What Is Platinum Rhodium Thermocouple?

A platinum rhodium thermocouple is a temperature sensor whose two legs are platinum-rhodium alloys, and it is bought for one reason that separates it from every base-metal couple: it is the instrument other instruments are checked against. The measurement is accurate, the output is stable over long periods at high temperature, and the price that comes with those two properties is a very small signal, a limited choice of alloys, and a material that loses its calibration when its surface is attacked.

The page below covers the three noble-metal types and what their legs are, the accuracy the standards allow, what the small signal means for the instrument the couple is wired to, what actually limits the life of one of these sensors, and the armoured form in which most of them are bought.

The letter fixes the pair of alloys and nothing else. Three noble-metal types are in general use, and the difference between them is the rhodium content of the legs, which is what sets the output, the range and the behaviour in a contaminated furnace.

The Type B point is worth stating in full, because it is the reason a buyer picks it. With platinum-rhodium on both legs there is no pure platinum limb for a contaminant to attack and no pure platinum limb for a deposited oxide to change, and published material on thermocouple selection gives that as the reason Type B tolerates contaminated atmospheres better than the other two noble-metal types. The trade is output: the same alloy pair produces a smaller electromotive force than Type S or Type R at the same temperature, and it is not normally used below 600 °C for that reason.

The tolerance of a noble-metal couple is published as an accuracy class, and the class is a band on the electromotive force curve and not a different alloy. A certificate that names a class is stating the deviation the batch is accepted within, and a tighter class is a narrower acceptance band on the same curve.

Two qualifications belong with the table. The ASTM equivalents of these classes are the special and the standard limits of error, so a certificate may name either system, and the two are not identical numbers. And the class describes the couple, not the measurement: the cold junction, the extension cable and the transmitter each add to the deviation, so the accuracy delivered at the control system is looser than the class on the wire certificate.

A noble-metal couple produces a fraction of the output of a base-metal one, and everything about how it is wired follows from that. Two consequences are worth having on paper before the sensor is ordered.

The first is the cold junction. A thermocouple measures a difference, so the temperature at the terminals has to be known and compensated, and with a small signal the compensation error is a larger share of the total. The second is the cable between the couple and the instrument, and there are two choices with different meanings. An extension cable is made from the same alloys as the couple, so the pair is simply continued to the instrument. A compensating cable uses different alloys chosen to give the same output over a limited temperature range, and it is the cheaper option where the run is short and cool. Either way the cable is part of the circuit and its own alloys decide whether the reading survives the journey.

Type B carries one practical advantage that the other two do not, and published type data states it plainly: since the output of the pair is negligible over the range from 0 °C to 50 °C, cold junction compensation is not normally required. In a plant where the terminals sit in an ordinary control cabinet, that removes a whole link from the measurement chain.

A platinum-rhodium couple does not fail by melting. It fails either mechanically or by losing its calibration, and the two mechanisms are published and separate. Both of them act on the metal, and the sheath and the packing around it can slow one of them and cannot touch the other.

The drift limit is the one that catches users of sheathed probes, and the distinction is worth stating once. A sheath protects the couple from mechanical damage and from contact with contaminants. It does not protect the wire from the loss of rhodium at its own surface, because the couple still reaches the furnace temperature, and a sheathed probe that has been in a hot oxidising furnace for a long period therefore drifts in the same direction as a bare one.

The armoured thermocouple, also called the mineral insulated or MI thermocouple, is not a couple assembled from loose parts inside a protection tube. The wires, the insulation and the sheath are built as one cable, drawn down through dies until the ceramic powder inside is compacted around the wires and the whole assembly is solid and flexible. That construction is where the properties of the product come from, and the three parts each carry a decision.

Four things follow from that build, and published guides to the construction state them in these terms. The compacted powder fills the space around the wires and shields them from moisture, gases, and contamination far better than loose ceramic insulators inside a tube, which is one route by which drift is slowed. The solid assembly resists vibration, shock, and crushing far better than an assembled probe, which is the mechanical answer to the brittle limb described above. The cable can be bent and routed around obstacles, which a rigid ceramic tube cannot be, and it can be drawn in very small diameters, so an MI probe fits where a conventional one does not. Response time follows from the sheath diameter, because a thin sheath has little thermal mass, and the compacted powder conducts heat to the junction efficiently.

