I have an issue with a Type K thermocouple setup. I have about 10-15°C difference between what is read at the thermocouple and what I read at the compensation cable before it connects into the DCS card. It is old WX comp cable over a hot float glass line. The T/C reads 105°C but getting 120°C on DCS. Now I believe the cable is degrading and the insulation is failing as it is brittle in parts. the Thermocouple/comp cable junction is at a high temperature. also the dissimilar metals of the T/C and comp cable may be giving us problems?

I'm not an expert with this so any info would help.

 

what are the displayed millivolts at the t/c and what are the displayed millivolts at the dcs terminals?

 

a further question is this. What is the temperature at the terminals you indicate are at high temperature. T/C extension wires are generally designed for ambient temperature compensations and are not meant for correction at process temperatures.

 

Electrical noise due to HT cables... Rans near to thermocouple cable....

mV may Different at Thermocouple leads and DCS connection leads in panel

 

T/C 3.7mV
DCS 4.1mV

T/C terminal head is at 55deg the cable runs through a steel ducting at 50 deg.

 

A 5°C error (maybe greater) is possible with some handheld digital thermocouple thermometers because of where and how the cold junction measurement is sensed. When the CJ sensor is buried under the plastic housing and does not really reflect the terminal temperature, CJ error can be significant.

For instance, suppose a handheld were stored in a poorly heated building in the winter so the handheld has equilibrated at 10°C. The handheld is taken and connected to a hot area where the thermocouple head gets up to 55°C. The terminals heat up to about 40°C fairly quickly, but the CJ sensor is buried and has a thermal lag of 10 or 15 minutes. Then CJ measurement could be much colder than the actual terminal block, say 25°C, resulting in a lower than expected temperature reading (40 - 25 = 15).

Thermocouple thermometry is only as (un)certain as its CJ.

It has been so long since I'd run into WX extension cable, I had to look it up.
For those as forgetful as I am, it is compensating cable for a K type thermocouple.

But the site I went to noted it with an asterisk*. The asterisk note says,
"* The compensating cable WX for K thermocouple included in previous ASA C 96.1 964 rules, has been excluded from ANSI MC 96.1 1975 rules"

Now I haven't a clue as to why WX was dropped from the 1975 rules, but guys older than I am might know.

Assuming your measurements are not unduly influenced by poor CJ measurement, I can make two observations:

1) If it were merely a short in the extension wire due to failed insulation, you'd see lower than expected temperatures, not higher than expected.

2) Error can be due to inhomogeneity of the extension wire, which creates junctions along the wire. It is sometimes called drift.

This link expounds on the topic. See pages 65 - 69; or 32-34 (pdf) on thermocouple inhomogeneity.
http://www.ahmetozkurt.net/MEC515/week6_temperature.pdf

Replacement is the only viable option.

 

you've an excellent discussion with many very good points, about the only issue that remains is to check your extension wiring. It may have been installed with junctions or splices of which you are not aware.

To have different mv at the T/C than at the DCS (if representative of the original posters installation) suggest that the wiring contains un-documented junctions.

 

Many good points I read..

Check for grounds on both K leads as close to the probe as possible "disconnected".

Then check for grounds in both wires of the extension cable. With a Fluke.

Is one of the extension wires magnetic?

If you are using an intrinsic barrier temporarily remove it. Zener may be leaky.

At the DCS is the TC being changed to 4-20 ma? Most thermocouple Inputs com with a temperature compensator. It may be disconnected or bad. Try another DCS input if possible. There is usually one thermistor close to the copper junction for each 8 channel TC input.

Use your calibrator to read both the TC in the field wait long enough that the TC calibrator sees the ambient. Then do it in back of the DCS panel. The measurements are disconnected from DCS not parallel.

Some calibrators have a table with junction temp offset.

If you do find from previous mention that there are terminals any where "double" the wires together and tied them down temporarily together or twisted together to eliminate any difference junction.
Are the actual readings in the middle of the K type range fairly hot. Ambients can sometimes be more suitable to J.

I have seen J extension runs for K. oops.

Your instrument must be using K leads.

 

one other diagnostic is to turn off the dcs compensation, and then check the readings.

 

I thank you all for your input and will be revisiting the issue this week.

But to add some more information when I had a few minutes. I did check another in a similar place and it too is reading higher at the T/C then DCS but one that is a in a lower ambient temp is reading exactly the same at T/C as DCS using the same tester with similar meter internal CJ readings. I will look into the questions raised with regards to joins and have a look at the DCS card CJ which is within the cabinet. These signals are not changed to 4-20 but straight to the DCS T/C input card after a run of at least 70m of extension cable.

Will keep you posted as to progress.

 

>Electrical noise due to HT cables...
>Rans near to thermocouple cable....
>
>mV may Different at Thermocouple leads
>and DCS connection leads in panel

Can i know, what will be the temperature reading in the DCS when Thermocouple burnout occurs?

 

First: Replace the K Thermocouple with its cable
If the measure is not accurate then there is a problem with your reading conversion process, and for that you should measure the voltage at your Com Pin data Sent-Receive and Gnd

Second: See if you’re K Thermocouple cable is not influenced by the near environment than your sensor.

Have in mind that the reading temp is only the voltage measurement at the end of your sensor plus the accuracy of your system to convert this voltage to the temp dimension.

Regards

 

> Can i know, what will be the temperature reading in the DCS when Thermocouple burnout occurs?

Not without checking to see
a) if the burnout detection is enabled
b) whether burnout is configured up upscale or downscale.

For heating control loops, input burnout is upscale, which drives the output low (shuts off the heat)

For cooling control loops, input burnout is typically downscale, which drives the output high (call for cooling continues, keeps the cooling full on)

What is 'typical' is not always the way the things get done or don't get done.