Anyone know alleged accuracy of this Kodak thermometer?

Paul Verizzo · Jun 19, 2015

I tried searching on the intertubes with both catalog and part number and got no hits. Lots of noise....

It's Kodak part number 491539, catalog number 112 2167. A glass tube, no metal backplate, a bit less than a foot long. Fahrenheit only, black fluid in the glass which has the temperature markings.

The markings allow, at best, a half degree of interpolation. Nevertheless, I'd like to know what Kodak claimed.

Thanks!

ic-racer · Jun 19, 2015

The specifications of your thermometer are as follows:

The Kodak Deluxe Darkroom Thermometer is a 10 inch glass fahrenheit scale thermometer, with an accuracy + 1/2°F at 75 to 68°F, and + 1°F at all other points. This thermometer contains alcohol.

Just for the sake of comparison the coveted Kodak Process Thermometer Type 3 has specifications as follows:

The Fahrenheit scale is accurate to +/-0.25 degree at 59, 68, 77, 86, 104, and 122 F and to +/-0.50 degree over the rest of the scale.

Ken Nadvornick · Jun 19, 2015

I use my Kodak Process Type 3 mercury thermometer as my darkroom standard. I calibrate my Kodak Deluxe alcohol thermometer against the Type 3. That calibration shows that the Deluxe reads exactly 0.5F too high at 68F.

However, I don't have a calibration constant for the Type 3 itself. I simply define it to be correct for my purposes.

For context, in general at 68F a difference of plus or minus one degree equates to approximately half a minute less or more development time. Click here for a handy time/temp adjustment calculator on the digitaltruth website.

Ken

Chan Tran · Jun 19, 2015

I use the Kodak type 3 process thermometer as my standard. I calibrate my many electronic thermometers and temperature controller using the Kodak as standard.

ic-racer · Jun 19, 2015

Ken Nadvornick said:
I use my Kodak Process Type 3 mercury thermometer as my darkroom standard. I calibrate my Kodak Deluxe alcohol thermometer against the Type 3. That calibration shows that the Deluxe reads exactly 0.5F too high at 68F.

However, I don't have a calibration constant for the Type 3 itself. I simply define it to be correct for my purposes.

For context, in general at 68F a difference of plus or minus one degree equates to approximately half a minute less or more development time. Click here for a handy time/temp adjustment calculator on the digitaltruth website.

Ken

Yes, I have both also. Compared to the Type-3, My Kodak Deluxe reads about one degree too high at 75F.

semi-ambivalent · Jun 19, 2015

Ken Nadvornick said:
I simply define it to be correct for my purposes.

Ken

The alpha is right here. Everything else comes after.

s-a

Sirius Glass · Jun 19, 2015

It is accurate enough to product consistent results for black & white and color processing.

Paul Verizzo · Jun 20, 2015

APUGuser19 said:
It's actually a blue fluid. The thermometer comes in a black cardboard tube, and was originally called Kodak Color Thermometer. I use mine all the time. Good enough for me.

Mine looks black, so it's "actually" black. I admit I do not shine a Kleig light through it which might show its blue-ness.

Paul Verizzo · Jun 20, 2015

ic-racer said:
The specifications of your thermometer are as follows:

Just for the sake of comparison the coveted Kodak Process Thermometer Type 3 has specifications as follows:

Exactly what I was looking for, thanks. I presume it's not "+1/2 degree" but +/-. Of course, that's the maximum error, by the same token it could be less or even zero.

To the extent one can interpolate those very close markings, it agrees with my new digital process thermometer/relay thing. Close enough, for sure.

I came across a lot of search results for the Process 3, but I don't think I need that, um, degree of accuracy. As long as a given unit is consistently off, just adjust the processing time.

GRHazelton · Jun 20, 2015

APUGuser19 said:
It's actually a blue fluid. The thermometer comes in a black cardboard tube, and was originally called Kodak Color Thermometer. I use mine all the time. Good enough for me.

I have one of these - somewhere! It worked fine for Ektachrome, back in the day.

ic-racer · Jun 20, 2015

Paul Verizzo said:
Exactly what I was looking for, thanks. I presume it's not "+1/2 degree" but +/-. Of course, that's the maximum error, by the same token it could be less or even zero.

To the extent one can interpolate those very close markings, it agrees with my new digital process thermometer/relay thing. Close enough, for sure.

I came across a lot of search results for the Process 3, but I don't think I need that, um, degree of accuracy. As long as a given unit is consistently off, just adjust the processing time.