The published working range for the construction depends on the sheath and the insulation together. Magnesium oxide insulated probes are published for use up to 2,400 °F, which is 1,316 °C, and MI thermocouples in general are described as serving from cryogenic temperatures into the range of 1,000 to 1,200 °C and higher where special alloys and designs are used. For a platinum-rhodium couple the practical ceiling in air is set by the drift described above and not by the cable, and the same distinction applies here as everywhere else in this material.

Four limitations belong in the same document as the advantages, because all four are decisions a buyer has to make and not defects to be avoided.

The fourth row is the one that is usually missed. A claim that the wires sit in a gas-tight, insulating state describes the design intent of the construction and the condition at the working temperature has to be checked against the datasheet, because the ceramic that insulates perfectly when cold conducts far better when hot. That is a property of the insulation and it applies to every maker of these probes.

One further choice belongs with the construction and not with the couple. The measuring junction is formed at the tip in one of three ways, and the three trade response against protection and against electrical isolation.

The legs. Type S is platinum with 10 % rhodium against pure platinum, Type R is platinum with 13 % rhodium against pure platinum, and Type B is platinum with 30 % rhodium against platinum with 6 % rhodium. Type S and Type R are used for reference and calibration work, and Type B is the one chosen where the furnace atmosphere is contaminated, because with platinum-rhodium on both legs there is no pure platinum limb for a contaminant to attack.

Because of the two properties in the first paragraph of this page, and because the standards state them as numbers. The tolerance is published as an accuracy class on the electromotive force curve, with ±1.0 °C over the 0 to 1,100 °C range for Types S and R, widening above that, and ±0.5 °C for Type B over its range. A base-metal type on the same standard carries a wider band, typically ±1.5 °C, which is why one is used for control and the other for reference.

It is a couple built as a mineral insulated cable and not assembled from loose parts: the two wires, a compacted ceramic insulation that is most commonly magnesium oxide, and a seamless metal sheath, drawn down together until the assembly is solid. Sheath materials range from stainless steel and Inconel through to specialty alloys and noble metals, and the construction buys protection against contamination, resistance to vibration and shock, bendability, small diameters, and a response time set by the sheath diameter.

It slows one route and cannot touch the other. The compacted insulation limits the ingress of gases and moisture to the wires far better than loose insulators, which slows contamination. It does not stop the loss of rhodium from the surface of the wire, because the couple still reaches the furnace temperature, and above about 1,100 °C rhodium migrates as a metallic vapour and the volatile oxides carry metal away. The published guidance for use in oxidising atmospheres restricts Types R and S to continuous operation up to about 1,300 °C for that reason.

The published guideline is a minimum bend radius of about five times the sheath outside diameter, and it is more conservative for thin sheaths and for repeated bending. Bending too sharply can crack the insulation, distort the wires or damage the sheath in a way that is not visible from outside, and repeated bending of the same section work-hardens the sheath. Bend away from the junction tip and the connector transition, and form the final shape before the probe goes into the installation.

Not normally, and published type data gives the reason: the output of the pair is negligible over the range from 0 °C to 50 °C, so the terminal temperature does not have to be corrected. Type S and Type R do not have that advantage, and both are wired with either an extension cable of the same alloys or a compensating cable matched to the type over a limited temperature range.

A platinum rhodium thermocouple is a noble-metal sensor, and the three types in use are separated by the rhodium content of their legs: Type S at 10 % rhodium, Type R at 13 %, and Type B with 30 % and 6 % rhodium on its two legs. The accuracy is published as a class on the electromotive force curve, at ±1.0 °C over the 0 to 1,100 °C range for Types S and R and ±0.5 °C for Type B over its range, and the class describes the couple and not the whole measurement, since the cold junction, the cable and the transmitter each add to the deviation. Type B has the practical advantage that its output is negligible between 0 °C and 50 °C, so cold junction compensation is not normally required.

