Jump to content
  • Welcome to the eG Forums, a service of the eGullet Society for Culinary Arts & Letters. The Society is a 501(c)3 not-for-profit organization dedicated to the advancement of the culinary arts. These advertising-free forums are provided free of charge through donations from Society members. Anyone may read the forums, but to post you must create a free account.

Reference Thermometers and Calibration Methods for Cooking Sous Vide


PedroG

Recommended Posts

Here is the place to present your reference thermometer you use to calibrate your sous vide equipment.

Your description should include model and manufacturer, URL to the manufacturer's website (if possible, deep link to the model), price in USD (ca.), resolution, accuracy, probe type, your comments and experiences (e.g. result of repeat calibrations, cost of recalibration).

After eventual discussion ("peer review"), the thermometer model may be included in the market overview in the wikiGullet article Reference thermometer.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Greisinger GMH 3710 high precision thermometer

Model: GMH 3710

Manufacturer: Greisinger electronic GmbH, Germany

URL: http://www.greisinger.de/index.php?language=en&task=2&wg=26&artikel=653#112100

Price ca. USD: 370 (incl. GTF 401 1/10 DIN Immersion probe Pt100 for gases and liquids and and 2-point-calibration 0°C/70°C)

Resolution: 0.01°C/F

Accuracy: ±0.03°C

Probe type: Pt100 1/10 DIN Immersion probe with cable and handle

Comments: my GMH 3710 has been in the water for over two years without any problem. Initial calibration stated deviation -0.05°C at -18.00°C, +0.01°C at 5.01°C and +0.05°C at 65.08°C. Recalibration after one year stated +0.004°C at 5.06°C and +0.017°C at 65.15°C. (Calibration was not done at the factory, but at the testo™ calibration lab located near my home).

If I had known there was the GMH 3750 with data logging capability, I would have ordered this one, but www.conrad.ch did not offer it.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

TRACEABLE Control Company

Model: 4000

Manufacturer: Control Company

URL: http://www.control3.com/4000p.htm

Price: $378.88 US

Resolution: 0.001 degrees

Accuracy: ±0.05°C between 0 and 100°C

Range-50.000 to 150.000°C, -58.000 to 302.000°F

Probe type: Various available, see URL.

Traceable to NIST reference, calibration certificate included.

Data logging capability available. Min/max and average readouts. 50 reading memory.

Padded carrying case, AC Adaptor available

Comments: I've been extremely pleased with this unit, and consider something at this level to be indispensable for a serious sous vide enthusiast, and certainly for a professional chef.

My calibration expired 7/30/2010 -- I need to send it in for recalibration

Link to comment
Share on other sites

Geratherm Basal Thermometer (mercury free)

Also referred to as an ovulation thermometer

Manufacturer: Geratherm (Germany)

URL: http://www.geratherm.com/en/

Price: $12.29 from Amazon

Resolution: Markings at 0.1°F, to 0.02°F by visual interpolation

Accuracy: No specific claims, but my unit agreed with my Traceable 4000 reference thermometer as closely as I could read it -- 0.02°F.

Range: 96.1°F to 100.9°F

Probe type: Clinical thermometer. Must be shaken down before use.

Comments: Padded plastic case, with built-in magnifier. By far the most accurate thermometer for the money I have ever found. Well suited as a backup for a calibrated reference thermometer. Buy two at that price, and compare them.

Link to comment
Share on other sites

Not really for 'calibration', more for sanity checking against a validated instrument ...

Superfast Thermapen made by ETI http://thermometer.co.uk/2-thermapen-thermometers.html

UK price ~£50

Resolution 0.1C

Accuracy ±0.4C (from -50 to +200C)

Calibrated against traceable standards.

Certificated calibration error for this thermometer when calibrated: 0.0C at 0C and 100C (but that was almost a year ago)

When I get a big chunk of time for idle play, I might try cooking some eggs at different (indicated) temperatures as a specific test to see how things go.

An ovulation thermometer would provide an additional reference.

