Thermocouple Polynomial Equation

I thought, could be wrong but hey, that any thermocouple of a certain type would follow a strict voltage pattern over temperature difference. That means that each K-type probe should result in the same graph shape, but with slightly different set-off point. You could test this theory by making the graph for a few of them given the data you find and compare the graph shapes. If I am right they would fit exactly over each other and that would prove my theory. If they do not fit at all you either have made a mistake or I am wrong.

If I am right all you need to do, after you have proved I am right, to take the means averages of all data you find and produce your own graph.

Good luck and please register and follow up on this thread, we love to hear back from you.

_{even if I am wrong,}

Thanks for registering and for coming back. Much appreciated.

I will look at it I promise. I already have had too many beers and am in my relaxed mode right now so any conclusions I make would have to be rendered void and useless by alcohol.

One tip I can give you right now tho is that if you have your linked url or ftp or http site in your "copy" button, you can highlight the text and click on the globe with the chain link picture to insert a hyper link. That way people will be able to just click your linked text instead of having to copy and paste in their browser. It makes for faster use of the forum so it is only to help us lazy old tech heads.

haha. K done deal the link has been fixed. Not to rush you but i have been lazy and my assignment is due tomorrow so if you have time tomorrow morning ish to look at it that would be awesome.

Thanks a ton,

Mike

Whoaa, no pressure man are you crazy?

I have had a fleeting glance at it but it is too demanding for now and almost bed time here. I have a job to go to in the morning early so this is it am afraid.

I have however found 2 links to graphs that seem to agree with eachother on the Ktype couples anyway.

http://users.ticnet.com/mikefirth/thermocp.htm

http://www.picotech.com/applications/thermocouple.html

Hope this helps you in understanding your own work sheet. I am sorry but it looks to me you have all you need and I am sure with a clear head I would have been able to help you along further but......

Hope you have better luck with the next question, sorry

Thanks for your effort and input i appreciate your time. Ya i should be able to figure it out. I only left it this long because i play for a school team and we were playing championships this weekend.

Seems like a pretty helpful board though i am glad i found it during my search.

This is not in my assignment but from your second link it appears that the graph has more linearity above 0°C than below it. is this true?

It does seem to be that way although it does not go completely linear and will retain a little of the hyperbolic shape it shows below zero.

I am slightly colour blind and have difficulty telling exactly which one is K type in this graph but they all seem to do the same thing.

Are your graphs approaching these shapes?

My Range for my graph is 0 - 100 C in type K of course and yes, it is showing very linear, not completely but close.

So you are done on this than?

ya i just completed it, it was due today.

Glad to see you signed up Wanny. Temperature is my life. In the old days, we used a galvanometer type instrument and read the output of TC's in millivolts. We had a book to convert the MV's to temp, then used a glass thermometer to compensate for room temp. Of course, I've not done it this way in years, everything is automatic now. I may, in fact, still have one of those MV to temp books in my desk at work. Somewhere, those tables may still be avalable online. I'm home now, but if you need any help tomorrow morning, drop me a line

Awesome thanks for replying. My instructor, however, wants the temperature to voltage polynomial equation. So we insert a knowen temp into the polynomial equation and get the mV output.

I think we have use the mV to temp in our lab where we would take a voltage from the thermocouple and look up the temperature from that then take the ref. temp of our fluke 744 to compensate for room temp.

Here is a link to a type K thermocouple reference table.

http://www.omega.com/temperature/Z/pdf/z204-206.pdf

Ya those are the kind i have used. Thanks for the link.

ok i think this is what i need. It is from a reliable source. The only problem... i don't know how to use that equation.

http://srdata.nist.gov/its90/type_k/kcoefficients.html

Expression for type K thermocouple for 0 to 100 deg C

where e= emf in mV and t= temperature in deg C

e = 0.04 t

There, you got someone to do your homework, now you can go play.

I am not looking for some one to do my HW just some confirmation on a proper equation.

I know that this graph is not linear so that equation is not what i am looking for, it may be the short way. Click on the link in my last post to see the equation to be used.

Wanny,

T₉₀ in that equation from NIST is your temperature. That equation is very similar to the second form of equation in your earlier post from the instrument company. I suspect the second term in it cuts the error to a lower amount. The big Greek letter ∑ means "sum of the terms for each i with i ranging from 0 (the value given below the ∑) to n (the value given above the ∑). The table of coefficients tells you how big n can be. Of course, this sort of table is based on original research using other methods of measuring temperature. Thus the thermocouple is a secondary standard which has been calibrated against a primary standard.

Regards--JMM

Thanks JMM, I think i know exactly what you mean. The second part of the term.... is that just added on at the end of the entire thing? I did all the calculations with it and they come out very similar to the Temp vs Voltage tables so i hope that is a good sign.

Thanks very much for your reply JMM

Mike

All TC's have an error, and type K is not the most accurate. I buy 12,000 ft of 20 ga, type K, and 8000 ft of type N every month. We compare it to a type S standard ( platinum , platinum 10% rhodium ). So, really, your only as good as your standard.

