Meat Thermometer Puzzle

[TromboneAl]

I have a ChefAlarm meat thermometer and a traditional instant read thermometer.

When I calibrate them in a glass of water, both are accurate and agree on temperatures. They read 32 degrees in an ice bath, and 212 in boiling water.

But in use, I get wildly different readings:



In that picture, the traditional probe is correct. The two probes are touching.

Why do you think that is?

My theory is that it's caused by the exact position of the probe, but a 30 degree discrepancy?

 

[Unpaintedhuffhines]

Interesting. I'd say it absolutely has to do with the exact placement of the probe. That cut of meat looks thin (tough to say exactly from that angle), and if it is, your margin of error is quite small. A thicker cut will give you a larger sweet spot.

This also reminds me of the old adage: "A man with one watch always knows what time it is. A man with two watches is never sure..."

 

[braumeister]

I agree that the picture makes it look like the meat is only about half an inch thick, maybe ⅝", so there is no way I would use anything but my Thermapen to check it.

 

[growing_older]

The two probes are touching. Could one of them be more sensitive to the conductivity of the other probe? Do they get the same readings when probing separately, say take a measurement with one then a few seconds later with the other?

 

[NW-Bound]

Due to a typo error, the thread title reads "Meath Thermometer...". At first glance, I thought I saw "Meth", and wondered what T-Al was up to.

Anyway, the thinness of the pork chop (I think that's what it was) caused the error, as these thermometers have relatively large probes with a significant thermal mass. Also one can never be sure if the probe tip has penetrated to the other side with a thin piece of meat.

I usually probe from the edge to be sure that an inch of probe is centered in the meat, even when working with an 1" thick beefsteak. Of course this is easy with a Thermapen, but I often sear steaks on a pan, then put in a broiler to finish and need to use a thermometer with a remote probe that can go in the oven.

 

[gauss]

Could it be that the response time of the two devices is different?

-gauss

 

[ERD50]

...
My theory is that it's caused by the exact position of the probe, but a 30 degree discrepancy?

Yes, easily. I agree with others on this.

In addition, the mechanical probe probably needs to be submerged more in whatever you are measuring. The electronic ones are typically more localized at the tip.

So there are really two 'sweet spots' to deal with - the 'sweet spot' in the meat, which is tiny in a thin cut like that, and the 'sweet spot' of the sensor.

Although I doubt it is a major problem, a two point calibration doesn't tell you all that much about the accuracy in the 150F range. You could have a linearity issue in the middle. A rough check is to put both probes in a hot water bath of ~ 150F, and see if they agree. Yes, they could both be off by the same amount, this isn't a 'traceable' sort of test, but it gives a reasonable confidence factor, especially if you have a third one to check against. Sorry if that was TMI, but metrology/calibration was a major segment of my working life, and it's hard to shake old habits!

-ERD50

 

[eytonxav]

Could it be that the response time of the two devices is different?

-gauss

That would be my guess as well

 

[ERD50]

That would be my guess as well

I assumed that T-Al did wait until both readings settled - but you know what happens when you assume!


-ERD50

 

[NW-Bound]

... and the time for him to take out the camera or the smartphone. I do not think the thermal lag can totally explain the 30F difference; the response of even the slow probes is measured in seconds, not minutes. The exact location of the sensing element inside the thin piece of meat is important.

And then, who knows how big the sensing element of the mechanical dial thermometer is? Is it a rod or coil that runs the length of the probe?

By the way, the Thermapen uses a thermocouple (junction of dissimilar metals), while most of the less expensive probes use a thermistor (a resistor with a negative temperature coefficient). A thermocouple can be made extremely small, hence has a very small thermal mass, but takes more sensitive electronics and also requires an accurate cold junction compensation. The thermocouple probe also allows the Thermapen to measure up to 572F.

Thermapens rule!

PS. See the following illustration from Thermoworks.

For more info, see: http://www.thermoworks.com/products/thermapen/thermapen_sensor.html

 

[eytonxav]

Another point regarding any measuring instrument, unless it has been periodically calibrated against a legitimate standard it can be off.

 

[NW-Bound]

T-Al claims that his thermometers agree at water freezing and boiling temperatures, so I do not think they can differ by 30F at somewhere in between.

In another thread, I reported seeing a 10F difference between two electronic meat thermometers. I checked them by immersion in a glass of boiling water, and recorded the readings through the range as the water was cooling down. The "standard" was an engineering thermocouple probe hooked up to a DVM. I may repeat this experiment with the Thermapen included.

 

[audreyh1]

I agree that the picture makes it look like the meat is only about half an inch thick, maybe ⅝", so there is no way I would use anything but my Thermapen to check it.

+1

The precision of the probe is everything. The temp change from edge to center can be significant in meat.

 

[NW-Bound]

The temp change from edge to center can be significant in meat.

But of course!

Like 30F or more easily.

That's how one can sear the outside, and the center stays pink.

 

[ERD50]

T-Al claims that his thermometers agree at water freezing and boiling temperatures, so I do not think they can differ by 30F at somewhere in between. ...

No, I would not expect that much delta in the mid-range, I was just throwing it in for completeness. It might account for a few degrees though.

Nice diagram of the thermapen. I think the mechanical type like he shows are even more extreme. Those have a bi-metal (or maybe just something sensitive to thermal expansion?) strip that is twisted (like an auger) so it turns the dial as it expands/contracts. That strip is likely an inch long, I seem to recall taking a broken one apart one time.

