Specifications

B-field Exposure From Induction Cooking Appliances 31
4.3 Uncertainty Budget
Table 10 gives a list of probable sources of error and an evaluation of the uncertainty budget,
as given in the IEEE 1528-2003 standard [9]. The uncertainty was assessed in the region of
the test specification limit (6.25 µT) and for the full range of measurements. Moreover, during
all experiments, the random components of uncertainty were reduced by performing repeated
measurements (averaging N = 4, random components of uncertainty divided by
√
N = 2). The
maximum uncertainty limits have been evaluated as follows [7, 8, 9]:
• Narda probe ELT-400 measurement uncertainty. The specification given by the
manufacturer is ±4 % or ±0.34 dB for a signal in the frequency range 50 Hz to 120 kHz.
• Error due to non-perfect frequency response of the probe. The Narda probe
ELT-400 frequency response deviation is less than ±0.80 dB or ±9.6 %.
• Error due to non-perfect spherical isotropy of the probe. As described in Chapter
4.1, the Narda probe ELT-400 spherical isotropy is very good, and the deviation is less
than ±0.2 dB or ±2.3 %.
• Error due to non-perfect linearity of the probe. The deviation from linear response
is < ±0.05 dB for B-fields exceeding 0.3 µT (see Chapter 4.1). The error increases for lower
fields and equals 2.80 dB for B-fields of 0.06 µT. However, most of the B-fields measured
exceeded 0.3 µT, and the uncertainty ±0.05 dB or ±0.6 % is considered.
• Centering of the pot on the hob. As discussed in Chapter 6.0.6, for pot displacements
of ±4 mm (in every direction relative to the hob center) a maximum error of 11 % or
±0.90 dB is introduced.
• Water volume and evaporation. As shown in Chapter 6.0.7, the B-field variation due
to changes in the volume of water in the pot is smaller than ±2.45 % or ±0.21 dB.
• Probe alignment with respect to the appliance edge. The probe alignment with
respect to the appliance edge was verified before for each measurement with a worst-case
accuracy of ±4 mm. It was calculated (Figures 24 to 29) that the corresponding B-field
variations are smaller than ±2.2 % or ±0.19 dB when measuring in front of each hob.
• Probe positioning on the DASY4 robot arm. The mount made especially to fix the
Narda probe ELT-400 on the robot arm has a notch which very precisely defines the probe
position (±0.5 mm). This error was considered to be negligible.
• Robot arm positioning. The DASY4 robot arm positioning and repeatability is better
than ±1 mm. The corresponding measurement uncertainty was considered to be negligible.
• Variation in the mains power supply. The transfer function between the mains power
supply and B-field variations was not assessed and assumed to be negligible.