A comparison on WS2P microphone with both coupler comparison

and electrostatic actuator methods

Longbiao HE1; Feng NIU; Xiujuan FENG; Huan XU; Ping YANG
Department of Mechanics and Acoustics, National Institute of Metrology, China

ABSTRACT
To verify the calibration capability of different acoustic measurement laboratories in China, a comparison on
the sound pressure sensitivity at reference frequency and its frequency response of working standard (WS)
microphone was carried out from 20 Hz to 20 kHz. Seventeen laboratories were divided in two groups and
B&K 4192 microphone was used as the comparison sample. Both coupler comparison method and
electrostatic actuator method could be used in the comparison. Most of the results from the seventeen
laboratories agreed with each other. However, it also found some helpful information for some laboratories,
such as no cavity volume correction when using piston calibrators, improper couplers used in high frequency
calibration above 16 kHz. And extra error could be led in sequential comparison method without the monitor
microphone due to the volume variation during the measurement.
Keywords: Comparison, WS2P microphone, Coupler comparison method, Electrostatic actuator method
I-INCE Classification of Subjects Number(s): 72, 78

1. INTRODUCTION
Working standard microphone is the transfer standard in acoustic measurement. IEC 61094 Part 4,
Part 5, Part 6 and Part 8 describe the specification of working standard (WS) microphones, using the
coupler comparison method, the electrostatic actuator method and free-filed comparison method to
obtain their sensitivity level and frequency response respectively(1,2,3). In China, there are also some
technical specifications for calibration of working standard microphones, such as JJG 1019-2007 and
JJG 175-2015(4,5).
There are more than 30 provincial metrological research institutions in China. It is necessary to
verify the calibration capability of airborne sound pressure measurement, as well as the calibration of
WS microphones. So the comparison of WS2P microphone was carried out.

2. COMPARISON ARRAGMENT

National Institute of Metrology (NIM) is the pilot laboratory in the comparison with 7
laboratories participated. 7he transferring route chart of the comparison samples was shown in Figure
1. The 17 laboratories were divided into two groups. The samples were circulated to three
laboratories in each cycle, and then returned to NIM for stability checking up. The next cycle was
continued until all the participants completed in one group.
A1 A1 A1 A1

A3 A2 A5 A4 A7 A6 A9 A8

(a) Group A

1
helb@nim.ac.cn

1
6605
 B1 B1 B1 B1

B3 B2 B5 B4 B7 B6 B9 B8

(b) Group B

Figure 1 – Transferring route chart of the comparison samples

The coupler comparison method and electrostatic actuator method were recommended to obtain
the frequency response of WS2P microphone units (the preamplifier included) in one -third octave
band. For the electrostatic actuator method, the pressure sensitivity level at reference frequency 250
Hz could be got by pistonphone and its frequency response should be obtained by the correction
value provided by the manufacture, which is also verified by the pilot laboratory.
Table 1 shows the methods used by each laboratory. The pressure sensitivity at 250 Hz was
measured by pistonpohone or coupler comparison method. The frequency response was almost
measured by coupler comparison method and electrostatic actuator method. And there are two
laboratories using multi-function sound calibrators to get the frequency response.
Table 1 – Methods in Group A
Reference frequency point Frequency response Coupler or calibrator
Coupler comparison method
A1 Coupler comparison method WA 0817
Electrostatic actuator method
A2 Coupler comparison method Coupler comparison method WA 0817
A3 Coupler comparison method Electrostatic actuator method WA 0817
A4 Coupler comparison method Coupler comparison method WA 0817
A5 Pistonphone Electrostatic actuator method B&K 4228
A6 Pistonphone Electrostatic actuator method B&K 4228
A7 Coupler comparison method Coupler comparison method WA 0817
A8 Pistonphone Electrostatic actuator method B&K 4220
A9 Pistonphone Multi-function sound calibrator B&K 4226

Table 2 – Methods in Group B

Reference frequency point Frequency response Coupler or calibrator
Coupler comparison method
B1 Coupler comparison method WA 0817
Electrostatic actuator method
Pistonphone/ Coupler comparison method
B2 WA 0817
Coupler comparison method Electrostatic actuator method
B3 Pistonphone Coupler comparison method B&K 4226
B4 Pistonphone Electrostatic actuator method B&K 4228
B5 Pistonphone Electrostatic actuator method B&K 4228
B6 Pistonphone Electrostatic actuator method B&K 4228
B7 Coupler comparison method Coupler comparison method WA 0852
B8 Coupler comparison method Coupler comparison method WA 0817
B9 Pistonphone Coupler comparison method B&K 4228

3. COMPARISON RESLUTS
3.1 Stability of Transfer Samples
7o minimize the effect of sample stability on the evaluation of the comparison results, the samples
were monitored by NIM before and after each of the comparison cycle. The sensitivity stability at

