K thermocouple help

This section helped me,

"Can I use any multimeter for measuring temperature with thermocouples?

The magnitude of the thermoelectric voltage depends on the closed (sensing) end as well as the open (measuring) end of the particular thermocouple alloy leads. Temperature sensing instruments that use thermocouples take into account the temperature of the measuring end to determine the temperature at the sensing end. Most millivoltmeters do not have this capability, nor do they have the ability to do non-linear scaling to convert a millivoltage measurement to a temperature value. It is possible to use lookup tables to correct a particular millivoltage reading and calculate the temperature being sensed. However, the correction value needs to be continuously recalculated, as it is generally not constant over time. Small changes in temperature at the measuring instrument and the sensing end will change the correction value."

But even after correction for reference temperature, the numbers don't seem correct. How do I convert a table with a reference temp of 32*F to a reference temp of 78*F? I doubt that is enough to correct the error I have, but...

Thanks, Mikek ===========================================================

Not trying to be sarcastic, but sometimes the answer is to just buy the solution and stop spending your time:

----- Regards, Carl Ijames

Den torsdag den 10. september 2015 kl. 22.38.37 UTC+2 skrev Carl Ijames:

or spend $6.99 and get something like this:

any kitchen should have one anyway

-Lasse

YES, incorrect! The wires are the same material as the thermocouple right up to the junction. So, in this case, one wire is Chromel, the other is Alumel. These wires should be continued with the same material types all the way to the cold junction. So, the blades in the connector will ALSO be Chromel and Alumel, even the screws in the connectors ought to be made of Chromel and Alumel. The cold junction makes the conversion from the two dissimilar metals at the SAME temperature, so that any additional thermal offset is supposed to cancel out. Typical temperature controllers and readout devices allow the cold junction to be at room temperature, and measure that uncontrolled temp with a calibrated thermistor that is then read and added into a correction algorithm.

Jon

Hi Carl, No problem, I almost clicked on it, but I've went 60 years without a K type thermocouple, it would probably get one use and that's it. I've already thought about using another method, but I still want to understand why I'm having difficulty with the numbers. When I get more time I'll check the 200mv scale on my meter. For now the best references I have are boiling and ice water to see if I can make the numbers work. Am I just stupid about the 32*F to 78*F conversion or is everyone stumped like me? ;-/ Mikek

Not sure that will work for me, the manual says the cord is good to

428*F and the thermometer can be set to 266*F, no mention it it reads higher. Mikek

Great, Thanks, Mikek

If you're willing to do an ice & boiling water test, that will give you the thermocouple's output per 100C - and that's all you need to know for your oven measurement.

NT

hmm, the manual I looked at says 480F/250C that should cover most of the range for a regular oven

-Lasse

The extremely low reading suggest your cold junction is much hotter

The average temperature of those two junctions where the Chromel and Alumel wires transition to copper is where you need to know the exact temperature.

--sp

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8 
Microchip link for 2015 Masters in Phoenix: http://tinyurl.com/l7g2k48

Being picky, the Vtc arrows should be slid over near the reference junction. Or inside, actually. It's incorrect to think of the measurement junction generating a voltage; the voltage happens all along the wires.

This is a box that we sell as an accessory to our t/c acquisition and simulation products.

The aluminum l-bracket on the left forces the thermocouple wires to take a longer path to the barrier strips, to reduce heat conduction from the outside. There's a 100 ohm surface-mount RTD on the bottom of the board. The D25 in the right goes to an ordinary 25-lead copper shielded cable.

Right, through the temperature gradient down the wires.

Looks like it should perform better than most integrated arrangements.

--sp

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8 
Microchip link for 2015 Masters in Phoenix: http://tinyurl.com/l7g2k48

The irony in the accuracy of the thermocouple is that it is entirely dependent on the accuracy of measuring/setting the temperature of the cold junction. In anything other than a lab setting that means you have to have some other technique of measuring the cold junction temperature that could potentially be applied to measuring the temperature of interest directly if the range is not too extreme.

Glass encapsulated thermistors anyone?

--

Rick

Yes, we seem to be able to measure or simulate to +-0.2C or better. Delta-sigma ADCs have made this fairly easy, both for acquiring the t/c voltage and for measuring the RTD.

Standard thermocouples are spec'd for numbers like +-2 C, but in real life they seem to be much better.

Oh, can I jibe on that? So (on a different forum) I was talking about using heat conduction in a silicon wafer to determine the sign of the charge carrier's. The typical trick is to heat one side of the Si wafer with a soldering iron, and use a DMM to measure the voltage between the hot spot and some cold edge. (one DMM lead placed on the soldering iron is typical... though I've also done it with a heat gun a piece of foam to block the heat from one side.)

So I thought, huh, maybe I can do the same thing with a metal. A piece of copper clad didn't show much of an effect. But a piece of bus wire (22AWG) worked fine. With my 800F weller tip I measured a voltage ~100-200 uV. But the dang sign was wrong! The hot end became negative! WTF. (Maybe someone else can confirm it..?) (maybe I should try some other metals.)

George H.

Only if people insist on using T/Cs where they are inappopriate.

respond really fast, withstand dunking into molten aluminum, or work at moderate cryogenic temperatures such as 77K there are few competitors.

Trying to measure near room temperature? No great demands on response time? Use an RTD or semiconductor sensor (or a thermistor, if you must).

If you really don't like cold-junction compensation you can always use type B- it's so nonlinear around room temperature that it doesn't really need compensation- in fact it's not even strictly monotonic. Of course it's only appropriate for very high temperatures anyway.

--sp

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8 
Microchip link for 2015 Masters in Phoenix: http://tinyurl.com/l7g2k48

I'm not following. Wouldn't being very *linear* mean you don't need compensation? Being non-linear makes compensation very important as well as maintaining the same temperature at both cold junctions.

--

Rick

The idea that you need to consider the thermal gradient along the wire may be technically correct, but of little practical value. Any time you attempt to measure it by contact with another material, you are back to having a "cold" junction at the temperature of the wire which totally swamps out the the effect you are trying to measure.

You would need to measure the voltage by a non-contact method. Even then to what end would this serve? All the data involved in measuring temperatures with thermocouples already includes this voltage.

--

Rick

Follow-up.

1.) If you try this at home be careful, I melted a nice 22Awg size slot into my pomona grabber.

2.) I was scratching my head trying to find a different wire, Ahh, I've got constantan. (for TC's) And that also shows a thermal voltage. And it has the right sign! The hot end becomes positive.

What's up with copper? I do recall reading that it has some strange thermo-electric properties, I guess I just never bumped into them before.

George H.

Hmm, A technique mistake on my part. (It wouldn't be the first time.) (I'm going to have to think more.) It works great with silicon.

I was thinking it would be a quick and dirty way to get the sign of the charge carrier, without having to do the hall effect.

George H.