Subject: Re: Calibrated Thermometers
>All - many thanks for replies.
>I apparently provoked a bit of a discussion about the theoretical
>temperature of ice water with my question. I reasoned (based on my 8th
>grade physics skills) that ice turning to water would result in the loss of
>a large number of calories at 32 F before the temperature of the surrounding
>water increased from 32F to 32F+ (latent heat of fusion?), so an insulated
>mug of ice water would stay at exactly 32F while there was significant ice
>in the water. I guess I will go looking for a physics book.
>Here's a link to a site about "latent heat":
That may or may not be the best discussion I've ever seen of the subject,
but it certainly *does* detail the significant points.
>Note that what I was thinking about is point "b" on the heat curve at the
>bottom of the page.
>Also note there are some other substances listed with their melting points.
>It appears that most are out of range and of no use in checking calibration
>near 20C except for Ethanoic acid at 17C.
I'll stick with H2O... ;-)
>I like the suggestion of checking against a good digital thermometer but of
>course that begs the question - how do I know the digital thermometer is
>accurate? I need a couple of reference temperature points near the useful
>operating range of the thermometer to calibrate it, and your comments have
>got me thinking more about it.
Unless you can get the digital thermometer calibrated, you are
simply going to trust it!
Digital thermometers may or may not be either more accurate or
more precise than a typical dial thermometer though, so the
trick is to find a "good digital thermometer". Of course, good
enough for darkroom work is not really very good. I'd think the
point would be to simply have one thermometer that is used as
the standard, and is typically stored away where it will not be
damaged. That could also be a good mercury or alcohol
I've set up arrays of a dozen or so digital temperature sensors
and charted the differences between them over a period of days.
It was very interesting, and lead me to the conclusion that I
would need to use three sensors to get sub 1F degree accuracy.
I averaged the three, tossed out the value farthest from the
average of the three, and used the average of the other two
values! It turned out that these sensors would drift back and
forth within just about the limit of their specification (plus
or minus 0.5C), and over a week's time they *all* seemed to
eventually wander near the limits. These were relatively
inexpensive ($2-3 or some such each, Dallas Semiconductor type
DS18B20), and may or may not be indicative of the accuracy that
might be available from various digital thermometers.
>"Floyd L. Davidson"
>> "Pieter Litchfield"
>>>Dumb question time:
>>>I have dial thermometers with a dial that can be rotated for
>>>"calibration" purposes. If I put the thermometer in a glass of
>>>water full of ice cubes, should it not read 32 degees F (or 0
>> Use an insulated container. A foam cup snuggly fitted into a
>> thick coffee cup would do fairly well, for example.
>> The idea is to *isolate* the water from heat loss or gain to the
>> surrounding (warm) air. What heat is picked up necessarily goes to
>> melting ice rather than raising the temperature of the water,
>> and hence the temperature will be constant at the melting point
>> of the ice.
>> It will be very close to 32F for most ice cubes... :-)
>>>Since the scales don't go to boiling, I can't use boiling
>>>water. Are there any other useful points?
>> Nope. Or, at least not using water.
>>>I would think that if there is a way of setting the calibration near 68
>>>degrees F to 75 degreesF this would be better than calibrating at 32
>>>F, so are there any natural processes that happen in this range that could
>>>be the confident basis of a calibration?
>> The thermometers should be linear, so they will almost certainly
>> be accurate at 75 degrees even if they are calibrated at 32
>> degrees. However, that should probably be verified for a new
>> It isn't easy to check them accurately, but if you have more
>> than one (and the more the better) it can be close enough. Put
>> them all into a relatively good size (half gallon or more)
>> container of water at room temperature. Ideally this would be a
>> room where the temperature doesn't change, but it won't really
>> make much difference because the thermometers should all react
>> faster than the water can change temperature. When the water is
>> at room temperature, all of the thermometers should read close
>> to the same. Simple as that.
>> In this case you want the water in the container to lose or gain
>> heat from the air around it, so don't use an insulating
>> container. An aluminum container would be best perhaps, but
>> another metal or glass is okay too. Just don't use a foam
>> insulated jar! The ideal container would probably be a double
>> boiler from the kitchen... with the outer container basically
>> buffering the inner one, which might take several hours to
>> stabilize too.
>> And, to top all of that off... buy any half decent digital
>> thermometer to use as a reference to calibrate all the others
>>>"Nicholas O. Lindan"
>>>> You can always stick it up your a---. Should
>>>> be 98.6 +-.
>> Not necessarily true. That is just the average, but
>> individuals vary greatly.
>>>> Fever thermometers are accurate to +- .2 degrees
>>>> and make good calibration standards: put both
>>>> thermomters in hot water in a styrofoam cup -
>>>> keep the water gently stirred.
>> Do that with a modern digital thermometer though, not
>> one of the traditional glass types that has to be
>> shaken down.
>> Regardless, it won't work well because it is too hard to
>> maintain a cup of water at 98F in a room that is 68F.
>>>> Only color photography is sensitive to
>>>> temperature, and that gets processed at 100
>>>> same as body temperature: they are both organic
>>>> B&W is a +-2 F game, and you only need
>>>> to have a repeatable temperature and the ability
>>>> to read the deviation to +-10%. A grossly
>>>> inaccurate thermometer will do as long as
>>>> it is linear.
>> That is true, and is the *most* significant thing to keep in
>> For example, lacking calibration... always use the *same*
>> thermometer for film developer. All that counts is whether it
>> has precision, not how accurate it is. It can be 5 degrees off,
>> but if it is always within 1/10th of a degree at the same
>> temperature, all is well. (Until a different thermometer is
>>>> Temperature calibration is tough to do with
>>>> great accuracy. A triple-point ice cell is
>>>> the only common temperature reference.
>>>> If you would rather have someone else calibrate your
>>>> Or buy a 'Kodak Color Thermometer' or 'Kodak Process Thermometer' on
>>>> ebay and be as Alfred E. Neumann: "What, me worry?"
>> Floyd L. Davidson
>> Ukpeagvik (Barrow, Alaska) firstname.lastname@example.org
Floyd L. Davidson
Ukpeagvik (Barrow, Alaska) email@example.com
View All Messages in rec.photo.darkroom
Calibrated Thermometers =>Re: Calibrated Thermometers =>Re: Calibrated Thermometers =>Re: Calibrated Thermometers =>Re: Calibrated Thermometers =>Re: Calibrated Thermometers =>
Copyright © 2006 WatermarkFactory.com. All Rights Reserved.