K-type hightemp and corrosion resistant thermocouple-but how to attach it?

[Leland_Gaunt]

Hello gentlemen

I need some help and hope i can find somebody who is familiar with k-type thermocouples (and extension wiring). Im not so much into electronics, but i wanted to build myself, for my chemistry reactions, a thermometer that is finally both able to handle at least 500 C and is also non reactive with bases or acids. Glassthermometers are first choice but they only reach 200-300C. I already have the regular, cheap "wire" k-type probes and also the 304 stainless steel probe rods. These can handle 500 C but they do react with certain acids. So i came over some probes that are made from Korund-Ceramics, Aluminiumoxid it is chemically.

So i just ordered one of these probes for 20 bucks and also a new K/J type Thermometer for 20 bucks from my trusted cheapo-brand Uni-T. Its not a boygroup, its a chinese brand that manufactures all kinds of testing equipment. While i was waiting for delivery, i realized i really dont even know much about K-types. The bottom of the probe has 6 screws, instead of 2, i asked the seller if its really a K-Type, he assured me it is. Theres also no cable coming with the probe, i dont know what i was thinking, probably i just wanted to cut of the end of one of these 2 dollar-wire thermocouples and screw it on the probe, using it as a cable. Now these cables are made out of different kinds of nickel-alloys, its probably not even possible or a good idea to use them as an extension cable. If i dont get a somewhat accurate reading, i could have stayed with my IR-thermometer.

So my questions are: is this indeed a probe that i can make work on my K/J type thermometer? And if, what kind of cable shall i use? I can spare one of my old probes to get a yellow K-Type Plug that fits iinto the thermometer. But what kind of cable? Just regular copper wire?

[Kean]

A thermocouple uses two different types of metal to form a junction that creates a voltage potential that has a known relationship to temperature.

The K-Type thermocouple uses alloys called Chromel & Alumel (Nickel Chromium & Nickel Aluminium).
To avoid errors, you want to use the same type of alloys in your wiring all the way to your measuring device (DMM, etc).  The connectors are also made with these alloys.

On the probe you show, I imagine the two centre screws are for mechanical assembly.  Two outer screws connect to the wires from inside the tube.  The other two outer screws are for your extension cable.  Be careful with disassembly and reassembly as you cannot easily tell the type of alloy of the metal contacts, so you don't want to get the mixed up.  I think I can see plus and minus symbols on those connection points.

You can buy the connectors and extension cable from usual suppliers.  I bought some from element14 under the brand name Lab Facility.

Also note that there are different standards for colour coding of both connectors and cable insulation - IEC, ANSI, DIN, and JIS.  Your yellow connectors for K-Type are actually the ANSI (US) standard, and you might want the green connectors and cables per IEC/DIN standards.

Here are some links to PDFs with further info
https://assets.omega.com/landing-pages/colorcodes/tc_colorcodes.pdf
https://www.labfacility.com/media/productattach/t/h/thermocouple_product_guide_labfacility.pdf

I've also one attached below

J-Type uses different alloys to K-Type, so also make sure you know what you have and don't mix them up.
Annoyingly the different standards do re-use some of the same colours for different thermocouple types. 

You want to minimise the use of copper connections.  If curious, you can learn more by reading up on isothermal blocks and cold junction compensation.

[jpanhalt]

I am definitely not an expert on thermocouples.  The only time I ever worked at temperatures like your are using was making ketene from acetone many years ago.  My comment is made based only from what I see.

First, is it type K?  I would trust the seller.  You need to identify the two wires.  Alumel is magnetic and chromel is not.  In regular TC's, alumel usually has red insulation and is the negative.  Chromel (+ lead) can be some other color, usually yellow.  However, it's not surprising the wires from that probe don't have insulation based on its intended temperature.

What you show is a junction.  I suspect the center two screws are probably to hold the assembly together.  The TC wires are attached at one end, and the other end is for attachment to your instrument.  Can you just connect under that same screws used for the TC wires?  Probably, but that may not be as secure as using the other two screws.  Can you use ordinary copper rather that chromel alumel?  Yes, but then that junction becomes your cold junction and you will need to know its temperature (see attachment).  Are the metal bars connecting the leads plated copper or chromel/alumel?  If you use chromel and alumel, then your cold junction can be remote.  Read about the "Law of Intermediate Metals" as applied to TC's.

Attached is a picture from Omega of various ways to connect TC's.  Figure 1 seems most applicable to your situation, if you use copper.

[Leland_Gaunt]

this probe is rated 1300C i think. Theres not the usual cable inside, they claim its Rhodium-Platin or something like that. I couldnt find any cable for it, the sellers also dont offer a cable.

I cant imagine the wire has to be the same stuff than inside. It wouldnt make much sense to me. because there are already different metals in the circle, like the connection bracket for example. (Allthough i think i would try to screw it under the same screw)..but also the K-type connector...and the wiring inside the Thermometer right?


i mean what about those long 304 stainless steel rod probes? Theyre like 20cm long. Is the complete wire made out of the same alumel/nickel alloys? I was under the impression that these 2 alloys are only used inside the rod and that the cable from the rod to the thermometer is just regular copper cable...?


well meanwhile ill just hook up a copper table and make some comparisons to see if i get real or fantasy readings.

[nfmax]

You could use K-type compensating cable, rather than extension cable. Compensating cable matches the thermoelectric behaviour of the probe, but only over a limited temperature range about room temperature. It can be made more robust and flexible, at less cost, than extension cable. Use the ‘spare’ pair of screws to attach it. Provided the two connecting bars are at the same temperature, the intermediate metal should not introduce significant error.

