Okay to twist thermocouple wires?

[refind]

I am trying to make some K-type thermocouples. I have wire. Is it okay to twist the ends together at the sensing tip? There is no tension in the wires, so is it okay? I don't see why everyone says to weld them or solder them.
I don't have a welder and the solder won't stick (yes I used flux).

 

[Averagesupernova]

Crimp them with some kind of crimp-style connector.

 

[jim hardy]

You'll get by for some time with twisting them .

Cut off the part that has solder on it.

There exist nickel crimps for that purposehere's the best place i know of for thermocouple accessories.
And a good introduction to them.
http://www.omega.com/guides/thermocouples.html

but i didn't find those crimps.

 

[OCR]

here's the best place i know of for thermocouple accessories.

Great link Jim... thanks...

 

[jim hardy]

dont know what is your application

in old timey steam plants they'd sometimes peen them into a shallow drilled hole
like this
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20100014144.pdf

aha ! Found those nickel crimps !
http://marlinmfg.com/pdfs/_crimpjunction.pdf

old jim

 

[Mike_In_Plano]

The thermocouple is one of the most misunderstood pieces of lab gear. There are plenty of folks that swear it can only be make by welding, others accept soldering, another group accepts twisting... Then you get into junctions and the arguments start up all over again.
Fortunately, General Electric used to train their engineers and technicians in most everything, and they wrote a complete book on thermocouples. It wasn't a very thick book, but it put most all arguments to rest.

Simply put, if you succeed in making a clean bond between the two wires and the bond experiences the measured temperature uniformly, then it doesn't matter what incidental metals are involved. Thus twisting, welding, soldering, crimping - they all work, until... Until corrosion sets in between the connections. Thus, twisting is the least preferable over time because the metal is neither fused nor crushed together sufficiently to form what's termed "a gas tight connection."

That said, almost every experiment I ever run used twisted connections because it is easy.

If you're attempting to measure the temperature of electronic components for reliability and safety assessments, then you really want a tiny junction, a poor thermal conductor for the TC, and to pass the across the body of the part to reduce the error due to heat being sucked away by the TC wires.

 

[refind]

I don't want to add some crimp because it will heavily increase the time constant. I want to keep response pretty fast (don't we all).
How about I twist them and dip them in a little epoxy? I only need to go to 175C at the most. They will be at room temp 99.9% of their life. Good idea?

 

[dlgoff]

... the solder won't stick (yes I used flux).

I'm curious why you can't solder these. I've soldered hundreds of type K thermocouples with a 15 watt iron using a low silver content solder from Radio Shack. This would give you a smaller "time constant" than from twisting. And Epoxy = Good insulator => larger "time constant"

http://demandware.edgesuite.net/sits_pod26/dw/image/v2/AASR_PRD/on/demandware.static/Sites-radioshack-Site/Sites-master-catalog/default/v1421881229251/images/06400013_00.jpg?sw=350&sh=350&sm=fit

 

[jim hardy]

I was told by an old timer to never ever solder a thermocouple. So i never have.
The idea got reinforced when i had to replace an extension wire that had been soldered. It was creating a 2 degF error in a ~430 degree measuement,

I think the trouble with soldering them is some fluxes can contaminate the thermocouple wire where it's heated , by diffusion, changing its microvolts per degree. Who knows what flux the splices in my example above were soldered with.

Here's an Australian government publication that gives instructions for soldering thermocouples.
I note they specify what flux to use.
http://casa.gov.au/wcmswr/_assets/main/rules/1998casr/021/021c99s2c16.pdf

I expect also that with some practice you could weld them by touching the ends together while powering with a car battery.
That would be a handy skill to have.

dlgoff seems to have got along okay. Let us know how you come out...
old jim

 

[jim hardy]

ps Omega sells fast miniature thermocouples...

i'd try twisting the end tightly, then apply a bead of epoxy above the twisted tip to stabilize it against untwisting.