Meanwhile, I can factor in that the Thermapen could possibly be half a degree C in error.

My strong suspicion is that most cooking people (even most sv cooking people) would think that because the display says 131F or 54.6C, the bath must be exactly at the indicated temperature ... "Are you trying to tell me its faulty?"

The public's expectation is that the underlying accuracy is better than the precision displayed.

But that ain't necessarily so ...

Edited by dougal (log)

"If you wish to make an apple pie from scratch ... you must first invent the universe." - Carl Sagan

Link to comment
Share on other sites

Geratherm Basal Thermometer (mercury free)

Also referred to as an ovulation thermometer

Manufacturer: Geratherm (Germany)

URL: http://www.geratherm.com/en/

Price: $12.29 from Amazon

Resolution: Markings at 0.1°F, to 0.02°F by visual interpolation

Accuracy: No specific claims, but my unit agreed with my Traceable 4000 reference thermometer as closely as I could read it -- 0.02°F.

Range: 96.1°F to 100.9°F

Probe type: Clinical thermometer. Must be shaken down before use.

Comments: Padded plastic case, with built-in magnifier. By far the most accurate thermometer for the money I have ever found. Well suited as a backup for a calibrated reference thermometer. Buy two at that price, and compare them.

The thermometer needs to be shaken down so that the temperature reading goes back to what it should be, right?

Lot of Amazon reviews showing that the temp reading gets stuck.

Link to comment
Share on other sites

Geratherm Basal Thermometer (mercury free)

...

Probe type: Clinical thermometer. Must be shaken down before use.

...

The thermometer needs to be shaken down so that the temperature reading goes back to what it should be, right?

Lot of Amazon reviews showing that the temp reading gets stuck.

Once upon a time, this was standard for absolutely all clinical thermometers.

The column top stays at the maximum reading.

You need to shake it down to reset it.

Its perfectly possible that there is a significant (younger) population out there that simply don't know how to use such a non-digital instrument.

ADDED - "Shake" isn't quite the whole story. Its more of a wrist-whip that's needed. Hard to describe, easy to demonstrate. No problem if you've ever seen it done, but you could waste a lot of time "shaking" a thermometer to absolutely no effect at all...

Edited by dougal (log)

"If you wish to make an apple pie from scratch ... you must first invent the universe." - Carl Sagan

Link to comment
Share on other sites

Cyclotest Lady Digital thermometer for cycle control

Model: ®Cyclotest Lady

Manufacturer: UEBE Medical GmbH, Germany

URL: www.cyclotest.de/fileadmin/downloads/7_0620_001_A_01.pdf

Price ca.: from € 9 to CHF 30 (may eventually not be easily available in the Fahrenheit world)

Resolution: 0.01°C

Accuracy: ±0.10°C between 35.50°C and 42.00°C

Probe type: maximum thermometer

Min/max function: no

Comments: This was my first reference thermometer before I bought the Greisinger GMH 3710 high precision thermometer. Calibration against the GMH 3710 gave the following results: offset at 33.11°C -0.06°C, at 36.02°C -0.03°C, at 38.72°C -0.02°C, at 41.88°C +0.01°C, i.e. accuracy well within specification.

Inexpensive digital fever thermometers are not automatically inaccurate, but extrapolation from calibration at 37°C to 55°C may be inaccurate, as I have pointed out in my Wikia article on thermometer calibration.

A digital fever thermometer with 0.01°C resolution may be a backup: calibrate against your reference thermometer, and check e.g. every two months. Should there be change of 0.1°C or more, recalibration of the reference thermometer may be due.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Here is a template for convenience: click the quote button, then delete the (quote) including this line and the end (/quote).

