6J12 is the tighter grade and 6J13 is the wider window
Both are copper-manganese-nickel resistance alloys and both are sold as manganin, but they are not interchangeable on a shunt. One fixed a narrow coefficient in three separate classes over a narrow temperature window; the other fixed a single wider band over a wider one. Choose by where the element sits when the current is on, not by which number looks better on a quotation.
What actually differs in the two specifications
Both grades are 11 to 13 % manganese with copper as the balance. 6J12 caps nickel at 3.0 % and 6J13 allows up to 5.0 %, which is why two grades that look identical on a purchase order come back with different electrical values.
6J12 is specified at 0.47 µΩ·m. 6J13 is specified at 0.44 µΩ·m. At the same wire diameter that is a 6.8 % difference in resistance per metre, so a design trimmed around one grade will not land on the same value with the other. The element length has to be recalculated; substituting the wire alone moves the shunt out of tolerance.
GB/T 6145 calls 6J13 F2 manganin, and in Chinese mill catalogues F2 is the shunt grade while plain 6J12 is the standard-resistor grade. That naming is a fair summary of the two specifications, and it is the faster way to remember which is which.
The class number is the part buyers leave off the order
The 6J12 coefficient is not a single number. GB/T 6145 splits it into three classes over 5-45 °C: −3 to +5 ppm/K, −5 to +10 ppm/K and −10 to +20 ppm/K, with class 1 the tightest. An order that says 6J12 and stops there does not say which class the mill will ship, and the top of class 3 is nearly four times further from zero than the top of class 1.
6J13 works the other way round: one band, 0 to +40 ppm/K, guaranteed over 10-80 °C. There is nothing to grade, but the band is wider than any 6J12 class and it is one-sided, so the error at a hot element always runs in the same direction instead of cancelling.
Why a shunt often wants the wider window
A shunt at rated current is not at room temperature. The element sits somewhere between 60 and 90 °C in a busbar mount with no forced air, which is outside the 5-45 °C qualification of 6J12 and inside the 10-80 °C qualification of 6J13. That one fact decides most grade choices: if the element runs hot, the tighter room-temperature band buys nothing, because the guarantee does not cover the temperature the part actually reaches.
The ceiling above both grades is set by oxidation rather than by the coefficient. Isabellenhütte publishes +140 °C as the maximum for MANGANIN and advises keeping precision resistors below +60 °C. Past +140 °C the alloy anneals and the resistance change stops being reversible.
Thermal EMF, where the grades differ for free
6J12 carries 1 µV/K against copper over 0 to 100 °C, and 6J13 carries 2 µV/K. On a 75 mV shunt read at full scale, the 1 µV/K gap between them becomes 15 µV at a 15 K gradient across the element, which is 0.02 % of reading and irrelevant. Read the same shunt at 20 % of full scale and the same 15 µV is 0.1 % of reading, which is now larger than the coefficient error the tighter grade was supposed to remove.
That is the honest trade: the extra microvolt per kelvin of 6J13 is worth almost nothing on a full-scale reading and worth a lot on a small one, so measure at the fraction of scale you actually use before deciding that 6J12 is the safer purchase.
When neither grade is the right answer
Above +140 °C the question stops being about 6J12 or 6J13. Manganin anneals and its resistance moves permanently, so a genuinely hot element needs a different family. Constantan tolerates far higher temperatures, but it brings roughly 40 µV/K against copper, which rules it out for low-level DC measurement and is fine for AC or for a heater.
Grade plus class, not grade alone; diameter with the tolerance band; temper; and whether the certificate must state the measured coefficient and the temperature range it was measured over. A certificate quoting a coefficient without its range cannot be checked against the class you asked for, and that omission is the usual reason two certificates for the same grade disagree.
XIANYU ALLOY draws 6J12 and 6J13 manganin wire from 0.05 mm to 6.00 mm in soft, half-hard and hard temper, in orders from 1 kg with a 3-7 day lead time, and issues a material test certificate with every batch.
Is 6J13 a lower grade of manganin than 6J12?
Can I substitute 6J13 for 6J12 without changing the element design?
Which class of 6J12 should I ask for?
Does 6J13 drift more than 6J12 at room temperature?
- 6J12 manganin is specified at 0.47 µΩ·m under GB/T 6145, with 11 to 13 % manganese and 2.0 to 3.0 % nickel.
- 6J13 manganin is specified at 0.44 µΩ·m and is the grade Chinese mills designate F2.
- The 6J12 temperature coefficient is quoted in three classes over 5-45 °C: −3 to +5, −5 to +10 and −10 to +20 ppm/K.
- 6J13 is quoted with a single coefficient band of 0 to +40 ppm/K over 10-80 °C.
- 6J12 carries 1 µV/K against copper over 0 to 100 °C, and 6J13 carries 2 µV/K.
- MANGANIN's coefficient is ±10 ppm/K between 20 and 50 °C, with a published ceiling of +140 °C.
- Isabellenhütte advises keeping precision manganin resistors below +60 °C and states that exceeding +140 °C causes permanent resistance drift.
- XIANYU ALLOY draws 6J12 and 6J13 manganin wire from 0.05 mm to 6.00 mm and issues a material test certificate with every batch.
- Isabellenhütte MANGANIN data sheet (CuMn12Ni, 43 µΩ·cm, ±10 ppm/K between 20 and 50 °C, +140 °C ceiling in air, precision resistors below +60 °C, −1.00 µV/K against copper)
- GB/T 6145-2010 6J12 entry with the three coefficient classes over 5-45 °C, 0.47 µΩ·m resistivity and 1 µV/°C against copper
- GB/T 6145-2010 6J13 entry, designated F2 manganin: 0.44 µΩ·m, 0 to +40 ppm/K over 10-80 °C, 2 µV/°C against copper
- Manganin and constantan resistance alloy comparison table: 6J12 and 6J13 chemistry, resistivity, coefficient, thermal EMF and working temperature side by side
- 6J12 alloy properties under GB/T 6145-2010, class 1 to class 3 coefficient definitions and the 5-45 °C measurement window
- 6J13 manganin product data: 0.44 µΩ·m, coefficient 0 to +40, thermal EMF against copper up to 2 µV/°C, GB/T 6145-2010
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