What is the most cost-effective but accurate way to measure sous vide bath temperature?

[Caruso]

I'll hopefully be receiving an Anova Precision Cooker sometime in the next month, and after seeing all the complaints about temperature accuracy issues on their customer forums, now want a reliable and accurate way of checking its temperature accuracy for myself. I see a lot of people using the thermapen - is that the consensus choice? And once you have one, are you just supposed to have faith that it is calibrated correctly? I've read that temp inaccuracy is non-linear, so that measuring accuracy at the temp of ice water and/or boiling water won't necessarily tell you anything about accuracy at sous vide temps.

[AlaMoi]

ThermaPens are good equipment.  their forte is "instant" - soooooooooooooooo, could be overkill for something you're looking to cook in for hours.  but, what the heck.

 

here's the problem:  situations which require really good control have their instruments calibrated (at least) annually.

anything you buy comes / may come "calibrated" - but things change with time....electronics are horrifically notorious for that - I've got some glass&mercury lab thermometers that have not been calibrated in ten years, but in the preceding 20 years were never to be found "out of calibration" so I'm thinking they're likely still 'ok'

 

if you are looking at tenths of a degree, boiling / freezing water is iffy anyway.  atmospheric pressure, dissolved solids, etc.

[rob1234]

I'm also interested in how you would measure the bath temp accuracy. Specifically where in the bath you would measure (in the middle of the bath or near where the circulator is spitting the water out) and what volume of water you should use (if that matters)?

 

I have a polyscience professional and use it in the cambro container polyscience sells. I checked the accuracy in the middle of the bath using a thermapen, a manual mercury lab thermometer, and a digital lab thermometer that was recently calibrated to be accurate to within 0.1 at 35C and 0.6 at 95C.

 

All 3 thermometers showed the bath temp to be off by about 1C compared to set temp. I sent it back to polyscience for calibration and they said it was accurate as is when checking with their much more expensive and accurate thermometers.

 

So either all 3 of my thermometers are out by a degree or I'm measuring wrong.

[gfweb]

You might consider just how precise SV needs to be.  I know many of us are engineer types who compulse over this sort of thing, but 0.5 degree one way or another probably doesn't even make a difference with soft cooked eggs or delicate fish.

[Caruso]

I agree tenths of a degree are not important for most sous vide applications. I'd just like a way to make sure mine is not off by 2 degrees when aiming for perfect medium rare on an expensive cut of meat. I'm guessing 5 or more degrees and it will be obvious there's a problem even without a thermometer, but 2 degrees might not be as obvious, while still compromising what you're trying to achieve.

Edited November 21, 2014 by Caruso (log)

[btbyrd]

Dave Arnold used to suggest that you put the circulator in a really icy water bath (more ice than water) and set the temperature below 0C (so the heater doesn't kick on) and let it circulate for a while until the temperature readout stabilizes.

[nickrey]

I agree with gfweb's comment above. Moreover, you'll find variances amongst many writers on what temperature defines a "perfect" medium-rare steak. As you cook, you'll get to know what your preference is with your particular equipment. While I acknowledge that I have a fully calibrated reference thermometer, I rarely use it. Instead I prefer to use temperatures that I know work with the sous vide machine that I am using at the time.

[IndyRob]

 

So either all 3 of my thermometers are out by a degree or I'm measuring wrong.

 

I think there's a saying that goes something like "A man with one watch always knows what time it is.  A man with two never does."

 

I'm curious as to the perceived temperature differences reported on Anova's forums.  I've used a Presto Kitchen Kettle to good effect for sous vide in the past.  It does involve working through an equalization phase to find the correct setting for the desired temp, but then it could hold a temp within 1 degree F for an extended period (and that's about 1/2 a degree C).  That was measured by a Thermapen.

 

IMHO, circulation in a home application is overrated.  I've never measured any more than one degree of variation in my crude setup.

 

But in any event, you have to cook to the equipment you have.  Home ovens vary massively, but we can still figure out the best way to bake cookies in our own ovens.

 

And even if we could control temperatures completely, that still does not account for variations in product and preference.  And I think this sort of brings on the sense of derision for sous vide or modernist cuisine in general.  The sense that it can all be reduced to numbers.

 

That ain't so.  It only narrows the window a bit.

 

Just cook, man.