-ERD50

 

[CaliforniaMan]

My first thought was that the ChefAlarm probably has the sensor at the tip, while the conventional one measures behind the dial, with the rod being a thermal conductor. I would then suspect that the ChefAlarm was a more accurate representation of the meat temperature. T-Al, why did you think that the traditional one was more accurate?

Possibly one way to test this would be to place only the tips into boiling water (with the rest of the rod cooler), vs. placing the entire rod in the water. My guess is that with the entire rod, both would agree, and with only the tips, the ChefAlarm would be higher.

 

[NW-Bound]

Yes, when the thermal sensing element is bigger than the object being measured, well, one ends up measuring partly the air temperature as the probe is not entirely immersed in the object.

Talking about the outside of the meat being much hotter than the interior, when I stuck the meat with the Thermapen, I observed that its reading first jumped up as the tip traversed the outer layer, then settled down to a lower temperature as I put the tip to the center of the cut. And this was a 1"-thick cut, not a roast. That Thermapen's response is that fast.

And as the probe was removed, the reading dropped back to that of the ambient air within a few seconds. Good stuff!

 

[braumeister]

If we keep up our nagging, eventually most FIREes will own a Thermapen. But in the meantime, I can understand the call of frugality.

So I would like to point out that if you go to their website, you can sign up for their email list. I typically get an offer for discounted products at least once, often twice a year.

 

[NW-Bound]

Thermoworks does have less expensive models with very reasonable prices. I am thinking about getting one, just for the heck of it.

 

[audreyh1]

If we keep up our nagging, eventually most FIREes will own a Thermapen. But in the meantime, I can understand the call of frugality.

So I would like to point out that if you go to their website, you can sign up for their email list. I typically get an offer for discounted products at least once, often twice a year.

LOL! I confess I have two. TWO! Earlier this year I picked up a "refurbished/demo" unit of their new backlit Thermapen for my outdoor grilling. A modest discount, but what the heck!

 

[imoldernu]

So, to the deeper question... as in the 24 lb turkey.... What part of the "needle is sensitive? the tip or the average of the overall length of the sensor? exactly what part of the big turkey, or the 8 lb. roast is the thermometer actually sensing?

If the cook book says 160 degrees, and the turkey has been in the oven for three hours, won't the meat nearest the outside be warmer than than that 2 inches into the turkey breast? Conversely, if the piece of grilled steak is 1 inch thick, and the probe is inserted to 3/4 of an inch, what about the difference between that, and the center (1/2 inch)?

A mercury thermometer senses the ambient air temperature. Touching the "bulb" at the bottom will easily change the reading.

While this may seem to be a picky issue, consider that the "Safe" temperature for different kinds of meat generally ranges from 140F to 165F, a fairly narrow range.

http://www.foodsafety.gov/keep/charts/mintemp.html
My old style meat thermometer still costs under $2.

 

[TromboneAl]

It behaved itself for the hamburgers we grilled for lunch.

Yes, I think the problem was due to the thinness of the pork chop, and the positions of the probes.

I did give both probes time to come to a stop.

Thanks for the ideas.

 

[NW-Bound]

Ah hah! It was a pork chop as I thought. I have found in the past I tended to overcook thin cuts, due to wanting to be sure that they got fully cooked for sanitary reasons, and my meat was often dried and tough.

By the way, here's another inexpensive probe from Thermoworks with a very thin tip. And it costs but $24 after shipping. The speed is not as fast as the Thermapen, but the accuracy/speed is still a respectable 1F in 5 secs.

I would still stick the tip into the chop edgewise to be sure that the entire tip is in the center of the piece.

 

[ERD50]

Ah hah! It was a pork chop as I thought. I have found in the past I tended to overcook thin cuts, due to wanting to be sure that they got fully cooked for sanitary reasons, and my meat was often dried and tough.

By the way, here's another inexpensive probe from Thermoworks with a very thin tip. And it costs but $24 after shipping. The speed is not as fast as the Thermapen, but the accuracy/speed is still a respectable 1F in 5 secs.

I would still stick the tip into the chop edgewise to be sure that the entire tip is in the center of the piece.

In earlier posts, some of the Thermapen fans were lauding the thinness of the tip. It was hard to tell from the pictures if it was significantly thinner than the cheaper ones - the CDN one I have was ~ $10 or maybe $15, and the tip is fairly thin ( OK, trip downstairs to my hole gauge says 3/32" tip, for ~ 3/4", then a 5/32" body). The CDN is pretty good for accuracy and speed.

So is there a significant difference between these cheaper Thermoworks models, the Thermapen, and other brands?

-ERD50

 

[NW-Bound]

That CDN you have may not be bad, considering its price. I just looked at it on Amazon, and they said response time of 6-8 seconds, and nothing about the accuracy. On the other hand, the less expensive Thermoworks I cited earlier claims 1F accuracy within 5 secs, and for about 2x the price. The Thermapen on the other hand is 0.7F within 3 secs.

Is the Thermapen worth $96? I think I would be happy with the lesser Thermoworks, if I did not get the Thermapen as a gift.

About the probe dimension, I just took a photo of some in my arsenal as shown below: an old dial mechanical thermometer, a smaller pocket-size dial thermometer, two probes from two different electronic thermometers, and the Thermapen.





And see the close-up photo of the tips below.







I also took out my caliper to measure the diameter of the probes.

1) Big old dial thermometer: 0.250"
2) The two electronic probes: 0.155"
3) Pocket-sized dial thermometer: 0.150"
4) Thermapen: 0.070"

Note that the shank of the Thermapen is 0.130", tapering down to 0.070" at the last 1/2". The minuscule thermocouple junction is at the very end of that pointed tip.