Wow, i have not been in the field yet but that sounds like alot. Temperature is your life!

how do you mean you compare it to a type S standard?

Wanny,

A partial answer--most types of thermocouple junctions are likely to be damaged or degraded by oxidation or similar processes, and are also subject to problems with manufacture or quality control. The Platinum/Platinum-10%Rhodium junction is pretty immune to chemical attack and oxidation, and more stable. It also has a wider temperature range than many others. Therefore it has been studied more and is known to give more reliable measurements of temperature. For these reasons it would be a standard against which other types are calibrated.

Of course, it is also more expensive, so other types are often used instead. Your questions and responses show the thinking that can make a good scientist or engineer. If what you learn is: how to ask questions, where to find answers, and how to look for assumptions or errors in the information you are given, then you are well on your way to being "educated".

--JMM

I see, i didn't realize that he meant using it to calibrate others with it. thanks for the info on the Type S, i'll be sure to keep that type in mind. Does the type S have and short falls?

Looks like you're already helped, but here's a link to all types of tables

http://www.instrumentation-central.com/pages/thermocouple_reference_table.htm

Hi there Mike
Each type of thermocouple has a unique temperature-voltage relationship depending on what metals have been used for the junctions. The voltages generated are not linear. Look-up tables are available for each type. The most common thermocouples are a type K which uses an iron-constantin junction for the low temperature ranges and a type J which is chromal-alumal and can take you up to 1000 degrees C.

Google the company WIKA instruments and drop them a request. They are normally very helpful.

Kind regards
Flicka

Close, but you have it mixed up. Type K is chromal-alumel, and type J is iron-Constantin. Nobody uses Type J anymore. The new standard is type N, a nickle alloy, very similar to type K, but with a higher temp. range. We use type S standards, and calibrate all the other TC's to that. We have a small furnace, and run test TC's from about 800F to 2200F, every 100F, and then chart the errors. The only problem with type S is that they are hugely expensive. Our large high temp. furnaces can only use type S, for control, high limit, and recording. Platinum prices are off the charts. I just bought 1200 inches of reference grade platinum, and similar platinum/10% rhodium, it was about $22000!

hello, here a good link to Omega.com online book called Transactions it has alot of

info on temp measurments.

Omega

Good luck
-Charles

I have been using the formula shown below for many years without any problems. The polynomial below describes the characteristics of the thermocouple output voltage for the range of 0°C to 500°C. This is only part of a much larger equation which describes the thermocouple output voltage for the range of -200°C to 1370°C

T= v * ( (2.51E-2) + v * ( (7.86E-8) + v * ( (-2.50E-10) + v * ( (8.32E-14) + v * ( (-1.23E-17))))))

Don't forget about CJC if you are making any measurements.

Good luck.

Of course my previous response was for a K-type only.

Good Luck

He already knew that, it is mentioned in page 7 of his link in his 2nd response at the top.

His table actually states all values for the equasions for most TC as listed at the top of the table.

case491,

As I mentioned in my first response "the equation works for me, and has for many years". Therefore, if the calculated response (using the standard equation) is not what the 'Young Fella' expects, then an error is being introduced somewhere. I also added the Cold Junction Compensation remark just in case the 'Young Fella' was making measurements to confirm the calculations (another source of error if not conducted properly).

Good Luck

My instructor wants us to use the full eqution and all the coefficients (9 for the Type K). Thanks for the help though HVAC good to know there is an easier way to do it.

One thing I don't understand dear Mike?

What are you actually asking us? You seem to hava answered your own question in your first reply to my first post. I was not up to much last night with the beer and all that but now I see that you could have calculated all you needed to know right there from that table of data.

Am I missing something here?

So what i really needed for this assignment to get going was the full polynomial equation. That first link had a polynomial equation but it's coefficients were different than other i have found, i didn't feel that the site was very reliable. I think i have found what i was looking for in the end at this site.

http://srdata.nist.gov/its90/type_k/kcoefficients.html

Jmueller helped me understand the equation as well, so i think i now have it figured out.

Thank you to every one for there responses it was very helpful and a very interesting site. When i get my mark i'll let you guys know.

Thanks again,

Mike

I understand this site agrees with your nature, it does all the calculations for you.

One note of warning however is that these coefficients do not agree with any of the others stated in any of the links and also not with HVAC's earlier post.

I would therefor go with the safety of numbers and choose to believe the most similar values rather than the lazy man's way out and go with this site.

Make calculations your friend and the sums will end up telling you what you need to know. What do you think?

HVAC's is the inverse of what i am looking for. I have been looking for a temperature to voltage conversion. HVAC's is Voltage to temperature which have different coefficients.

I don't get how the one i chose has done the calculations for me? It has given me an equation and the coefficients, or am i miss reading you?

Fill in the temperature range in the box above the lists and the table changes for all the values you have just ranged.

Don't tell me you did not know that

Haha ya i did, but i had to make a complete excel spread sheet with my equation doing the work.

Ya this one can sleep. Thanks for all your help!

Mike

you can refer to IS 2054 for temperature vs. milli volt chart.