Edited November 22, 2014 by IndyRob (log)

[Caruso]

I'm surprised at some of the responses here that seem to imply that accurate temperature measurement is just not that big a deal. Of course after trial and error I could very likely figure out how to get good results with an out of whack immersion cooker (assuming it was at least consistent), but wouldn't it be better to ensure that it is accurate by testing and calibrating it with a trusted thermometer? Without that, you can't even say what temperature you like for a "perfect" medium rare because frankly you can't accurately say what temperature your circulator is setting the water at.

 

Generally speaking, I'm all for the "Just cook, man" approach, but for certain sous vide recipes, I think you need a more exact approach. For instance, there are times when you are aiming for minimum doneness for safety reasons, and two degrees can matter a lot then. If at all possible, I'd like accurate temperature control on my immersion cooker when cooking meat that way.

 

[AlaMoi]

there is no shame in getting one's thermometers calibrated.

 

ThermoWorks offers a calibration service; there are many many other labs that offer calibration services.  not all to many cooks do that.

for $200 you can buy a ueber high accuracy ThermoPen (+/- 0.04'C) and have it calibrated regularly and then you'll know.

 

as has been noted in this and a few trillion other threads, not every cc of water in a circulating immersion bath is _exactly_ the same temperature.

 

so first one has to look at the advertised accuracy of the "thermostatic" control in the bath - if the accuracy is +/- 1F degree, add in the physically not possible to eliminate volume variations, there's no sense in looking at checking for +/- half a degree.

 

a decent 2nd thermometer only assures you can check the sous vide equipment and have some confidence it is working properly.  if there was a 2 degree difference last week, and this week there's a 5 degree difference, well - then you know something is likely 'off'

 

a very valid point presented above:

if you set up a poll asking for the perfect temperature - of any verbal description ie rare thru charcoal - you will not get 100% agreement.  especially from people using sous vide because they are going by what their gear reads, not a calibrated thermometer....

 

if your bath reads some number different from a good thermometer, and that difference is stable and repeatable, the other point made above is: go with experience.  if the bath produces the result you want when set to X, then set it to X again.

 

taste&preference is a funny thing - assuming somebody somewhere has a perfect thermometer/temperature control and writes a recipe to sous vide Z to temp Y - and assuming you have a perfect thermometer/temperature control to duplicate that, you may still not care for the result.

 

yes, it needs to be close.  after that it's all experience and preference

[Shalmanese]

If you don't own a high quality thermometer, simply do the following steps

1) Calibrate at 0C in a circulating ice bath

2) Put in a pot of water on the stove and bring to a rolling boil to calibrate at 100C

3) Put in an egg at 62C in for exactly 75 minutes and calibrate using the Khymos charts: http://blog.khymos.org/2009/04/09/towards-the-perfect-soft-boiled-egg/

4) Plot the 3 lines on a graph, draw a straight line through them. This is the temperature response curve of your thermocouple. If the 3 points aren't close to being on a line, then something went terribly wrong somewhere.

If you perform all 4 steps, you'll have a very good idea how your sensor is performing across a range of temperatures and can calibrate accordingly. Typically, you want to calibrate so that it's accurate around 65C, that way, so long as your thermocouple isn't off my more than 1C/20C, it will be at most a degree off when cooking seafood at 45C and cooking vegetables at 85C (Apart from eggs, a difference of 1C is rarely noticeable).

Edited November 22, 2014 by Shalmanese (log)

[Shalmanese]

For example, let's say you receive an egregiously bad machine that reads:

1) 5C in an ice bath

2) 103C at a rolling boil (PS: forgot to mention this assumes you're at sea level. If you're at an elevation, look up the water boiling temperature at your location)

3) Your egg looks most like the 65.6C egg in the picture

This means your sensor is off by 5 degrees at 0C and -0.02 degrees for every degree rise. This means you should adjust the thermometer up 3.7 degrees so that it's accurate at 65C and you will be 0.4 degrees too warm at 45C and 0.4 degrees too cold at 65C (you can manually compensate for this if you're really anal).

[AlaMoi]

....Put in an egg at 62C in for exactly 75 minutes and calibrate using the

 

if you take the egg out of the freezer, does not work

if you take the egg off the counter, does not work

if you take the egg out of some part of a refrigerator at some unknown temperature, doesn't work.

 

science is based on science, not eggs.

[gfweb]

I'm surprised at some of the responses here that seem to imply that accurate temperature measurement is just not that big a deal. Of course after trial and error I could very likely figure out how to get good results with an out of whack immersion cooker (assuming it was at least consistent), but wouldn't it be better to ensure that it is accurate by testing and calibrating it with a trusted thermometer? Without that, you can't even say what temperature you like for a "perfect" medium rare because frankly you can't accurately say what temperature your circulator is setting the water at.

