which is more accurate digital thermometer or analog

[Jimbodaman]

In the middle of a brew day and my analog and digital thermometers are off by 5 degrees measuring my mash water. 167 says analog, 162 says digital

 

[BigMack]

Are they handheld? If so just test them in water with a bunch of crushed ice in it and that should give you an idea of which one is correct.

 

[xcdarrel17]

Put them both in ice water and see which one is closest to 32*

 

[broadbill]

Could be one of them is accurate, they could both be off of the true temperature. Are they calibrated?

 

[Jimbodaman]

By the way it's a thermoworks RT600C says it comes factory calibrated

 

[Jimbodaman]

Analog one is a mercury style one that obviously has no calibration. The red just rises up

 

[LandoLincoln]

Boiling water is 212° F, depending on your altitude.

I'd go with calibrating them with boiling water over freezing water, since mash temps are way closer to boiling.

 

[dyqik]

The Thermoworks one in fairly good condition is way more (like odds of at least 100:1) likely to be accurate. An alcohol bulb (analog) one almost certainly won't be, unless it came with a certificate of calibration to a NIST traceable standard.

Neither boiling water or ice water alone is sufficient to show that a bulb thermometer or dial thermometer is accurate at around 150F. Testing both ice and boiling water is better than one test, but really not sufficient. You really need at least three points to show that the calibration is linear in the range you are interested in, and that the top and bottom points are accurate. The closer these points are to the temperature range of interest, the better.

 

[nDub]

Boiling water is 212° F, depending on your altitude.

I'd go with calibrating them with boiling water over freezing water, since mash temps are way closer to boiling.

I do both. I use two cups, one with boiling water and one with crushed ice and water and test in both. Make sure you're testing the thermometers at the same time in the same water.

I'm kind of neurotic about thermometers though. I think I might have five different brands/styles.

 

[Iseneye]

The Thermoworks one in fairly good condition is way more (like odds of at least 100:1) likely to be accurate. An alcohol bulb (analog) one almost certainly won't be, unless it came with a certificate of calibration to a NIST traceable standard.

Neither boiling water or ice water alone is sufficient to show that a bulb thermometer or dial thermometer is accurate at around 150F. Testing both ice and boiling water is better than one test, but really not sufficient. You really need at least three points to show that the calibration is linear in the range you are interested in, and that the top and bottom points are accurate. The closer these points are to the temperature range of interest, the better.

That is all true but let us know how to add a middle calibration point. Ice water and boiling water are easily created by homebrewers and have known temperatures but not one in between.

 

[Jimbodaman]

Well thanks to the great response of all you wonderful HBT users I found my thermoworks to be the most accurate. It read 32.4 F in ice water while the analog read 35-36. That explains why my beers turned out so watery thanks all

 

[dyqik]

That is all true but let us know how to add a middle calibration point. Ice water and boiling water are easily created by homebrewers and have known temperatures but not one in between.

It's not actually that easy to create good ice and boiling water standards either. Boiling water is probably the easiest of the three, but you need to check the weather and altitude to get reliably within 1 F. But, yes, a third point is harder still to do.

One relatively simple method to add a third point is to go outside the range - a completely saturated salt (sodium chloride only, so avoid iodide added or "lo" salts) solution will freeze at 0 F (by the original definition of F), and your freezer should be able to get there. But then your calibration is reliant on points even further from the relevant range. Other stuff you might have about the house with well defined boiling and melting points tend to be things you shouldn't be heating up too much.

I guess body temperature isn't too bad a standard (and it's the other one that was originally used in creating the Farenheit scale) - measured correctly it's unlikely to be off by more than 1 F from 98.6F (98.2F for under the tongue measurement), unless you are noticeably sick, cold, warm, etc.. A cheap fever thermometer from the pharmacy can check that as well. This is probably only a bit worse than a home ice water test unless you are very careful with that - probably slightly worse absolute accuracy, but it's much closer to mash temperatures than ice water.

