Measurement - Thermocouples

A thermocouple consists of two wires of different metals, joined by a weld at one end. A small voltage is created between them (the Seebeck effect), rising as the temperature increases. Of course, as soon as you connect the other end to your circuit, another thermocouple is created, and this is compensated for by measuring the temperature at the instrument end as well, and compensating appropriately.
Thermocouples are classified according to the metals used - for use in a kiln, only the K, R and S types can cover the full temperature range. Although slightly different, the R and S types are to all extents and purposes interchangeable. Note that K and R/S thermocouples output different voltage levels, so you need to ensure that your kiln controller or pyrometer is set to match your thermocouple type.
Accuracy varies according to the type of thermocouple, with the table below giving the accuracy levels when brand new:

TypeASTM StandardASTM SpecialIEC Class 1IEC Class 2
KGreater of ±2.2°C or ±0.75%Greater of ±1.1°C or ±0.4%Greater of ±1.5°C or ±0.4% (max 1000°C)Greater of ±2.5°C or ±0.75%
R/SGreater of ±1.5°C or ±0.25%Greater of ±0.6°C or ±0.1%Greater of ±1°C or ±(1 + 0.003(t-1100))%Greater of ±1.5°C or ±0.25%

Dismissing the Special and Class 1 standards for K type thermocouples, as kilns operate well outside of their rated temperature range, this gives the following errors at 1300°C:

TypeASTM StandardASTM SpecialIEC Class 1IEC Class 2

Note that ASTM (standard EM230) recommend that K type thermocouples are not used above 1260°C, and don't give output voltages above this temperature. This is with AWG8 wire - the upper limit comes down as higher AWG (thinner) wires are used. Similarly, IEC 60584-1 only gives output voltages up to 1200°C for Class 2, and 1000°C for Class 1 for this type. It should only be used in an oxidising or inert atmosphere, unless protected b a sheath.

Degradation and Errors

The wires of K type thermocouples start deteriorating above 750°C, and also deteriorate in low oxygen environments or in the presence of many fumes, especially sulfur. Between 800 and 1050°C, one of the wires is prone to chromium depletion ("green rot"), especially in low oxygen environments, leading to drift and eventually failure. If cooled faster than 100°C/hr between 400 and 600°C, an undesirable phase change occurs in one of the wires, giving errors of about 5°C in this range until heated up above 700°C.
Note that R/S thermocouples are sensitive to contamination by silicon in the sheath, especially in a reducing atmosphere, which can introduce errors of up to 10°C. This can be minimised by using thicker wires, as they have a lower ratio of surface area:volume, and wires of 0.35 or 0.5mmØ are available for this. Alternatively, an aluminium oxide sheath can be used in preference to a ceramic one.
Both types may suffer errors from the wires picking up electrical noise. The best ways of avoiding this are first routing the wires clear of any noise, and secondly twisting the + and - wires of both the thermocouple and any extension wires. In an electrically noisy environment, use of shielded wires may be beneficial.


Whilst R/S thermocouples do not suffer from drift, this can be significant with K type thermocouples.
Drift occurs because of changes to the thermocouple wires, which affect the Voltage:temperature response. This may be in the structure of the metals, oxidation, elements being given off, or material being absorbed from the atmosphere. Drift starts the first time the thermocouple is used, and increases with time. An error of more than 20°C after 500 hours firing is not unusual.

Extensions Wire

Generally the thermocouple wire is too short to reach the kiln controller. In this case, thermocouple extension wire matching the thermocouple type should be used - this is a heavier gauge wire, minimising signal losses. It is made of different metals than the thermocouple wire to reduce costs, and as a result has a lower maximum operating temperature, generally 200°C.
Extension wires may be run up to 100m, if of a suitable size to avoid voltage drops, though in general lengths should be kept as short as possible.
Use of an extension cable will introduce additional errors, though. For a Class 2 thermocouple at 1000°C, this may be 2.5°C.
The wire is available with a range of insulation types, designed for different upper temperatures, e.g.: PVA 105°C; PFA 250°C; glass fibre 400°C; and high temperature fibreglass 800°C.
There are also standard thermocouple connectors in standard plastic (to 200°C), high temperature
plastic (350°C) or ceramics (650°C).


The thermocouple may be placed in a ceramic sheath. This slows the response time from less than a second to a couple of seconds, but protects it from a reducing atmosphere or metallic or other vapours that may attack it. It also protects the thermocouple from being hit by kiln shelves or pots.

Wire colours

Standard wire colours are given in the table below (with - denoting not specified in the standard):

WireASTM ThermocoupleASTM ExtensionIEC (both)
K type
R/S type


The thermocouple only measures the temperature at the tip, and this should extend sufficiently into the kiln so that it is measuring the kiln temperature, rather than the temperature of the insulating bricks. Rules of thumb are that it extends into the kiln by four to ten times the sheath diameter, though I would suggest a longer distance so that it picks up radiant heat as well as the hot air of convective heat.
You can run a test to get the optimal thermocouple protrusion into the kiln: set the thermocouple a known distance out, then fire at your kiln's full rate for, say, an hour, logging temperature against time. Then extend the thermocouple further into the kiln, and repeat; if there is no change, then it is extended too far, whereas if the temperature is higher then the new position is better.