 

Generally speaking, I'm all for the "Just cook, man" approach, but for certain sous vide recipes, I think you need a more exact approach. For instance, there are times when you are aiming for minimum doneness for safety reasons, and two degrees can matter a lot then. If at all possible, I'd like accurate temperature control on my immersion cooker when cooking meat that way.

Sure, but if you are talking about a 0.5 degree error it won't have a detectable effect on your product. 2 degree error? Maybe not even then in most dishes.

[IndyRob]

I'm surprised at some of the responses here that seem to imply that accurate temperature measurement is just not that big a deal.

 

As an extremely frugal person who invested $80 in a Thermapen because I do think that temp is that important, I also realize that it's just one measurement.  It's one thing to have accurate readings, but it's quite another to become a slave to them.  Time and temp is simply not enough.  Especially if you're measuring a bath and not the product.

[PedroG]

If you just do short time cooking and don't need to pasteurize, the "just cook" trial and error approach may be OK. For long time cooking your equipment should be calibrated to make sure you are sufficiently above the "danger zone", and for pasteurizing as well calibration is necessary. See Temperature accuracy and stability in the wikiGullet Sous Vide page and Importance of temperature control on pasteurizing times.

See also the wikiGullet page Reference Thermometers

Interpolating to 55°C from a 0°C and 100°C calibration may be unreliable, see my calibration of a testo kitchen thermometer against a Greisinger GMH3710 high precision thermometer.

[weedy]

I checked my Anova against y Thermapen and they agreed within .02 F

 

that was close enough for me.

 

had they not agreed, then I suppose I would have had a dilemma as to whom to trust more

 

 

my old (now unused) SVS OTOH was never within 1.5 degrees of the Thermapen

[gfweb]

I think SVS needs a circulator to get that kind of accuracy

[Shalmanese]

....Put in an egg at 62C in for exactly 75 minutes and calibrate using the

 

if you take the egg out of the freezer, does not work

if you take the egg off the counter, does not work

if you take the egg out of some part of a refrigerator at some unknown temperature, doesn't work.

 

science is based on science, not eggs.

The egg equilibrates to the water temperature at the ~45 minute mark, regardless of starting temperature. 75 minutes is just to bring it in line with the Khymos chart. There are slight texture changes that vary by cook time but they're on the order of several hours.

[Simon Lewinson]

The simplest way to get an accurate temperature check near the sous vide temperature range is to use a cheap medical thermometer. They are required to be quite accurate and just about every house has one. I use one to check my sous vide rig at 40 degrees C. It is also independent of altitude and water purity issues.

Simon

[rotuts]

I have found that a Thermapen is a wise long term investment, esp. if you can hold out for a sale.

 

sales are common.

 

I have several others Ive accumulated over the years, but I now always reach for the TP.

 

BTW  the 'medical' grade thermometer for a single temp check at 40 C is a very good one.

 

http://www.amazon.com/Geratherm-20030-Thermometer-Mercury-Free-Oral/dp/B0013NE0RC/ref=sr_1_9/175-4670372-5171351?ie=UTF8&qid=1416759270&sr=8-9&keywords=clinical+thermometer

 

5 bucks or so for an analog ( scale ) one or so.

Edited November 23, 2014 by rotuts (log)

[haresfur]

I'm pretty sure that thermocouples like the thermapen probe are very linear in their temperature response.  I checked my thermocouple in ice water and in boiling water and found it was very close to 0 and 100 degrees so I'm confident it is accurate in the 60 degree range.  You might have to correct for elevation, depending on where you live.

 

My SV controller is off by about 1.5 degrees so I will adjust to match the thermocouple now that I have found the instruction book.

[nickrey]

If your rig is stable and if it is close enough to temperature, temperature doesn't matter particularly. Sous vide cooking is not rocket science nor does it depend on extremely fine differentiations in temperature. However, as PedroG said, if you want to surf the edge you'd better have an accurate thermometer.

 

If you're worried, get a reference thermometer and get it calibrated annually. If the reference thermometer is too expensive, perhaps the issue isn't as important as it may seem.

[haresfur]

Nick, have you noticed any drift in any of your equipment calibration vs the reference thermometer?

[PedroG]

I did experience sensor drift once or twice with my SousVideMagic.The actual sensor has been in the water 365 d/y for several years without any drift, but I regularly check it against my Greisinger GMH3710 high precision thermometer.