XIANYUAlloy · Changzhou
Technical note · 2026-10-26

Manganin or constantan at 60 °C ambient: pick on thermal EMF, not on TCR

At 60 °C ambient, use manganin for DC current sensing. Both alloys have a low temperature coefficient, so the ambient temperature itself decides very little. What separates them is the thermal EMF each generates against copper — roughly 1 µV/°C for manganin and roughly 40 µV/°C for constantan. On a 50 mV shunt that difference is larger than the drift you were trying to design out. Constantan becomes the right answer on AC, or when the element has to run above manganin's 140 °C ceiling.

The temperature coefficient is not where these two differ

Manganin is CuMn12Ni. Its temperature coefficient of resistance is specified at ±10 ppm/K between 20 and 50 °C, and its R(T) curve is a shallow parabola rather than a straight line. Constantan is CuNi45, with a coefficient of about ±40 ppm/K. Per degree, constantan looks four times worse, and it is. But both numbers are small enough that neither alloy fails a 0.5 % accuracy budget on temperature coefficient alone.

One honest qualification. Manganin's ±10 ppm/K is a guarantee for 20–50 °C. At 60 °C ambient the element is already outside that window, so you are relying on the typical curve, not on the datasheet limit. The curve is forgiving — across the whole 10–100 °C band the change in resistance stays inside 0.06 % — but it is a typical value. A supplier who quotes ±10 ppm/K and stops there has not told you what happens at your operating point.

Thermal EMF is the number that decides it

Every joint between the resistance element and a copper terminal is a thermocouple. If the two ends sit at different temperatures — and they do, because one is nearer the heat source — the pair generates a voltage that adds to the signal you are trying to read. Manganin produces about 1 µV/K against copper. Constantan produces about 40 µV/K, because constantan is a thermocouple alloy by design: it is the negative leg of Type T and Type J.

Put numbers on it. A 50 mV / 100 A shunt built from constantan, with a 5 K gradient between its two terminals, carries 40 × 5 = 200 µV of offset. That is 0.4 % of full scale — four times the error budget of a Class 0.5 shunt, produced by a temperature difference you cannot see. The same gradient on manganin gives about 10 µV, or 0.02 %.

This is not a subtle preference, it is the reason the two alloys exist separately. Reference texts on precision resistors describe constantan as suitable for AC resistors, where its thermal EMF is not a disadvantage, and manganin as the standard material for DC shunts. The distinction is thermal EMF. On resistivity the two alloys are within about 15 % of each other and behave almost identically in every other electrical respect.

What 60 °C ambient actually costs you

Ambient is not element temperature. The element self-heats, so 60 °C ambient may mean 75 or 90 °C at the blade. Manganin's maximum working temperature in air is 140 °C, above which an annealing process permanently shifts the resistance. Shunt makers work to tighter internal numbers: 80 °C as the point where resistance begins to drift, 125 °C as the practical maximum, 140 °C as the hard stop.

So at 60 °C ambient you have roughly 20 K of self-heating before you reach the 80 °C stability line. That is the real constraint in this application, and it is mechanical rather than metallurgical. The usual answer is derating: continuous current at no more than two thirds of the rated value, blades mounted vertically so convection works, and terminals sized to conduct heat out of the element instead of into it.

When constantan is the right call

Choose constantan when the measurement is AC. The thermal EMF shows up as a DC offset, an AC front end rejects it, and constantan's wider service range and lower cost then count for more than its thermal EMF does.

Choose it when the element genuinely runs hot. Constantan is quoted for continuous service between 350 and 400 °C depending on grade and atmosphere, well clear of manganin's 140 °C ceiling. If the shunt sits beside a contactor cycling at 120 °C, manganin is being asked to work at its limit and the drift is a maintenance item, not a fault.

Do not choose it because it is cheaper and the resistance reads the same on a bench test at room temperature with no current flowing. The failure mode only appears under load, when the terminals are at different temperatures, and it appears as a small constant offset that a zero-current calibration will never reveal.

What a wire mill can and cannot fix

Thermal EMF and the 140 °C ceiling are properties of the alloy family. Changing supplier will not change them. What a mill does control is that the grade is what the label says, that the resistance coefficient is measured on the actual batch rather than copied from a handbook, and that the certificate matches the spool.

The short version

  • DC current sensing at 60 °C ambient: manganin. Thermal EMF decides this before temperature coefficient does.
  • AC current sensing, or an element that runs above 140 °C: constantan.
  • 60 °C ambient leaves about 20 K of self-heating before manganin's 80 °C stability limit. Derate to two thirds of rated current.
  • Test the shunt under load, not at zero current. Thermal EMF offset is invisible without current flowing.
Questions we get asked
Is manganin or constantan more stable at 60 °C?
On temperature coefficient they are close. Manganin is specified at ±10 ppm/K between 20 and 50 °C and stays inside 0.06 % change across the wider 10–100 °C band. Constantan is about ±40 ppm/K but flat over a much wider range. At 60 °C ambient manganin is slightly tighter, though the margin is small.
Why is constantan not used for DC shunts?
Because of thermal EMF. Constantan generates roughly 40 µV/K against copper, against about 1 µV/K for manganin. A few degrees of gradient between the shunt terminals then adds an offset worth tenths of a percent of full scale, and no amount of gain calibration removes a voltage that varies with temperature.
What happens to manganin above 140 °C?
The resistance changes permanently. Above the maximum working temperature in air an annealing and oxidation process shifts both resistivity and temperature coefficient, and the shift does not reverse on cooling. Shunt manufacturers usually treat 125 °C as the practical maximum and 80 °C as the limit for holding a stable value.
Does 60 °C ambient damage a manganin shunt?
Not by itself. 60 °C is well inside the alloy's range. The problem is headroom: self-heating adds to ambient, and the element has to stay under roughly 80 °C to hold its value. At 60 °C ambient that leaves about 20 K, which is why high-current shunts in hot enclosures are derated and mounted for convection.
Citable facts
Sources

Send the specification — diameter, grade, temper, quantity.

We come back with a confirmed lead time and the documentation we can issue for the shipment. 100 g sample at cost; orders from 1 kg.

Request a quotation Order a 100 g sample