The magnetoresistance of 1/2 W, grade 1002, Speer carbon resistors with nominal room temperature resistances of 100, 200, and 470 Ω and of 1/8 W, 10 Ω, Allen-Bradley resistors was measured at various temperatures between 0.5 K and 4.2 K in static magnetic fields up to 140 kG. In this temperature range the maximum magnetic field induced temperature error of these thermometers amounts to 14%. The Allen-Bradley unit is to be preferred over the Speer resistors for use as a thermometer in a magnetic field for the following reasons: it has a more reproducible and regular magnetic field dependence, greater temperature sensitivity, and its much smaller physical size facilitates its incorporation in experimental apparatus where dimensions are restricted. One drawback of the 10 Ω Allen-Bradley resistor is its relatively high impedance at the lowest temperatures. In addition, the effects of magnetic fields up to 150 kG on the characteristics of Chromel P/Au+0.07 at.% Fe and Chromel P/constantan (type E) thermocouples were determined in the temperature range 4.2–45 K. A description is given of the measuring apparatus along with a discussion of possible measurement errors and the necessary precautions which must be observed. The magnetic field induced temperature errors for the Chromel P/Au+0.07 at.% Fe thermocouples are relatively large at temperatures below 20 K, and are nonreproducible between sections of gold-iron wire taken from the same lot. We do not recommend the use of these thermocouples below 20 K at fields greater than 50 kG. In contrast, Chromel P/constantan thermocouples have small, reproducible temperature errors up to 150 kG for 4.2≤T≤45 K. The data indicate that even in strong magnetic fields, at temperatures above 20 K, it is practical to correct for magnetic field induced temperature errors when using thermocouples. The reproducibility of the field induced errors in wires from different sources is still an open question and is worthy of further investigation.
Read full abstract