For B&W I suspect you will be fine with the one you have. I have a bunch of thermometers because they frequently came with other stuff or I saw them for sale for a buck or so.

Sirius Glass · Jun 20, 2015

Paul Verizzo said:
I tried searching on the intertubes with both catalog and part number and got no hits. Lots of noise....

It's Kodak part number 491539, catalog number 112 2167. A glass tube, no metal backplate, a bit less than a foot long. Fahrenheit only, black fluid in the glass which has the temperature markings.

The markings allow, at best, a half degree of interpolation. Nevertheless, I'd like to know what Kodak claimed.

Thanks!

If you purchase a Jobo CPP2 it will measure the chemical temperatures to 0.1 degree C.

Paul Verizzo · Jun 20, 2015

ic-racer said:
For B&W I suspect you will be fine with the one you have. I have a bunch of thermometers because they frequently came with other stuff or I saw them for sale for a buck or so.

I'm just using it as a reference to various other thermometers, some having nothing to do with photography. Just playing.

It is in agreement to a few tenths of a degree with the digital processing device I bought recently. And the latter has huge red numbers instead of those eye busting little lines.

And with hybrid processing, a bit off is kind of inconsequential, as long as the error is consistent. It can be adjusted for if necessary.

kreeger · Jun 20, 2015

Paul Verizzo said:
I tried searching on the intertubes with both catalog and part number and got no hits. Lots of noise....

It's Kodak part number 491539, catalog number 112 2167. A glass tube, no metal backplate, a bit less than a foot long. Fahrenheit only, black fluid in the glass which has the temperature markings.

The markings allow, at best, a half degree of interpolation. Nevertheless, I'd like to know what Kodak claimed.

Thanks!

Paul, I happen to own your (112 2167) thermometer and the Kodak Process Type 3, as well as two other Kodak thermometers, the Tank & Tray and the basic thermometer.
I lined them all up in this photo at about the same reading so you can see the variances between them.
This shows the 112 thermometer reading 4 degrees lower than the Type 3. In fact it reads lower than either of the least cost units.

The are also lined up least expensive to most expensive at the bottom.

RalphLambrecht · Jun 21, 2015

Paul Verizzo said:
I tried searching on the intertubes with both catalog and part number and got no hits. Lots of noise....

It's Kodak part number 491539, catalog number 112 2167. A glass tube, no metal backplate, a bit less than a foot long. Fahrenheit only, black fluid in the glass which has the temperature markings.

The markings allow, at best, a half degree of interpolation. Nevertheless, I'd like to know what Kodak claimed.

Thanks!

In the days of brick and mortar shops my system of buying an accurate thermometerwas to to a place where several were hanging r laying right next to each other,read them and quickly calculate the average, then find the one reading that average and buy it,works well:sorry,but no can do online

Paul Verizzo · Jun 22, 2015

@kreeger and Ralph: Great comments. In my playing with my thermometers, I've noticed that if I have them all within inches of each other on a table, the discrepancies can by huge on the digital ones. But if I wait some hours or even overnight, that all settle down close to each other. I have no explanation for this.

My thermometers include a $5 stick digital, a $10 large display with hygrometer, a $10 indoor/outdoor, and a $5 tiny dial analog like AC service people use. All of these are from Harbor Freight. Then I have the subject Kodak glass stick and a cheap large round analog. I used to have a "lab" digital but the display went sector bad after not too long....grrrrrrr....and I had a mercury "lab" stick but, boy, did I curse when I broke that! Still have the mercury to turn into the hazmat folks some day.

Oddly, even my new digital processor read a degree different from the Kodak thermometer when first fired up, but after a day, came with a few tenths.

A lot of observations I sure can't logically explain, but as far as developer temperatures, I know I'm "good enough."

phass · Jun 22, 2015

Paul Verizzo said:
I tried searching on the intertubes with both catalog and part number and got no hits. Lots of noise....

It's Kodak part number 491539, catalog number 112 2167. A glass tube, no metal backplate, a bit less than a foot long. Fahrenheit only, black fluid in the glass which has the temperature markings.

The markings allow, at best, a half degree of interpolation. Nevertheless, I'd like to know what Kodak claimed.

Thanks!

You can calibrate your thermometers with very easy procedure. You need to make enough ice cubes from distillated water to fill atleast 1/3 of bucket. When ice is ready ad water and wait until maybe half of ice will melt (water temperature will not increase above 0 C until all ice will melt), than try to test your thermometer. Do not forget to stir your cocktail well, otherwise you may find that all you thermometers off of +4 C (water has highest density at 4C).
Cheers.