Title

Model:

Manufacturer:

URL:

Price ca. USD:

Resolution:

Accuracy: ±

Probe type:

Min/max function: yes/no

Comments:

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Greisinger GTH 175Pt

Model: GTH 175Pt

Manufacturer: Greisinger electronic GmbH, Germany

URL: /www.greisinger.de/index.php?language=en&task=2&wg=27&artikel=37#100390

Price ca. USD: 77

Resolution: 0.1°

Accuracy: ±0.25°C at 55°C

Probe type: Pt1000 immersion, fixed cable

Min/max function: yes

Comments: This would have been my reference thermometer if I had not decided to opt for the GMH 3710 with 0.01°C resolution allowing more sophisticated PID-tuning experiments. Calibration can be adjusted with trim-pot.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Greisinger GTH 175/Pt - K

Model: GTH 175/Pt - K

Manufacturer: Greisinger electronic GmbH, Germany

URL: www.greisinger.de/index.php?language=en&task=2&wg=27&artikel=39#100410

Price ca. USD: 93

Resolution: 0.1°

Accuracy: ±0.25°C at 55°C

Probe type: Pt1000 insertion, fixed cable and handle heat resistant up to 250°C/480°F.

Min/max function: yes

Comments: The same as Greisinger GTH 175/Pt, but with Teflon handle and Teflon cable which are both temperature resistant up to 250°C/480°F, and the probe is insertable, allowing e.g. core temperature measurement in an oven or in a BBQ grill/smoker.

Greisinger is delivering worldwide (except crisis regions like Iran, Irak, Libya etc.).

To see price info on the Greisinger website, registration/login is necessary.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Greisinger GTH 175/Pt - WPT2

Model: GTH 175/Pt - WPT2

Manufacturer: Greisinger electronic GmbH, Germany

URL: www.greisinger.de/index.php?language=en&task=2&wg=27&artikel=543#107785

Price ca. USD: 158

Resolution: 0.1°

Accuracy: ±0.1°C at 55°C

Probe type: Pt1000 insertion, fixed cable

Min/max function: yes

Comments: The same as Greisinger GTH 175/Pt, but with certified 2-point calibration 0°C/70°C.

So far best price tag for ±0.1°C accuracy with calibration certificate.

Beware not to destroy calibration by turning the trim-pot!!

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Anybody aware of a good fast-response sensor that's easily soldered to wires? I want to use something other than an immersion probe for smoke temp.

The Maverick ET-73 has a successor: Maverick ET-732 dual probe remote BBQ thermometer which is said to have better range and out-of-range alarm; currently unavailable at amazon. Yard and Pool sells it at 60$.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

Not really for 'calibration', more for sanity checking against a validated instrument ...

Superfast Thermapen made by ETI http://thermometer.co.uk/2-thermapen-thermometers.html

At the moment, I also use my Thermapen for sanity checking of my immersion circulator. To be on the safe side, I've applied the the indicated offset of -0.5°C (i.e. all temperatures are reported 0.5°C lower than the IC thinks they are), even though this is about the accuracy limit of the Thermapen.

My strong suspicion is that most cooking people (even most sv cooking people) would think that because the display says 131F or 54.6C, the bath must be exactly at the indicated temperature ... "Are you trying to tell me its faulty?"

The public's expectation is that the underlying accuracy is better than the precision displayed.

But that ain't necessarily so ...

I think it's conceptually difficult for most people to read a temperature as being between 54.1 and 54.9°C when the thermometer displays 54.5°C. It certainly is for me. I'd really like to get something like the ETI Reference Thermometer:

ETI Reference Thermometer

Model: Reference Thermometer

Manufacturer: ETI

URL: http://thermometer.co.uk/919-reference-thermometer.html

Price ca. USD: ~315 (excl. VAT)

Resolution: 0.01°C

Accuracy: ±0.05°C at 55°C

Probe type: fixed Pt100 1/10th DIN

Min/max function: no

Comments: A 5-point UKAS calibration certificate is included.

Link to comment
Share on other sites

  • 4 weeks later...

