K thermocouple help

There's only one kind of type K thermocouple. Two types of J.

What is the color code? Yellow = K, Blue = T, Black = J (unless outside North America, then it's a crap shoot).

Perhaps it's a type T (Copper Constantan). There's also N but that's not so common.

--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

This is interesting to me as well.

Should the mV reading MikeK got be subtracted from the mV reading that would be obtained at room temperature?

At 177C, you should be seeing about 7.22 mV against an ice point reference, maybe 6.2 using a DVM at room temperature.

How and where are you making the TC-to-copper junction? That would have to be at 0C to have the measured temperature align with the published type K tables.

Low end DVMs aren't very good at resolving millivolts. I have at home a cheap Extech DVM that has a thermocouple connector. It seems to be pretty good.

Not if the meter isn't at room temperature.

The thermocouple just shows the temperature difference between the hot end and the cold end. Google "cold junction compensation".

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

Maybe this is mislabeled, but to be sure I don't think I understand the operation. The label says K Chromel/Alumel The probe looks like this, colors and all.

ISTR Mastech and Extech are rebranding of the same meter, ICBR. The way I'm using this is, plug the connector into my meter and read millivolts on the 200mv range. No reference to anything.

Thanks, Mikek

Look up, in the type K table, the TC voltage for room temp, the reference junction temp, and ADD that to the meter voltage. Then look up the sum in the tables.

As the ref junction temp approaches the tc junction temp, the meter voltage will approach zero.

Actually, it doesn't. The thermocouple curves are nonlinear, so the measured voltage isn't a function of the delta-T between the ends. The meter shows the difference in tc *voltages* at the two temperatures.

You are sticking a two lead thermocouple into some DMM that is different than what is was design for? If so I think all bets are off...

In the distant past I used two TC's and an ice bath reference.

George H.

Read JL's response to my post in this thread. Thermocouples work on a temperature difference (in, apparently, a nonlinear way -- thanks JL). So to find the temperature at the end you're measuring, you need to know the temperature at the end that's connected to your meter, and you need to do some math to get from a voltage to a temperature reading.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

Ok, I'll try that, need to run out for an hour right now. But, as I see it, both metals are at the same temp. Why this subtraction?

Ya, still confused. Back soon. Mikek

It would work if the DVM is accurate enough, and you estimate the cold junction compensation... plenty good enough to check a kitchen oven.

Yeah, probably the different tables have a different assumption for the cold junction temperature.

You have to look up the mV at the current room temperature** (on a given table) and then add that to the mV reading.

Take the reading (say 13.15mV, picking a number out of the air), add

** Actually the average temperature of the two T/C leads where they contact the multimeter leads, so keep your fingers off them.

--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

That's only an approximation (assuming the T/C is linear, which it is not). Good enough for a home oven though.

--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

That's type K, see my previous post on how to use the tables with a multimeter. You can resolve 100uV on even a cheapie meter, which is

auto zero to within far less than one LSD if they are 3 digit.

To get that accuracy you need to measure or control the cold junction reference closely. Ice-water slurry is the old fashioned method.

You do have to correct for the cold junction which is almost a 1:1 source of error (1 degree of CJC error = ~1 degree of reading error)>

For cooking in an oven a few degrees C makes little difference, of course.

--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

Ok, I'll try that, need to run out for an hour right now. But, as I see it, both metals are at the same temp. Why this subtraction?

Ya, still confused. Back soon. Mikek =========================================================================

Every time you have an electrical connection between dissimilar metals you get a "thermocouple". So besides the chromel/alumel junction at the end of your probe, you have two more junctions back at the meter, where the alumel touches the meter probe and where the chromel touches the other probe. I'm sure at some level of accuracy this doesn't work, but if you have alumel touching metal X and then metal X touching chromel, and all of that is at exactly the same temperature then you assume that these two junctions are in series and you treat it as a simple chromel/alumel junction at that temperature. Now, you have that junction, called the reference junction, either at a fixed, stable temp like 0C or sitting at room temperature (your case), and it is producing a thermocouple voltage that is in series with the voltage from the junction at the probe end at the unknown temperature so you have to subtract it from the probe voltage to get just the voltage at the probe so you can look that voltage up in the table to get the temperature. Now I'm not going to think hard enough to say for sure if the leads are +/- on the reference junction and -/+ on the probe junction or the other way round, so you might have to add instead of subtract, but the reason why is the fact that there is a second junction in the circuit :-).

----- Regards, Carl Ijames

Fair enough. My fluke DMM has a some special adapter and a degree's range. I hadn't thought about just using it directly. (one small thing I learned today, thanks)

George H.

It's an addition. The millivolts you read is:

V = f(T2) - f(T1) where T1 is the cold junction temperature and T2 is the hot junction temperature, where f(T) is the temperature to millivolt table (or function)

If you want T2 (hot junction temperature) you need f^-1(V + f(T1))

.. which is easy if youi know T1, the cold junction(s) temperature where the meter leads contact the thermocouple material.

Note: f^-1 represents the inverse function (millivolts to temperature), not the reciprocal of the temperature to millivolt table.

--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

Actually there are at least three junctions, so it is f1(T2) + [f2(T1) - f3(T3)]. Since both T1 and T3 are essentially the same, you can simplfy [...] to - f2(T1). I think this is the part Carl is confused about. Given the lack of a diagram here he likely still is. ;)

The two connections at the meter are at the same temperature, but they are different metals. In the end they equate to the same junction as the "hot" end of the TC, just in reverse so the voltage is subtracted.

--

Rick

This is the probe, note the bead on the end.

I assumed the wire was copper from the bead leads to the connector. Is that incorrect? Do we just extend your explanation further away from the meter and down the wire to the bead lead copper wire connection?

ALL VOLTAGES are in Millivolts. Type K thermocouple Voltage @ 212*F 4.096 Type K thermocouple Voltage @ 78*F 1.023 Calculated Voltage reading expected 3.073 Actual voltage reading on meter 2.400 (This not 3.073)

Second Test

Type K thermocouple Voltage @ 350*F 7.207 Type K thermocouple Voltage @ 78*F 1.023 Calculated Voltage reading expected 6.184 Actual voltage reading on meter 3.850 (This is not 6.184)

Ah! I see now my chart is referenced to 32*F.

Can I convert that or do I need a chart referenced to 78*F? Mikek

This is the probe, note the bead on the end.

I assumed the wire was copper from the bead leads to the connector. Is that incorrect? Do we just extend your explanation further away from the meter and down the wire to the bead lead copper wire connection?

Mikek ======================================================================================

Your probe is made from a piece of chromel wire and a piece of alumel wire. At the tip the two wires are welded together (spotwelded, tigged, gas welded, just twisted together, whatever) to make the probe junction, and at the meter end that cute rectangular plastic connector actually has legs made of chromel and alumel that the wires connect to so that there are no unnecessary junctions. The legs are different widths to polarize the plug to make it foolproof when you connect it to a thermocouple meter. The technical info at

is very good at explaining all of this, and how to use the tables.

----- Regards, Carl Ijames