Sirius Glass · Jun 22, 2015

phass said:
You can calibrate your thermometers with very easy procedure. You need to make enough ice cubes from distillated water to fill atleast 1/3 of bucket. When ice is ready ad water and wait until maybe half of ice will melt (water temperature will not increase above 0 C until all ice will melt), than try to test your thermometer. Do not forget to stir your cocktail well, otherwise you may find that all you thermometers off of +4 C (water has highest density at 4C).
Cheers.

Great idea. Welcome to APUG.

Maris · Jun 22, 2015

There are many subtle catches with precision liquid-in-glass thermometers. An important one is called "emergent stem error". My darkroom thermometers specify that they be used vertically and immersed 76mm into liquid. Laid horizontally and fully immersed in liquid the reading is different. Emergent stem error is calculated using the temperature difference between the liquid surrounding the thermometer bulb and the air surrounding the graduated stem of the thermometer.

Mercury-in-glass thermometers will read different if they settle by rising or falling. The mercury in the capillary moves in jumps because mercury does not "wet" glass.

I used to do molecular weight determinations in the laboratory by precision thermometry of freezing point depression or boiling point elevation and a good thermometer correctly immersed, with the emergent stem error calculated and compensated, fitted with a high power sliding loupe and a thermometer "vibrator", will give reliable readings to within 1/100 of a degree and in the case of a Beckman Differential thermometer estimates to 1/1000 of a degree can be done.

Paul Verizzo · Jun 23, 2015

phass said:
You can calibrate your thermometers with very easy procedure. You need to make enough ice cubes from distillated water to fill atleast 1/3 of bucket. When ice is ready ad water and wait until maybe half of ice will melt (water temperature will not increase above 0 C until all ice will melt), than try to test your thermometer. Do not forget to stir your cocktail well, otherwise you may find that all you thermometers off of +4 C (water has highest density at 4C).
Cheers.

I thought about doing that, or with boiling water - I'm 5 feet above sea level - but the mechanical thermometers, stick or dial, don't go to those extremes. I doubt if I'd get much better than the +/- half degree anyway.

phass · Jun 24, 2015

Paul Verizzo said:
I I doubt if I'd get much better than the +/- half degree anyway.

The general rule is the accuracy of measurement is 1/2 of scale grid, so if thermometer has 1 degree between the scale marks than accuracy of measurement is +/- 0.5 degree, and if it is 0.5 degree between marks - +/- 0.25.

Cheers.

Sirius Glass · Jun 24, 2015

phass said:
The general rule is the accuracy of measurement is 1/2 of scale grid, so if thermometer has 1 degree between the scale marks than accuracy of measurement is +/- 0.5 degree, and if it is 0.5 degree between marks - +/- 0.25.

Cheers.

That is what I have always heard and read. Digital thermometers are a special case though.

Roger Cole · Jun 25, 2015

phass said:
You can calibrate your thermometers with very easy procedure. You need to make enough ice cubes from distillated water to fill atleast 1/3 of bucket. When ice is ready ad water and wait until maybe half of ice will melt (water temperature will not increase above 0 C until all ice will melt), than try to test your thermometer. Do not forget to stir your cocktail well, otherwise you may find that all you thermometers off of +4 C (water has highest density at 4C).
Cheers.

This may calibrate them at 0C/32F but how many people process at such a cold temperature? You need them calibrated close as possible to the temperatures you use, or a range including those, say 65F-80F for B&W and maybe 95F-105F for color.

Roger Cole · Jun 25, 2015

Paul Verizzo said:
I thought about doing that, or with boiling water - I'm 5 feet above sea level - but the mechanical thermometers, stick or dial, don't go to those extremes. I doubt if I'd get much better than the +/- half degree anyway.

My mechanical dial thermometer vary so much I've come to consider them almost useless. Easy to read but just too variable and inaccurate. They can easily be off 3 degrees F tomorrow after calibrating at the same temperature today (using a liquid in glass - all of mine of that style seem to agree pretty closely so I mostly use my Patterson as it is easiest to read and seems accurate.)

phass · Jun 25, 2015

Roger Cole said:
This may calibrate them at 0C/32F but how many people process at such a cold temperature? You need them calibrated close as possible to the temperatures you use, or a range including those, say 65F-80F for B&W and maybe 95F-105F for color.

Very "interesting" idea. Try to remove the scale from your thermometer and measure the distance between adjacent lines on the bottom of scale and adjacent lines on the top of scale. Hint: the measurements will be the same.

Anyone know alleged accuracy of this Kodak thermometer?

Attachments

Privacy & Transparency

Privacy & Transparency