 

[Mike_In_Plano]

I'll point back to the sacred book of thermocouples by GE. What matters is that you have an electrically sound connection, even if you have transient metals in the connection region. As long as it's all the same temperature, the energy transitions will cancel out as you go from one material to the next. Solder's good if you have an alloy that will wet the two alloys.

 

[dlgoff]

The idea got reinforced when i had to replace an extension wire that had been soldered. It was creating a 2 degF error in a ~430 degree measuement,

For extensions, you need to use the proper metals. For K-Type, you would want to use something like these; from the Omega link.

What matters is that you have an electrically sound connection, even if you have transient metals in the connection region.

Exactly. The reason I brought up the soldering was to provide the OP with a smaller, less "time delaying" joint.

 

[Mike_In_Plano]

As a kid, I worked in a TV shop, and my senior tech advised me that "If I were told a pink elephant comes out of the set at 5:00 everyday, I didn't have to believe it, but I did have to show up at 5:00 with a pink elephant gun."
Since then, I've experienced all matter of "inconceivable" things. Which has me itching because my trusty training is at odds with Jim's observation and I tend to trust both.
That said, I'm itching to perform a most deliberate experiment in which I fabricate a twisted thermocouple and another thermocouple that deliberately has a section of solder between the leads. Then wire them opposing in series and place them on the same isotherm in an oven.
The difficulty being that I don't have a decent oven or a microvolt-meter, though I'm sure I can cobble the latter. I suspect a weekend project is brewing...

 

[jim hardy]

A small clean joint ought to be the key.

The microvolts are actually produced along the wires leading away from the joint, not at the joint itself.
So it's important to not alter the metal in the area where it experiences a temperature gradient.

The practical considerations of thermocouples
The above theory of thermocouple operation has important practical implications that are well worth understanding:

1. A third metal may be introduced into a thermocouple circuit and have no impact, provided that both ends are at the same temperature. This means that the thermocouple measurement junction may be soldered, brazed or welded without affecting the thermocouple's calibration, as long as there is no net temperature gradient along the third metal.

Further, if the measuring circuit metal (usually copper) is different to that of the thermocouple, then provided the temperature of the two connecting terminals is the same and known, the reading will not be affected by the presence of copper.

2. The thermocouple's output is generated by the temperature gradient along the wires and not at the junctions as is commonly believed. Therefore it is important that the quality of the wire be maintained where temperature gradients exists. Wire quality can be compromised by contamination from its operating environment and the insulating material. For temperatures below 400°C, contamination of insulated wires is generally not a problem. At temperatures above 1000°C, the choice of insulation and sheath materials, as well as the wire thickness, become critical to the calibration stability of the thermocouple.

The fact that a thermocouple's output is not generated at the junction should redirect attention to other potential problem areas.

3. The voltage generated by a thermocouple is a function of the temperature difference between the measurement and reference junctions. Traditionally the reference junction was held at 0°C by an ice bath:

The ice bath is now considered impractical and is replace by a reference junction compensation arrangement.

http://www.capgo.com/Resources/Temperature/thermocouple/thermocouple.htmlWhat we found in that extension i mentioned was a soldered splice about every fifty feet.
It was the correct thermocouple extension wire.
We replaced it with a single piece, no splices, and our measurement improved to almost perfect.

Decalibration
Decalibration is the process of unintentionally altering the makeup of thermocouple wire. The usual cause is the diffusion of atmospheric particles into the metal at the extremes of operating temperature. Another cause is impurities and chemicals from the insulation diffusing into the thermocouple wire. If operating at high temperatures, check the specifications of the probe insulation.

http://www.azom.com/article.aspx?ArticleID=1208

I think that our extension wire had been soldered with a gas torch and got way too hot.
But i really don't know. Soon as i saw those soldered joints the old-timer's words came back to me so we pulled new extension wire, about 400 feet of it..

Mike: what does the GE book say about soldering ? Any precautions about overheating?

Refind : how much precision do you need? Are you after precise numbers, or response time?ps thanks all for the kind words...

old jim

 

[dlgoff]

... it's important to not alter the metal in the area where it experiences a temperature gradient.

