No video
EEVblog
Everything you need to know about how Thermocouples work.
K type thermocouples, the Seebeck effect, the Seebeck coefficient, and cold junction compensation.
Along with some practical measurements with a multimeter to demonstrate the effect.
Seebeck effect on a single conductor:
www.dataforth.c...
NIST Tables:
srdata.nist.gov...
Forum Topic: www.eevblog.com...
EEVblog Main Web Site:
www.eevblog.com
EEVblog Amazon Store:
astore.amazon.c...
Donations:
www.eevblog.com...
Projects:
www.eevblog.com...
Electronics Info Wiki:
www.eevblog.com...
Пікірлер: 193
Hi Dave - very helpful. BTW you misunderstood the NIST table. Everything in the whole table is referenced to 0 degrees C. The 0-10 columns just give you the last digit of the temp in degrees C. For example if you want to know the mV value corresponding to 1005 degrees, go to the 1000 row and look under column 5.
An interesting thing about thermocouples that many may not realize is that although they generate a relatively small voltage, they can generate a high current if their load resistance is low, and the temperature differential is high. The current can be more than half an amp, maybe up to an amp or more. An example of this is a furnace that uses a thermocouple to monitor the gas pilot flame. In this application, it must generate enough current to hold a solenoid open to allow gas flow.
@ndrsg3013
5 жыл бұрын
We use them and change them very often at work (for similar purposes you mentioned)...
@sbybill3271
4 жыл бұрын
@tzampini What sort of temperature difference would produce high enough current?
@sylviam6535
2 жыл бұрын
Interesting. I thought the thermocouple current controlled a larger current via a transistor/relay. I didn’t realize it actually kept the solenoid open by itself.
@tzampini
2 жыл бұрын
@@sylviam6535 The reason they don't amplify the current with a transistor/relay is because this would require power. In the event of a power outage, this safety system would not work. By having the thermocouple current drive the gas valve solenoid directly results in a simple fail-safe system that doesn't rely on house power to operate.
@brynnrogers5081
Жыл бұрын
ThermoPILES are used in that case, and they have like 11 or 22 thermocouples in series/parrellel wiring. The hot end sticks in the flame, and the cold end is 2 inches away near where the leads come out.
Hi The bit about the compensation table is wrong. The scale along the bottom is just to provide the 1 Deg. steps. The vertical scale goes up in 10 deg. steps so you go the the 10 deg step that is lower than the temperature that your cold junction is at and the use the horizontal table for the 1deg finer steps. OK ?
@AdamSkubel
10 жыл бұрын
That makes more sense, thanks for pointing it out.
@lollandster
9 жыл бұрын
I noticed that too. You have to be careful when listening to self-proclaimed experts, always check the source.
@TheOysterjam
8 жыл бұрын
+lollandster ya, i sat there for a while trying to figure out why as the 2 temps come closer together, the volt difference diverged. glad to have this sanity check right below
@blahbleh5671
6 жыл бұрын
I thought so!!
@BYFav
3 жыл бұрын
I think so too. But what's the intention to go from 0 to 10 horizontally? I expect 0 to 9. Anyway, it's 1° step since the 10 values are equal to the next lines 0
Amazing, I simply became captivated when you performed your Seebeck experiment with the looped wire. These kinds of things are really what sets apart "just a tutorial" from "an explanation"
Great Tutorial. Very good explained. The great thing about you dude is you opened up your instrument for viewers to increase our knowledge. That's the great Seebeck Effect we got :-) . Thanku !
The Seebeck effect does apply to a single wire, and that is precisely why thermocouples work. A thermocouple is two different metals with two different seebeck effects. It's just that a single wire seebeck effect is not practical, but it does actually exist. See the link in the video description.
@TheKitsuneDen
2 жыл бұрын
Even though this is a while ago mate, I wouldn't go for the Fluke brand... I have a MM 300 Kleins, (not the best) but also I have a DL479T (much better- has a clamp and many other uses 🦊for my HVAC-R equipment 🙂)
Brillant! This video and the tear down of the fluke before it are just excellent. I love how this video tied in the mechanics (iso thermal block) of the tear down. Really love this style or combination of videos. I've got a EE degree and been in the buisiness (mostly software side of things) for around 18 years and your knowledge of the subject is expert and your ability to present it in easy to follow format is brilliant. Topics I've not seen since my college days are coming back to me
In the new videos about thermocouples with multimeters, you mentioned this old one - great and informative as usual. However I think I found a flaw at around 12:01 Minutes when you explain the NIST K-Type table.. I think it’s not cold on the X-axis and hot on the Y-axis. It’s a linear list from -270 - 0 °C in one degree steps. If you want to know the value for -13°C , you go the „-10“ line in the „-3“ column and get the value of -0.508 for -13°C.