Two mechanisms limit the life of a couple of this family, and they are grain growth in the limbs and the loss of rhodium from the surface, which deposits oxide on the platinum limb and moves the calibration. The armoured form answers the first and slows the second by keeping contaminants away from the wire, and it does not stop the drift, which is why the published continuous limit for Types R and S in oxidising atmospheres is about 1,300 °C even where the couple itself survives higher. Bending is limited to a radius of about five times the sheath diameter, the open end has to be sealed because the magnesium oxide is hygroscopic, and insulation resistance has to be read at the working temperature and not at room temperature. The couples and the armoured probes are supplied from Changzhou, Jiangsu in Type S, Type R, and Type B, in the sheath diameter, length, and junction form the installation calls for.

IEC 60584-1, thermocouples, reference tables and tolerance classes, with ASTM E230 and its limits of error, and the ASM Handbook treatment of thermocouple materials. Source for the leg compositions of the three noble-metal types, the accuracy class bands quoted for Types S, R, and B and for the base-metal types, the equivalence between the class and the special and standard limits of error, and the comparative behaviour of Type B in contaminated atmospheres.

Published type data for the noble-metal thermocouples, including the temperature limits of Type B and the statement that its output is negligible over the range from 0 °C to 50 °C so that cold junction compensation is not normally required, together with the note that the type is not normally used below 600 °C. Source for the range figures and for the cold junction position.

Bureau International des Poids et Mesures, Guide to Thermocouple Thermometry, Part 1. Source for the migration of rhodium as a metallic vapour above 1,100 °C, for the formation of volatile oxides of platinum and rhodium in the presence of oxygen at high temperature, and for the statement that Types R and S may be used continuously in oxidising atmospheres up to about 1,300 °C.

Pearce and co-workers, published work on the thermoelectric drift of platinum-rhodium thermocouples above 1,200 °C and on the mass loss of platinum-rhodium wires at 1,324 °C. Source for the two named failure modes of a Type R or Type S couple, which are excessive grain growth in the platinum limb causing wire breakage and temperature drift caused by the deposition of rhodium oxide on the platinum limb.

Published technical guides to mineral insulated thermocouples, together with the published technical paper on magnesium oxide insulated sensors. Source for the construction of the cable from the wires, the compacted insulation and the seamless sheath, for the use of magnesium oxide and of aluminium oxide, for the sheath materials named, for the protection against moisture, gases, and contamination, for the vibration and shock behaviour, for the bendability, the small diameters and the response time following the sheath diameter, for the working range to 2,400 °F and the range of 1,000 to 1,200 °C for MI thermocouples generally, for the minimum bend radius of about five times the sheath outside diameter and the work-hardening caused by repeated bending, for the hygroscopic behaviour of the insulation at an unsealed end, for the statement that its insulation resistance drops dramatically as the probe reaches operating temperature, and for the grounded, ungrounded, and exposed junction options.

Published guidance on extension and compensating cables, and product data held by DLX Alloy. Source for the distinction between an extension cable, which uses the same alloys as the couple, and a compensating cable, whose alloys differ and are matched to the output over a limited temperature range, for the bright annealed condition in which these grades are supplied, and for the types, sheath diameters, lengths, and junction forms supplied from Changzhou.

Every figure above is a published value or a published requirement for the material and the construction named, and the accuracy classes are the bands published in the standards and not measurements on a delivered sensor. Tolerance classes apply to the couple, and the measurement delivered at the instrument includes the cold junction, the cable and the transmitter. The limits on bending, moisture, and insulation resistance are properties of the construction, and the figure for the sheath and diameter ordered should be confirmed against the manufacturer datasheet. Temperature limits in air are set by drift for this family and not by the melting range of the alloys, which is why the published continuous figures are lower than the survival figures, and a probe chosen for a furnace duty should be specified against the drift limit and not against the maximum temperature on the type sheet.