When the damaged (stressed) portion of a thermocouple wire is used in a large
temperature gradient, your validation data will have errors. How much error is
determined by the manufactured quality of the wire.

bold by me

http://www.ge-mcs.com/download/validation/Validation_TC_Datasheet_final.pdf

I suspect a weekend project is brewing...

Note the bold. It does make a difference.

 

[Bystander]

What we found in that extension i mentioned was a soldered splice about every fifty feet.

And with NO thermal "lagging," I'll bet. The headaches in keeping all joints, connectors at constant T sell lots of aspirin.

 

[jim hardy]

Here's what Omega has to say

http://www.omega.com/temperature/z/pdf/z021-032.pdf

Z-29
Poor Junction Connection
There are a number of acceptable ways to connect
two thermocouple wires: soldering, silver-soldering,
welding, etc. When the thermocouple wires are
soldered together, we introduce a third metal into the
thermocouple circuit, but as long as the temperatures
on both sides of the thermocouple are the same, the
solder should not introduce any error. The solder does
limit the maximum temperature to which we can subject
this junction. To reach a higher measurement
temperature, the joint must be welded. But welding is
not a process to be taken lightly.⁵
Overheating can
degrade the wire, and the welding gas and the
atmosphere in which the wire is welded can both diffuse
into the thermocouple metal, changing its
characteristics. The difficulty is compounded by the very
different nature of the two metals being joined.
Commercial thermocouples are welded on expensive
machinery using a capacitive-discharge technique to
insure uniformity

Sounds to me like solder should be okay provided one is attentive to cleanliness and uses modest heat. Anecdotal evidence for problems exists so technique must be a factor.OP's difficulty was in soldering to Chromel and Alumel

I don't see why everyone says to weld them or solder them.
I don't have a welder and the solder won't stick (yes I used flux).

Here's an article that describes how one person silver-soldered Chromel - Alumel extension wire for his airplane.

http://www.aeroelectric.com/articles/excerpt.pdf
page 7 of 12

 

[Mike_In_Plano]

Unfortunately, my handy GE thermocouple book has been lost through the years. It didn't cover the underlying physics in any case. There is a book I found by Daniel D. Pollock that seems to explain more than anyone healthy would care to know - it goes for over $300.

If there's a three paragraph explanation to the underlying physics behind these things, I'd love to hear it. I've constructed them in thin film and still don't have a good explanation for why they behave as they do.

 

[jim hardy]

My understanding is not deep.
The simple mental image i keep in my head, and which i use to explain to folks, goes like this:

We know that any metal contains a sea of electrons that are loosely bound, in the very outer shells of its atoms.
That's why metal conducts so well.
We also know that temperature is molecular motion.
Thirdly we know that different atoms have different affinities for electrons.

Pardon the jargon here, trying to paint a simple mental picture:
Heating one end of a wire rattles the atoms there
loosely bound electrons in the hot region are shaken out of their proper place and migrate toward the cool end
causing a small potential difference along the part of the wire that's experiencing the temperature gradient.
Not surprisingly that potential difference depends on the makeup of the wire.

So: if i could measure the voltage between the ends of a single wire experiencing a temperature gradient,
i could plot its microvolts per degree.
Of course that'd be impractical because one of my my voltmeter's wires, the one touching the hot end, would also experience a temperature gradient.

But if i joined two dissimilar metal wires and heated the joint
at their far end i'd measure the difference between their temperature induced voltages, and at room temperature.

That's stated more eloquently here:
http://www.mstarlabs.com/sensors/thermocouple-cold-junctions.html

The terms hot junction and cold junction, as applied to thermocouple devices, are mostly historical. You don't need to have any junctions to get thermocouple effects. If you heat one end of a metal conductor and hold the other end at a constant reference temperature, two important things occur.

1. Heat flow. There is a thermal gradient, so heat flows from the hot end to the cold end. With small-gage thermocouple wire, very little thermal energy actually reaches the cold end, and the thermal gradient is typically not constant along the wires because of heat loss.
   