The easier method is to just buy a NIST traceable calibrated or tested digital thermometer. Or even a simple PT100, which will likely be more accurate than a cheap alcohol bulb, mercury bulb or analog dial thermometer calibrated at home (PT100s are typically ±0.6C at 60C, so just over 1 F, for the lowest B grade sensors. Grade A sensors are ±0.25C, and better accuracies are available).

 

[Iseneye]

It's not actually that easy to create good ice and boiling water standards either. Boiling water is probably the easiest of the three, but you need to check the weather and altitude to get reliably within 1 F. But, yes, a third point is harder still to do.

One relatively simple method to add a third point is to go outside the range - a completely saturated salt (sodium chloride only, so avoid iodide added or "lo" salts) solution will freeze at 0 F (by the original definition of F), and your freezer should be able to get there. But then your calibration is reliant on points even further from the relevant range. Other stuff you might have about the house with well defined boiling and melting points tend to be things you shouldn't be heating up too much.

I guess body temperature isn't too bad a standard (and it's the other one that was originally used in creating the Farenheit scale) - measured correctly it's unlikely to be off by more than 1 F from 98.6F (98.2F for under the tongue measurement), unless you are noticeably sick, cold, warm, etc.. A cheap fever thermometer from the pharmacy can check that as well. This is probably only a bit worse than a home ice water test unless you are very careful with that - probably slightly worse absolute accuracy, but it's much closer to mash temperatures than ice water.

The easier method is to just buy a NIST traceable calibrated or tested digital thermometer. Or even a simple PT100, which will likely be more accurate than a cheap alcohol bulb, mercury bulb or analog dial thermometer calibrated at home (PT100s are typically ±0.6C at 60C, so just over 1 F, for the lowest B grade sensors. Grade A sensors are ±0.25C, and better accuracies are available).

Thanks for that. I won't be doing any of that but good to know.

 

[muels]

It's not actually that easy to create good ice and boiling water standards either. Boiling water is probably the easiest of the three, but you need to check the weather and altitude to get reliably within 1 F. But, yes, a third point is harder still to do.

One relatively simple method to add a third point is to go outside the range - a completely saturated salt (sodium chloride only, so avoid iodide added or "lo" salts) solution will freeze at 0 F (by the original definition of F), and your freezer should be able to get there. But then your calibration is reliant on points even further from the relevant range. Other stuff you might have about the house with well defined boiling and melting points tend to be things you shouldn't be heating up too much.

I guess body temperature isn't too bad a standard (and it's the other one that was originally used in creating the Farenheit scale) - measured correctly it's unlikely to be off by more than 1 F from 98.6F (98.2F for under the tongue measurement), unless you are noticeably sick, cold, warm, etc.. A cheap fever thermometer from the pharmacy can check that as well. This is probably only a bit worse than a home ice water test unless you are very careful with that - probably slightly worse absolute accuracy, but it's much closer to mash temperatures than ice water.

The easier method is to just buy a NIST traceable calibrated or tested digital thermometer. Or even a simple PT100, which will likely be more accurate than a cheap alcohol bulb, mercury bulb or analog dial thermometer calibrated at home (PT100s are typically ±0.6C at 60C, so just over 1 F, for the lowest B grade sensors. Grade A sensors are ±0.25C, and better accuracies are available).

If you're using a pt100, you should have an accurate meter that is calibrated to read it though. Is there another way to read a pt100 sensor besides ohms? The ohm differences on an a385 are pretty tight to use a Home Depot multimeter.

 

[dyqik]

If you're using a pt100, you should have an accurate meter that is calibrated to read it though. Is there another way to read a pt100 sensor besides ohms? The ohm differences on an a385 are pretty tight to use a Home Depot multimeter.

You can't really accurately read a PT100 on the end of a cable with a multimeter, calibrated or otherwise, because of the lead resistance.

A PT100 readout circuit (like those in a $20 PID controller) will put the PT100 in a bridge and use 3 wires to measure it, so that the resistance from the leads can be subtracted.