Wow more involved than I expect rd this to be, thanks for taking the time to teach some of us, Dave
No doubt, Mr. Jones is god. I could sit 24/7 and watch this man but I've got my own miracles to pull. Thanks for the videos, Mr. Jones.
Thanks a lot for your post. This is the most professional explanation about the work of thermocouple I could find on KZread.. Great job!
Dave, thanks, I have pondered ocer this cold junction comp for about 3 months now, I didn't have the time to figure it all out myself. You cleared it all up in 30 minutes. These type of dare I say it "educational" videos is where you excel, clear concise and maybe the odd mistake but hey, they happen and people should take it on the chin. Thanks again
Have watched a lot of videos on TC's and the people making a big deal about the bead on the business end of the TC wire. They go to great extents to make sure there are those same two types of wires through the entire length from the tip back to the meter. All you really need to care about is making only one transition. When I worked for a US defense contractor in the 80's, I used to make thermocouples for the various temp tests from TC wire on spools of wire. I would cut off 10' of wire, strip and twist one end, and give it a blob of solder, then run the wire back to the patch panel, which was all standard nickle plated bus bars. No need for all the fancy plugs, extension cables, etc. Can use a 16 AWG extension cord if you need to. Besides, what you connect that TC to will not be made of the dissimilar metals.
At the begining of the video I think the demonstration of the seebeck effect with its explanations could be wrong. The soldering iron was placed near centre of the wire and the only reason for the voltage going up was because the soldering iron was not dead centre. I bet if the iron was place 30% of the way along the wire where the conducted heat would not really heat the banana jack the voltage output would be higher than in your example.
@simonbaker9909
9 жыл бұрын
That was my impression also.
@TheOysterjam
8 жыл бұрын
+Neil Bradshaw i think you are almost right, not quite. he placed the iron on a junction of Cu-Cu. even if he placed it at 5% down, it's still Cu-Cu (less thermal mass on one side, so you would see more change). the junction needs to be 2 different metals for best results.
I've read about this effect in a physics book, and it says that of course you will get a voltage across the wire if you are using just one metal, but it will be an inner field, which you can't measure from outside with a probe. So if you are using just one wire and getting a voltage reading, it will be due to other materials (metals) in your measuring circuit. Thats why you have to have the other metal with, of course, a different seebeck coefficient, to not zero out the voltage drop.
Thx to who suggested to make a video about Thermocouplers and of course thank u Dayv
Note that in order to have seebeck effects you must have TWO types of metals and they MUST be of different types, one should be p-type and the other n-type. The electrons' flow from the n-type to the p-type is possible because the p-type material has holes, which are positively charged spaces.
The seebeck effect does happen on a single conductor, it's just not called a thermocouple in this case. A thermocouple by definition uses two dissimilar junctions.
Funny that I keep coming across thermocouple stuff right now that it's relevant to my interests... I've been looking at making a reflow oven and am collecting parts at the moment.
Worth mentioning that the Curie Kink at about 300c was used to regulate early Weller soldering irons..
Nice tutorial. Lots of "stuff" in there I've never thought about before! Thank's Dave. THUMBS UP!
In theory, yes, it should cancel out, as it's effectively a thermocouple with two identical metals with identical seebeck coefficients. In practice, it can produce a small dynamic effect like you saw. You may have noticed it actually started going back down the longer the iron was on there.
Thanks Dave, I love videos like this! They provide a practical insight on things you hear people talk about a lot, but you might be uncomfortable with asking how they work because you don't want to look like a berk. I really enjoy videos where you share insight and experience, they're inspiring and I've learned a lot from them (for instance, I now have screw-you money stashed). I'd love to see more like this.
Good tutorial. Note: only some thermocouples work direct in liquids, like PVC or waterproof types, some soak up liquids and affect the result. Just a tip, use non-conductive fluids with TC's in a tube an immerse it into the liquids with the reference probe, keep in mind temperature ratings on all devices used.
Very clear and interesting, especially the temperature reference technique.