[Kean]

 

this probe is rated 1300C i think. Theres not the usual cable inside, they claim its Rhodium-Platin or something like that. I couldnt find any cable for it, the sellers also dont offer a cable.

Are you sure that is a K-type then?  Rhodium would lead me to think it is type R or S.  (Edit: probably not at that price, and 1300C should be possible with K).

I cant imagine the wire has to be the same stuff than inside. It wouldnt make much sense to me. because there are already different metals in the circle, like the connection bracket for example. (Allthough i think i would try to screw it under the same screw)..but also the K-type connector...and the wiring inside the Thermometer right?

i mean what about those long 304 stainless steel rod probes? Theyre like 20cm long. Is the complete wire made out of the same alumel/nickel alloys? I was under the impression that these 2 alloys are only used inside the rod and that the cable from the rod to the thermometer is just regular copper cable...?

well meanwhile ill just hook up a copper table and make some comparisons to see if i get real or fantasy readings.

If you don't keep the temperature at those junctions at a fixed or measurable temperature then they can induce errors.  Those errors may be significant or insignificant depending on what you are doing, as well as due to environmental factors.

This is all well described in any article about thermocouples - e.g. the Reference Junction and Circuit Construction sections of https://en.wikipedia.org/wiki/Thermocouple

[Leland_Gaunt]

well i opted for the laziest way, just cut off the tip of a wire probe and attached it....i didnt really believe it would work, but it does somewhat. tested in water, both cooking and room temperature, difference to my other thermometers is less than 1 C.

Doesnt look nice but it works. The probe is very slow though, meaning it really needs a while.

it looks to me like the material of the wire only really matters where the temperature-difference is occuring. Where it stays cold and even-tempered, the Thermometer doesnt seem to give a fluff what the wire is made of.

[Kleinstein]

The wires from a relatively cheap type K probe can work fine as an extension wire. No problem with this and a sensible solution as type K probes are sometimes cheap.

That the TC reads OK at room temperature is no surprise - here the alloy don't matter as there is zero voltage, no matter what wire. The real test is the boiling water.

[Kean]

The wires from a relatively cheap type K probe can work fine as an extension wire. No problem with this and a sensible solution as type K probes are sometimes cheap.

That the TC reads OK at room temperature is no surprise - here the alloy don't matter as there is zero voltage, no matter what wire. The real test is the boiling water.

Yes, although the insulation near the junction may melt once the probe is subjected to high temperatures.

The test in an ice bath or boiling water will also confirm if you have the wire polarity correct.

[Leland_Gaunt]

nah, looks good. In cooking water its totally in line with my glassthermometer. The digital meat Thermometer is 2 degree off, but thats 99 cent. Literally. But im only gonna use it in case of very corrosive chemicals, otherwise ill use the metal probe. 5mm vs 8mm diameter is a big difference. 8mm is pretty difficult to insert into the reaction flasks or apparatus in some cases.

Why cant they just make something thats only 3mm wide, corrosionresistant, working to at least 500-600 C. and comes complete with cable and plug ill buy it..

[Hydron]

Haha, I was going to post "just be lazy and butcher one of the cheap ones that came with the Uni-T" but was too slow and you've done just that

For the best accuracy just try and avoid temperature gradients across the connection points (incase the metal of the plates are not correct for type K).

[Leland_Gaunt]

hi, i just wanted to give an update that might help some people with the same hobbies i am..i built myself a probe now for 7 bucks and its the best thermometer i ever had. I just bought a k-type wire in glasfibre-fabric, rated for 480 C. (5 bucks) and put it into a cheap 1ml-Glaspipette. (2 bucks). The open tip can easily be closed by melting, you dont even need a bunseburner, a good storm-lighter with a pointy, blue flame is enough. 500C is about the maximum this glas should handle, so the 2 parts fit together pretty well. Its made for very acidic or alkaline fluids anyway and the hottest i can think of is hot sulfuric acid that gets slightly over 300 C. hot when you concentrate it.

The corundum-thermocouple had 2 flaws, first its very thick, it doesnt even fit through a regular thermometer-Glasadapter, because its slightly thicker than 8mm. Its also very slow responding, while the glas-thing has a good reacting time, once you rub your fingers on the tip, it only takes a few seconds and the temp. starts rising.

[rogerggbr]

Excellent solution. As an ex-'tiffy (Instrument Artificer) I would always suggest just using the appropriate compensating cable (extension cable) for the whole thing, protecting the end as required, exactly as you have. Type K is tricky sometimes as there are 3 different types available, some more popular than others (KX, KCA, KCB). For my use I prefer type T (<300C), in my previous life we always used type J (<500C), both readily available.

[Kleinstein]

The compensating cables are just a cheaper alternative for a limited temperature range. Using the TC wire pair all the way to the connector / cold junction is perfectly fine, just possibly a little more expensive when the wire is long.

The thermal contact inside the glass tube can be a bit tricky and may not be that good. It may help to have the TC wire(s) slightly coiled inside the glass to get a bit better contact and bit more thermal resistance for the wire part.

One can usually get plain glass tubing by the meter rather cheap. It still depends if there is a shop as shipping could be expensive.
Shops with chemistry glass ware usually also tend to have just tubes to bend as needed.

[Leland_Gaunt]

you dont need a good contact inside. Its just a very small capillar with maybe 2mm diameter, the air inside will take the temperature of the glas in a matter of a few seconds. A very tight diameter that fits the thermocouple perfectly would just make for a even better response time. Not that important for me- The 0.5ml Pipette does have an even smaller diameter. Its just i only had the 1ml at hand.


i didnt use any extension. Just the 1m probe "SMPW- GG-K-30-SLE" from AliExpress. Dont know how accurate it is. But will be more than enough for me.