility to calibrate, if there is get hold of a mercury or alchohol thermometer, best to do it mid range ie, room temp, the adjust to match, if you need accuracy you realy need something that is able to be calibrated, this is especialy true with humidity meters, the tool to use is a whirling hygrometer, this has 2 mercury thermometers one of which has a wet sock over the bulb, so as you swing it round the moisture evaporates cooling the bulb, the resultant difference in temp between the 2 thermometers calculates accuratley relative humidity

 

[nev]

which also brings about the question, is my *insert any measuring type tool* correct? Is a cheap steel rule accurate? or any more accurate than a 'quality' one?
My own measuring and cutting skills are so dire it makes no difference to me, but the mind boggles at the amount of opportunities for non accuracy during the manufacturing process of a measurer, the expansion and contraction qualities of the materials, accuracy of the machines that build the machines that build the machines etc

 

[bugbear]

... the mind boggles at the amount of opportunities for non accuracy during the manufacturing process of a measurer, the expansion and contraction qualities of the materials, accuracy of the machines that build the machines that build the machines etc

Here's the book you need:

http://www.mooretool.com/publications.html

"“Foundations of Mechanical Accuracy” Wayne R. Moore – 1970

BugBear

 

[nev]

... the mind boggles at the amount of opportunities for non accuracy during the manufacturing process of a measurer, the expansion and contraction qualities of the materials, accuracy of the machines that build the machines that build the machines etc

Here's the book you need:

http://www.mooretool.com/publications.html

"“Foundations of Mechanical Accuracy” Wayne R. Moore – 1970

BugBear

I have no need for such a book, ''about a mil'' is as accurate as i need but youd have thought that the manufacturers of each of Robbo's thermometers were adhering to the same standard but it seems not to be.

 

[bugbear]

I have no need for such a book, ''about a mil'' is as accurate as i need but youd have thought that the manufacturers of each of Robbo's thermometers were adhering to the same standard but it seems not to be.

Maybe they're all made to "plus or minus 5" tolerance...

BugBear

 

[Kalimna]

If there is an option to calibrate the digital thermometers, then checking, as Jacob mentioned, against two known values (freezing and boiling point of water, adjusting for atmospheric pressure as required) should give you a fairly accurate reading.
I assume that the digital probes have fresh batteries in them?

Adam

 

[Gerry]

Good question on calibration. I have two thermometers at work, both are calibrated to UKAS (United Kingdom Accreditation) standards. These are calibrated to national standards in a liquid bath. I have attached a snippet of the calibration certificate below.

Gerry

 

[Spindle]

Hi

Checking thermometers using any method, (boiling point of water - melting point of ice etc.) will only indicate their accuracy at that / those specific temperature(s) - not across the whole range.

To fully check the calibration it would need to be compared to a standard across the range.

Regards Mick

 

[bugbear]

Hi

Checking thermometers using any method, (boiling point of water - melting point of ice etc.) will only indicate their accuracy at that / those specific temperature(s) - not across the whole range.

To fully check the calibration it would need to be compared to a standard across the range.

Regards Mick

If we're being picky, you also need pure water, and to regulate air pressure.



BugBear

 

[mjcann]

Spindle wrote:
Hi

Checking thermometers using any method, (boiling point of water - melting point of ice etc.) will only indicate their accuracy at that / those specific temperature(s) - not across the whole range.

To fully check the calibration it would need to be compared to a standard across the range.

Regards Mick


If we're being picky, you also need pure water, and to regulate air pressure.



BugBear

_________________
Avoid any up and down movements; this is the chief fault of the beginner.

I agree with you both on this, if it was a process controler or similar, but we are talking weather station, probably made in the middle of china for about 20p a dozen. :lol:
martin

 

[Robbo3]

Thank you all.

Perhaps I should say that the digital thermometers are similar to a digital clock & are battery powered. Although classed as outdoor, any sensors are hidden within the electronics & wouldn't take too kindly to being immersed in liquid, either hot or cold.

I think Nev's first post was closest to understanding my question. The point being that it doesn't matter how many thermometers I have, they may all be inaccurate. If only one of them is accurate I have no way of knowing which one it is.

 

[Jacob]

.....
I think Nev's first post was closest to understanding my question. The point being that it doesn't matter how many thermometers I have, they may all be inaccurate. If only one of them is accurate I have no way of knowing which one it is.

Yes you have - the one that's most often closest to the local weather forecast.

 

[mind_the_goat]

I brought an electronic medical thermometer from boots, Took it back because it had us all down as being dead for several hours. Glass types seem hard to find now.

Yes you have - the one that's most often closest to the local weather forecast.

Are you serious ? You expect a forecast that attempts to predict an average temperature over hundreds of square miles for a period of several hours would give a precise measurement of the temperature in one spot in your garden at a specific time of day?

Still think the best thing is to stick your glass one (assuming it has the range) in ice and boiling water, I guess there is a chance the capillary has been made badly but it should be linear otherwise.
It's still hard to measure air temperature without careful placement and a Stevenson screen. The impact of wind and infra red radiation will be different on each device.

 

[Robbo3]

Yes you have - the one that's most often closest to the local weather forecast.

In that case I wouldn't need a thermometer. I can keep dashing in & out to look on either the PC or the TV. :?

Still think the best thing is to stick your glass one

That was my feeling. Although ancient & quite small I have much more confidence in its readings although perhaps wrongly.