Yes, it's all related the thermoelectric effect. There is the "seebeck effect and the "peltier effect", sometimes joined and called the "peltier-seebeck effect".
You surprised me a lot!! When you had that cobber cable in the multimeter, I did think that it will never work! Seebeck's first rule was "Two like metals won't work" or something like that. :-)
I love the BTTF reference at 20:16!
Nicely explained but there is an error there, I think, about the NIST chart. I believe that the horizontal line is actually the subdivision of temperature rather than the cold junction temperature. You can easily confirm that by checking that the last column value is identical to the first column of the previous row when negative and the to the next row when positive.
Sometimes you are just amazing didactic, others, just brilliant. Thanks Dave.
get a large bundle of thermocouples in an array and DIY your own thermal imaging :)
Dave I always enjoy the tutorial videos.
Is there a shortest wire consideration for building a K type thermocouple? In my application I need 20cm probe lenght at most. Temperature to measure up to 250 max, mostly 220-230 range. I figure out that silver solder works good for hot juncture.
There is also a thermo couple that is copper filled with mercury that is used in water heaters and old floor furnaces. They actually mechanically push valves with the pressure created by the expansion of the mercury. Kinda confusing me by the use of the name over and over. But in these types the resistance change in the wire caused by the heat don't play a factor in the measurement? Thanks for the video, I always find them very enlightening. Well done as always.
Thanks so much Dave - Part of my job is validation of thermal processing of foods using T-type thermocoupes inside what is effectively a giant pressure cooker. I had some understanding of all of this, but your concise explanation cleared some concepts up and added a couple more. It's always easier to troubleshoot problems when you have some background right? Keep up the good work mate, love your videos.
I believe that another reason for the cold junction to have a high thermal mass is to limit the rate of change of temperature when the ambient temperature does change, so that the sensor that is measuring the cold junction can keep up with the rate of change. This brings me to another advantage of thermocouples - rapid response to temperature changes.
Quiet a heap of info on TC's. Thank you
Hi there, from a fan in Dominican Republic!
I would like more videos like this, it was very educational.
also resistive welding works well too. there are a few web pages where people have home made spot welders. For small stuff, not the automotive type There are spot welders made for jewellery work that can spot weld thermocouple wire together.
At the hot end of the thermocouple, why is there always a blob of something? I understand that the manufacturer needs to attach the two materials together but the blob seems a little excessive and that blob can be coated in something. Why is this?
@wildonpriddy1800
6 жыл бұрын
08Ultrasonic The blob is the welded junction,we made our thermocouples up by welding junction with acetylene torch,we ran temp.survey in hi temp.furnaces used for heat treating & annealing. The results were recorded on graph recorders for production tracking,also customer proof records. (It was thermo couple wire twisted at the end and then welded to form t.c. Hot junction)
If you heat the centre of the yellow loop with the iron you'd expect very little potential difference between the meter cable ends, or am I missing something?
One of your best Dave, cheers.
I wonder how Thermoworks are getting an accuracy of +/- 0.4 deg C with their Thermapen, which uses a K-type TC? Are they calibrating each device carefully, to characterise the individual sensor in each unit? How much error is random, and how much can be calibrated out?
Very informative. Was hoping you would to touch on PID control though
Another application is in Radioisotope Thermoelectric Generators (RTGs) where thermocouples are used to convert the heat from a decaying isotope into electricity. Wiki RTG
well, yes the effect don't go away, but your measuring at the end of a conductor, the temperature along the conductor don't matter, if it did the whole idea of the thermocouple would not work. but your right that it is good to know of the effect, as smd resistors also has dissimilar junctions, and and a small temperature difference from one end too the other, can give an error if you are working with precision electronics.
I did an essay for the IB at school on thermocouples. Ice and water on one side and different really hot stuff on the other. Crazy for my school poor man's lab.
I would love to see one of those LASER Thermometers taken apart and explain how those work... they look like magic!!..
Which one is the best multimeter for everything (professional, indesrtrial, home, electrical and electrician)
Hi Dave. I bought a Uni-T K/J type that came with two wires. Both have K Type embossed on the plug but they both have one red and one white cable. Are these K or J type thermocouples?
what is used in high temp situations, like foundry or metal processing(2000c to 3000c)?