2. Seebeck effect. Energetic electrons at the hot end diffuse toward the cold end, pushing less energetic electrons along with them, resulting in a higher static potential at the hot end relative to the cold end. The larger the temperature gradient, the larger the potential difference. (There are additional contributing effects when dissimilar materials are joined.)

In practice, it is difficult to measure the Seebeck effect directly. When you attach measurement probes, there is a thermal difference across the probe leads, producing additional thermocouple effects that interfere with the measurements.

Classical thermocouple loop configuration
To make the thermal effects measurable, two different metal conductors are used. They must be chemically, electrically, and physically compatible. They produce different electric potentials when subjected to the same thermal gradient.
more at source

For me the light came on when i remembered as a kid shaking a tree to get the fruit.
Different varieties of tree have different affinities for their fruit, as do different metals for their conduction electrons.
I grew up in subtropics shaking mango and loquat trees.
Surely you folks from temperate zones noticed the same thing?

I hope this preposterous oversimplification is some help.
We must oversimplify and exaggerate to get our thinking started along a right path
and refine from there.
When our microscopic and macroscopic views come into agreement, and our thought experiments based on the two predict the same results, is i believe when we are beginning to understand.

Imagine those poor guys trying to figure it out before discovery of the electron...
http://thermoelectrics.caltech.edu/thermoelectrics/history.html

Thermoelectric Effects - Early study of Thermoelectricity 1820-1920

In the 100 years before the world wars thermoelectricity was discovered and developed in western Europe by academic scientists, with much of the activity centered in Berlin.Seebeck Effect

In 1821-3 Thomas Johann Seebeck found that a circuit made from two dissimilar metals, with junctions at different temperatures would deflect a compass magnet [1]. Seebeck initially believed this was due to magnetism induced by the temperature difference and thought it might be related to the Earth's magnetic field. However, it was quickly realized that a "Thermoelectric Force" induced an electrical current, which by Ampree's law deflects the magnet. More specifically, the temperature difference produces and electric potential (voltage) which can drive an electric current in a closed circuit. Today, this is known as the Seebeck effect.

old jim

 

[jim hardy]

i have to take a break
my internet has become unusable with PF

ignores post button for ten minutes then posts nine copies
will try again in a week or so

old jim

 

[Tenngineer]

My understanding is not deep.
The simple mental image i keep in my head, and which i use to explain to folks, goes like this:

We know that any metal contains a sea of electrons that are loosely bound, in the very outer shells of its atoms.
That's why metal conducts so well.
We also know that temperature is molecular motion.
Thirdly we know that different atoms have different affinities for electrons.

Pardon the jargon here, trying to paint a simple mental picture:
Heating one end of a wire rattles the atoms there
loosely bound electrons in the hot region are shaken out of their proper place and migrate toward the cool end
causing a small potential difference along the part of the wire that's experiencing the temperature gradient.
Not surprisingly that potential difference depends on the makeup of the wire.

So: if i could measure the voltage between the ends of a single wire experiencing a temperature gradient,
i could plot its microvolts per degree.
Of course that'd be impractical because one of my my voltmeter's wires, the one touching the hot end, would also experience a temperature gradient.

But if i joined two dissimilar metal wires and heated the joint
at their far end i'd measure the difference between their temperature induced voltages, and at room temperature.I hope this preposterous oversimplification is some help.
We must oversimplify and exaggerate to get our thinking started along a right path
and refine from there.
When our microscopic and macroscopic views come into agreement, and our thought experiments based on the two predict the same results, is i believe when we are beginning to understand.

Imagine those poor guys trying to figure it out before discovery of the electron...
http://thermoelectrics.caltech.edu/thermoelectrics/history.html
old jim

I hate to bring an old thread back to life, but what an amazingly simple explanation! I have been wrestling with the TC in my gas fireplace, and struggling to figure these things out. Thank you for this post Jim!

 

[jim hardy]

Well ! Thank you for the kind words.

Good luck with your gas burner.