I found this to be an interesting article. I've been reading a bunch of articles, at NIST and on Wikipedia and other places, to determine how temperature scales are currently defined. There is something called International Temperature Scale of 1990 (ITS-90) which is designed to allow a laboratory to independently recreate a temperature scale, which for out uses would range across three standard points, the triple point of Vienna Standard Mean Ocean Water (VSMOW), the melting point of Gallium (29.7646C) and the freezing point of Iridium (156.5985C).

I vaguely recalled that the freezing point of water/boiling point of water were no longer being used, but at this point they are no longer being used to define the size of the Celsius degree, either.

This article was also interesting to me: http://www.nist.gov/customcf/get_pdf.cfm?pub_id=830919 - especially a table on page six which compares the tolerance of several different sorts of measuring devices, and the best case calibration tolerance.

As I read the table, it seems that the "Grade A" PRT has a tolerance of over .1C at freezing, rising to .5C at sous vide temperatures. (Vertical scale is a log scale, that is my best guess). While calibration error is well below that (and reflect repeatability and adjustability, as I read the article) there is also some discussion of error caused by repeated reading of thermistors and such.

All that said, I use a Taylor 9842 that I have checked with a freezing water bath and which was likely within half a degree - the water bath varied from about 31.6 to 32.4 depending on the amount of agitation. I didn't adjust it in the bath, since it was not distilled water. I validated it at sous vide temperatures by cooking an egg and looking at how much of the white and yolk had set, it seems to match what other people get at those temperatures.

This thermometer reads in 0.1 degree F. While I doubt the thermometer is accurate to that degree (as it were) the last digit does let us know what the relative temperature is - if you measure something, and leave the thermometer in, then once the thermometer stabilizes, you can determine if there is a drift in temperature - is something warming or cooling? If the thermometer read in whole degrees only, the change in temperature over time would not be apparent until the reported temperature crossed a degree line. This should allow you to determine change in temperature faster than you otherwise would be able to.

It is true that many people don't understand measurement accuracy, nor do they understand how calculations affect that. I don't believe that most people take physics, and if they do, they don't remember it. Your measurement accuracy can't be increased by performing calculations on it - an example would be that you have a wire that is a mile long and you want to cut it into one foot lengths, how many will you have? The answer is, probably more than 5000, but it is impossible to say for sure. The answer that a lot of people will give is, 5280, but is it likely that a 1 mile measurement is carried out to a 1 foot tolerance? Even if it were, if each piece is cut 1% too long (or there is a 1% cutting loss, actually really low), and everything else is accurate to four decimal places, we get 5227 pieces, plus a fraction. The rule of thumb is that you presume 2 places of accuracy in a measurement, so the fewest you are probably going to get is 5175 or so. Or that was how I was taught in high school physics, anyway. It is probably OK for high school, but people will do a series of calculations and carry every decimal place that their calculator will allow without looking at precision, or where to round, and they will believe that those decimal places are meaningful.

How accurately can you measure when cooking? I'd bet that 1 decimal place would be pushing it for a recipe.


SousVideOrNotSousVide - Seller of fine Artificial Ingredients such as Lactisole through Amazon.Com....

Link to comment
Share on other sites

I agree. Pseudo precision is pointless.

I calibrate my dig. thermometers with boiling distilled water and iced dw. If its accurate in that range isn't that good enough?

Link to comment
Share on other sites

  • 3 weeks later...

I calibrate my dig. thermometers with boiling distilled water and iced dw. If its accurate in that range isn't that good enough?

That is certainly better than not calibrating! If you calibrate at 0C and 100C (or thereabouts) you are trusting that the device is linear at the "important" temperatures around 55C where a degree one way or the other can make a huge difference on pasteurization time. I tend to place health issues high on my list, so even with reasonable confidence about the temperature of my rig, I assume it is running cold when looking at time achieve a specific reduction of pathogens.

Even if you chose to trust the linearity of your thermometer, you also need to be careful when doing the calibration.

Thermoworks has a nice video on making a meaningful ice bath. Many believe that you should have the probe in a steam bath, not in the water itself. Additionally, the boiling point of water varies significantly depending not only on physical altitude, but also normal weather-induced changes in pressure. (I can't comment on the accuracy of the tables provided in the second link.)