Question: Does the length matter??? I wanted to measure/calibrate my soldering iron temps. I was going to buy a 191 tool, but later thought getting a proper 2 channel device would have more uses. So, if I cut the leads down at the plug end of one of my probes, so I just have a little stub sticking out, that I can touch my hot soldering iron tip to, will it still read the same? Love all your videos, Thanks for sharing. I know this video is from 2013, hope you still check comments...
Some want to know how to use the K-type sensor that came with an AstroAI DMM to measure a temperature. Could you hold the wire sensor at the base of a spark plug to get a cylinder head temperature of a running engine? The sensor wire would hardly be in good contact with the metal cylinder head.
did you read that Application Note you linked to, "Seebecks experimental work showed that if a voltmeter circuit wiring and the conductor under test were both the same material, then the net loop voltage is zero because [Sa*(T1-T2)+Sa*(T2-T1)=0]." Yes the effect is there in a single wire, but you can't measure it because either your probes are the same material as the wire, and cancel the effect out, or a different material and then they are the second material wire.
for the purpose of making a calibration of a thermocouple I want to know the existing methods of calibration please thanks
This was a beautiful demonstration of thermocouples and their physics. I feel however, RTDs, are the industry standard for high precision temperature measurements. I'm somewhat surprised that thermocouples are used so much when keeping track of junction temperatures and metal composition is such a pain in the ass in comparision. I suppose cost (which have dropped in RTDs recently) of the element is driving it, but it seems the effort in "reading" a thermocouple is much more complicated.
Yes, but both alligator-metal junctions will be at the same temperature (ambient temperature) so you won´t get any voltage. You need two junctions at diferent temperatures. If you put one alligator-metal junction at ambient temperature and the other at the temperature you want to measure, you will have a thermoelectric voltage, but you won´t know what temperature correspond to that voltage, because for that junciont you have not tables. The types of junctions are normalized.
Nice! I work with thermocuples and resistive sensor everydays!
16:00 that is a very clever analysis
I love your channel thanks, could you discuss how strain gauges work and the method for reading the output? Thanks again
Great, love to hear the basics from you..
Great video Dave. Very interesting indeed.
seems 4 viewers of the video forgot their monitor and speakers turned off while trying too watch the video. anyway, If you want a cheap multimeter with no built in thermal sensor but comes with a thermocouple, check out the craftsman 82345 multimeter (I think it just assumes that the ambient temperature is at 70F (21.11C) (purchased it for $15 a few years ago mainly to use as a battery tester, especially for the cordless drill which which does not have any overcharge protection).
If you want to produce some usable power get a peltier block and heat up one side while keeping the other cold. You can actually get enough power to light up small lamps and stuff.
Anyway as usual, a very nice tutorial, learned something again today. Thanks mate.
Back to the Future quote
Are naked probes that come with multimeters safe to immerse in liquids, assuming they don't corrode the metals? Somebody warned not to immerse the cable of a meat thermometer. How do meat and medical thermometers work? The don't show any internal temperature when unplugged.
If the thermocouple has to be compensated with a temperature sensor inside the meter, what is the temperature sensor's junction referenced to? Obviously, there cannot be an infinite number of junctions referenced to other junctions referenced to other junctions.
Brilliant teaching. Excellent!
Yup lol i love all his other vidoes and i find them quite informative..... and i must congratulate Dave on his other videos as i have prob spend alot of time watching most of them and learning alot along the way....... the copper wire and to soldering iron was just proving that the wire he was using had some small inhomogeneities in the wire...... Its was nice to see the ITS90 polynomial coefficients as i had to type all those into excel macros :).... Thank you for the video :) Dave
The reason they didn't solder it is you can't solder those metals. Ultrasonic welding is another option.
Back in the day we actually had ice bath units as the cold junction reference. As I recall the bath was controlled by monitoring the volume of ice water so that the system was maintained with some ice and some water. If it was allowed to get colder the volume would increase as more water turned to ice so by making sure the volume didn't increase or decrease beyond certain limits then you knew there was a solution with ice and water so you could know for sure you were at zero Celsius.
Awesome video!, thanks. 👍👍
Yes, you are right, oops. Will correct in annotation.
Can this be used to test circuit boards for showing, let's say and fuse or on board capacitor or small diod.
I know you're an engineer Dave but how do you remember all that stuff? Percentages types of metals.....wow!
And also the voltage you will get will be very small, because the types of juncions normalized were choosen because they produce the biggest thermoelectric voltage compared to other junctions. The bigger the voltage you get, the bigger the precision you get.
what about the diameter of the cables, could changing that, influence the outcome?
