As pointed out by Zheng et al.  the reduction of magnetic noise in the design of electric thermal demagnetizers is an important consideration. There are generally two types of noninductive heater windings, a single wire noninductively wound or a coiled single wire noninductively wound [Zheng et al., 2010, Figure 1]. In tests that they carried out Zheng et al.  claim that the magnetic noise generated along the axis of a standard MMTD80 thermal demagnetizer is more than 3 orders of magnitude larger than generated along the axis of their new Sogo Fine-TD thermal demagnetizer. If this were true then the new instrument would indeed represent a significant advance in thermal demagnetizer design.
2. Magnetic Noise Measurements
The MMTD80 heater element (which has been in use for almost 2 decades and is now used in the new high-resolution MMTD80A, http://www.magnetic-measurements.com/mmtd.html) was carefully designed to minimize magnetic noise. It is well known that the magnetic field generated by a toroid is contained within the toroid and the field generated by a uniform solenoid is contained within the length of the solenoid but escapes at the ends. The MMTD80 heating element is designed to incorporate some features of a solenoid and a toroid. The heating coil is a continuous solenoid even where it changes direction at the ends of the heater element (Figure 1). It is essentially made of solenoids that are connected by half toroids to prevent magnetic field leakage. This “elongated toroid” is not a perfect toroid and so we expect some flux leakage at the ends of the element but along the main straight sections of the heater element there should be negligible magnetic flux leakage (or magnetic noise).
Using the technique described by Zheng et al.  we passed a direct current of 1.0 A through the heater winding in order to detect magnetic flux leakage with a fluxgate magnetometer. This magnetometer can only measure constant magnetic fields so a direct current must be used. The measured magnetic field will be very similar in magnitude to the alternating magnetic field generated by a peak to peak current of 1.0 A. We measured the three components of the magnetic field along the full length of the heater element using a calibrated Bartington Mag 01 fluxgate magnetometer and double checked the measurements with a Bartington Mag 01H. As anticipated the axial magnetic noise in the sample region of the MMTD80 furnace is very low (Figure 2) but increases toward the ends of the heater element due to flux leakage as the element coil is bent through 180°. For the MMTD80 the RMS average axial magnetic noise in the hatched sample region for a current of 1.0 A (Figure 2) is 48.4 nT, much lower than the single measurement values quoted for the Sogo Fine-TD (213 nT) or the ASCTD48 (7180 nT, extrapolated from the 0.5 A value). For a current of 1.0 A the nonaxial magnetic field components (X and Y) are very small in the sample region (Figure 3), the RMS average X (horizontal) being only 2.7 nT with a maximum measured value of 6 nT and the RMS average Y (vertical) being 1.7 nT with a maximum measured value of 4 nT. These values are more than 2 orders of magnitude lower than the measurements made by Zheng et al.  using the same model MMTD80 thermal demagnetizer.
Zheng et al.  quote a very high axial magnetic noise figure for the MMTD80 furnace, 107,100 nT at 0.3 A, which is equivalent to a very large value of 375,000 nT for a current of 1.0 A. Since this is more than 4 orders of magnitude larger than those presented here it is worth considering how such a large field could possibly be generated inside the MMTD80.
The MMTD80 thermal demagnetizer is equipped with a calibrated solenoid which is designed to generate an internal magnetic field for paleointensity experiments. The calibration factor is normally about 170 nT/mA. If a current of 1.0 A was passed through this internal solenoid it would only produce a field of 170,000 nT, less than half the equivalent value reported by Zheng et al.  for a 1.0 A current passing through the heater element.
It is clear from Figure 2 that the data for the MMTD80 presented by Zheng et al.  are not correct and they should investigate the reason for this. If their data for the Sogo Fine-TD and ASCTD40 are correct then the measurements made on the MMTD80 and presented here show that with its elongated toroidal type heater element it has the lowest magnetic noise of the three thermal demagnetizers.