Those TC's typically make only tens of millivolts(some as much as 150mv) to operate a very sensitive solenoid coil inside the gas valve. Ohmmeter check will probably find your trouble. Ace hardware carries replacement TC's.

A friend of mine had a gas log that had somehow got water into the gas valve and corroded its innards terribly . Thermocouple made millivolts but the valve wouldn't open even from a AA battery. If your valve is stuck closed don't try to fix it, replace it.

 

[Averagesupernova]

If your valve is stuck closed don't try to fix it, replace it.

Wiser words have never been spoken.

 

[Baluncore]

Unfortunately, my handy GE thermocouple book has been lost through the years. It didn't cover the underlying physics in any case. There is a book I found by Daniel D. Pollock that seems to explain more than anyone healthy would care to know - it goes for over $300.

Author is: Daniel D Pollock. Title is: The theory and properties of thermocouple elements (ASTM special technical publication).
You can find second hand copies from about US20.00 + shipping. http://www.bookfinder.com/search/? If bookfinder finds one on amazon or ebay, etc, then buy it through the appropriate PF link on this page.
https://www.physicsforums.com/threads/support-pf-buy-on-amazon-com-from-here.473931/

As a kid, I worked in a TV shop, and my senior tech advised me that "If I were told a pink elephant comes out of the set at 5:00 everyday, I didn't have to believe it, but I did have to show up at 5:00 with a pink elephant gun."

Reports of pink elephants at 5PM make perfect sense, and you don't need a gun.
In late afternoon the sun is setting lower and so shines through the new TV room's lovely green curtains. Even a B&W TV looked pink in that situation. Was it a coincidence that at 5PM, the children's cartoon “Babar the Elephant” was on TV.
Pink pictures did appear on B&W TVs due to room lighting, they were reported by at least one serviceman as an unusual service call.The problems I have had with thermocouples have been failures at high temperatures over time.

When solder is used, the solder melting point lowers the upper temperature possible. For low temperatures, solder is acceptable as it protects the surfaces from corrosion.

A fusion weld results in a bead of alloy that has a lower melting point than the wire materials. At high temperatures over time, the alloy bead will fail first.

A quick twist or a secure crimp presses the two wires together and so creates a surface contact. If, over time, the surfaces in contact corrode, then fusion welding or soldering would have been a more appropriate termination solution. In any case, diffusion over time at high temperatures will form an alloy with the same characteristics as a fusion weld bead.

 

[jim hardy]

A quick twist or a secure crimp presses the two wires together and so creates a surface contact. If, over time, the surfaces in contact corrode, then fusion welding or soldering would have been a more appropriate termination solution. In any case, diffusion over time at high temperatures will form an alloy with the same characteristics as a fusion weld bead.

that's why we preferred Chromel-Constantan over Iron-Constantan, neither Chromel nor Constantan rusts .

 

[Barry Kahn]

Very small sterling crimps seem to work well based on a few hours testing. I'm using a thermocouple to monitor the surface temperature of our wood stove. I bought a welded-end TC wire initially but it is only 3 feet long and I wanted more distance between the stove and the electronics, so I bought 50ft of type K wire from Omega and cut a longer piece. Tried twisting the wires to join the ends but that didn't work very well. Then this morning I thought of using sterling crimps. (I own a bead store and sterling crimps are used in making jewelry.) Tested it and the 1mmx1mm crimps work great with the 30 gauge wire. It's been on the stove for about four hours now and based on that brief trial, I'd say this is an easy and inexpensive way to join the ends. Can't say anything about longevity yet but as far as ease of assembly, this was dead simple.

 

[Averagesupernova]

If your welded end thermocouple was type k there is no reason you couldn't have just added on with whatever length you want.

 

[SensorJP]

You can add extension grade TC wire to make it longer. You'll have to make sure the loop resistance doesn't get too high or you can start having secondary junctions and decreasing your sensor accuracy.

We mostly resistance weld or TIG the sensor wires together. If you can do it at all, soldering type K is a poor process because it typically requires a highly acidic flux that tends to reduce reliability and, if not cleaned well enough, eats away at surrounding material.