Link to comment
Share on other sites

  • 3 months later...

Greisinger GTH 175Pt

Model: GTH 175Pt

Manufacturer: Greisinger electronic GmbH, Germany

URL: /www.greisinger.de/index.php?language=en&task=2&wg=27&artikel=37#100390

Price ca. USD: 77

Resolution: 0.1°

Accuracy: ±0.25°C at 55°C

Probe type: Pt1000 immersion, fixed cable

Min/max function: yes

Comments: This would have been my reference thermometer if I had not decided to opt for the GMH 3710 with 0.01°C resolution allowing more sophisticated PID-tuning experiments. Calibration can be adjusted with trim-pot.

I finally got GTH 175Pt-K (i.e. with the heat-resistant penetration probe) with an ISO calibration by Testo. However, I've got some difficulty reading the calibration certificate. They give separate values for actual deviation at the measurement points, allowed deviation according to the manufacturer's specification and for total uncertainty of measuring. So what's my actual accuracy when using this thermometer? The "total uncertainty of measurement"? So why is that given as an absolute value instead of ± like the allowed deviation?

BTW, the GTH 175/Pt doesn't have a min/max function.

Link to comment
Share on other sites

  • 1 year later...

I've had my Sous Vide Supreme Demi for two years now and suspect it is off by at approximately 1 degree Fahrenheit above its displayed temperature - would love to confirm or rule this out! So here's my possibly stupid question: how exactly do you go about using a thermometer like the Geratherm to measure/calibrate the temperature of your sous vide water baths? Is the process different for a non-circulating sous vide oven versus something like the Polyscience or Sidekick?

Edited by Caruso (log)
Link to comment
Share on other sites

Calibrating at 37°C is better than nothing, but you have no idea how the sensor of the Sous Vide Supreme Demi deviates at different temperatures. See my Wikia article: a sensor that is correct at 55°C may well be 0.5°C off at 37°C (or vice versa) as was the case with the old SousVideMagic 1500B. You might calibrate your bath and one or two other thermometers against a Geratherm at 37°C and compare again at 55°C. If they differ, you know you have to get an ISO- or NIST-calibrated reference thermometer; if they do not differ, you might rely on this calibration. Anyway, with an unstirred bath, you better stay on the safe side with your temperatures, and forget about "natural convection", this happens only during ramping up at full power, but not at steady state.

Peter F. Gruber aka Pedro

eG Ethics Signatory

Link to comment
Share on other sites

  • 5 weeks later...

I have both units. I won't comment on the accuracy as the specs are posted on both sites and I haven't independently verified them.

The Sper unit is a little more ergonomic but the probes it comes with are not useful, nor are any of the accessory probes from Sper Direct useful. It is not backlit as the photo suggests. I've been trying to return one and can't get them to respond to my emails.

With Thermoworks, you get a much better selection of probes that are suited for cooking and the service has been great. You might just give them a call.

Link to comment
Share on other sites

  • 1 year later...

I'm rounding out my new sous vide setup and am on the hunt for a thermocouple for measuring protein temps. I'm under the impression that there are many variables at play when determining the doneness of a protein and the only sure way to confirm that the target temp is reached is to either leave the item in the water bath for longer than necessary or monitor it with a probe.

I am a bit cautious about poking my vacuum sealed item with a sharp probe and potentially damaging the bag. I know that standard practice is to stick a piece of foam tape to the bag before inserting the probe. Does this really work without damaging the seal? Can the probe be inserted at this location multiple times? Alternatively, could I leave this probe inserted for the duration of the cook without worrying that it will find its way loose as the bag bobs freely in the water bath?

Also, I am interested in the ThermoWorks Therma K meter and miniature needle probe (113-181). Is anyone here using this arrangement who can vouch for its suitability for sous vide.

Thanks in advance!

Chris

Link to comment
Share on other sites

×
×
  • Create New...