Great video, I really like the tutorial videos.
When you used the soldering iron to heat the centre of the wire I expected the meter to show no change. Does the effect of the temperature differential between the centre of the wire and each end not cancel out ?
Brilliant. You live and learn
Dear Sir, Thanks a million times for your explanation, it was very simple and straight to the point. I have this question please; If we have a TC type K connected to a Temp. PID, Can we use two copper wires connected to Fluke DMM form one end and the other end of these two wires is connect at the PID exactly at the same TC terminals. So the PID can read the temperature from the TC and at the same time the Fluke Digital Multi Meter can read the Temperature from the TC as well. Do you think it will work, or the two copper wires will affect the reading ? as i do not know the right gauge and length of these two copper wires ? Please help Sir,
Looks like you've misinterpreted calibration tables.Horizontal axis is just a least significant digit of temperature. Such form is used for compactness. Temperature value is a delta between cold junction and hot junction. Usually cold junction is held at 0 degrees during measurement of these tables.
Can i use these thermocouple to detect animals skin temperature? Thanks
I've got a ZD 937 soldering station. It has an atmel at24c02 2KByte IC, now I'm wondering, perhaps it's used to store the nist data table, hmm....
@lolman123401
9 жыл бұрын
bergweg Read the IC and post the results on the EEVBlog Forums!
@bergweg
9 жыл бұрын
***** Do you also have a zd-937?
@lolman123401
9 жыл бұрын
bergweg Naw, I'm poor, I use a cheap ass thing with the PSU in the handle
Finally some tutorials :) Go on brother :)
The cool thing is that this Peltier-Seebeck effect is reversible !! Like in Peltier cells, by powering the Thermocouples, you get a Heat Pump !! In fact, Peltier Cells are many Thermocouples joined together.
Brilliant experiment, Thumb UPed already....
Why there is no temperature probe calibration should I say compensation in the Expensive Agilent U1252B, meter? My probe is +3C (checked in water bath with calibrated mercury thermometer), so if I can adjust various settings of the multimeter why I can't set the +3C to the temperature readings? I sometimes forget to compensate it. Why Agilent is so dumb not to include this?
Everyone else is watching the Super Bowl. And I'm sitting here watching Dave Jones be AWESOME.
I know you explained noise but it was not satisfactory. How do you reduce the effect of noise here? When you trun on the multimeter basically you will get 0-2mv noise from environment. How do you know your signal from thermocouple is accurate in that small range?
@Arek_R.
8 жыл бұрын
That case you talking about when DMM show few mV is because you have not connected probes to anything and it pick up some crap. But when we close the loop(connect to something) then noise dont have enough power to interfere readings.
WOW, its both, magic and complex ... how many coeff take the poly ?. Brilliant vid Dave.
Yep got pne pf thoes on top of a wood burning stove running a fan.
Look at peltier modules for that purpose.
Dave, what do you think of the cheap little yellow thermometers you can find on eBay by searching for "k type thermometer" and then sorting by Location > Australia Only? Do you think these would have features such as the thermal masses near the sockets like the Fluke one you showed in this video? Or would they just be useless crap?
Could someone tell me the best way to attach that point end of thermocouple to the measuring surface. It is so difficult to keep it in place without a fixture.
@tomrestis4533
4 жыл бұрын
Just use some thermally conductive epoxy like Item# OB-100-1/4 from OMEGA. More importantly, to get an accurate reading the thermocouple should lay flat on the surface for roughly 10 x the wire diameter so that enough heat can be transferred into the thermocouple to balance the heat wicked away up the length of the thermocouple.
@sbybill3271
4 жыл бұрын
@@tomrestis4533 Very helpful. Thank you
Very, very good video indeed - thx !
Great video, very helpful.
Great video as usual Dave, thank you! But how can I test a thermocouple? How can I determine whether is giving a correct reading or not? Is it just about checking continuity between the two junctions?
@NiftyFingers
9 жыл бұрын
Alberto Silvestrini I would guess maybe voltage and temperature for the metal were correlated and put into a table by some other means, and then the voltmeter uses some sort of table lookup.
@martinda7446
9 жыл бұрын
Alberto Silvestrini I suppose you can test with ice water for 0 and boiling for 100. the boiling depends on pressure but your ice water will be absolute (and safe)