6606
reference frequency of the samples before the comparison is less than ± 0.02 dB. The reference value
was considered as the average of five measurement results before comparison, after each cycle
completed and after the last cycle. The sensitivity stability of the transfer samples at 250 Hz is shown
in Table 3.
Table 3 – Sensitivity stability of transfer samples at 250 Hz
Reference value, dB Variation during the comparison, dB
A-1 37.56 -0.02̚+0.01
A-2 38.24 -0.03̚+0.06
B-1 -37.47 -0.09̚+0.06
B-2 -37.82 -0.04̚+0.04

The uncertainty of the reference value was evaluated as 0.10 dB (k=2). Here sample B-1 was
touched by fingers in the first comparison cycle and its sensitivity at 250 Hz was changed from
-37.56 dB to -37.46 dB. After further stability evaluation, it was confirmed to be used again.

3.2 Evaluation of Results

The comparison results was evaluated by value of En which is expressed by formula (1),
xX
En (1)
U 2
lab  U 2 ref

Where x is the measurement result of the participant, and X is the reference value, Ulab is the
uncertainty of the participant and Uref is the uncertainty of the reference. If the confidence level is
the same and |E n | is not larger than 1ˈ the results is satisfied and acceptable, otherwise not
acceptable.

3.3 Comparison Results

The comparison result of sensitivity level at 250 Hz was shown in Table 4, in which the data of
Laboratory A2 and A8 were obtained from re-measurement. It showed that all the sensitivity results
at the reference frequency agreed with each other. In Group B, the sensitivity results of some
laboratories were higher than the reference value, as shown in Figure 2.
The frequency response comparison results were shown in Figure 3. Here B2-1 means coupler
comparison method by laboratory B2 and B2-2 means electrostatic actuator method. Similar
frequency responses were obtained both in Group A and B, while some difference presents in high
frequency range especially above 10 kHz.

Table 4- Comparison Result of Sound Pressure Sensitivity Level at 250 Hz (|E n | value)
Lab No. Sample A-1 Sample A-2 Lab No. Sample B-1 Sample B-2
A2 0.36 0.22 B2 0.27 0.38
A3 0.64 0.92 B3 0.54 0.44
A4 0.01 0.23 B4 0.74 0.95
A5 0.61 0.67 B5 0.78 0.50
A6 0.17 0.41 B6 0.83 0.79
A7 0.20 0.37 B7 0.20 0.81
A8 0.33 0.04 B8 0.25 0.08
A9 0.47 0.67 B9 0.68 0.87

6607
 (a) Sample A-1 (b) Sample A-2

(c) Sample B-1 (d) Sample B-2

Figure 2- Comparison results of sound pressure sensitivity level at 250 Hz

(a) Sample A-1 (b) Sample A-2

6608
 (c) Sample B-1 (d) Sample B-2
Figure 3- Comparison results of frequency response

4. DISCUSSION
4.1 Sensitivity level at reference frequency
The measurement of the sound pressure sensitivity level at reference frequency was obtained by
coupler comparison or acoustic calibrator method. If the coupler comparison method is used, it is
necessary to consider the insertion loss of the preamplifier for reference microphone. When LS-type
sound calibrator was chosen, such as the B&K 4228 pistonphone, the actual sound pressure level
inside the calibrator is equal to the sum of the nominal sound pressure level, static pressure
correction, and cavity volume correction. Static pressure correction was usually considered, while
the cavity volume correction will result in a sensitivity deviation of 0.08 dB if it was ignored
(Removing the protection grid of B&K 4192 and putting on an adapter, the cavity volume could be
considered as the same to B&K 4180). As shown in Figure 2 (c) and (d), a considerable portion of
the laboratories used the calibrator method to provide the sound pressure sensitivity levels, which
were about 0.1 dB higher than the reference value at 250 Hz. It was verified that some laboratories
indeed ignored the cavity volume correction of the calibrator.

4.2 Frequency response at high frequency

The frequency response measurement method included electrostatic actuator method and
coupler comparison method. In the latter method, different couplers was used, such as B&K WA0817
and B&K 4226 etc. When using sequential comparison method with multi-function sound calibrator,
larger uncertainty was introduced at frequency higher than 10 kHz, as the comparison result of
laboratory A9 and B3 shown in Figure 3.
In contrast, for the electrostatic actuator method, the frequency response at high frequency was
more consistent. When the frequency response of the microphone unit is obtained by the coupler
WA0817, it had poor repeatability at high frequency. Figure 4 showed repetitive results of the
microphone unit A-1 with 20 times measurement using WA 0817 compared with a LS microphone.
The repeatability standard deviation is 0.25 dB at 20 kHz without excluding the outliers and 0.08 dB
at 16 kHz. Here relative frequency response at 20 kHz was -1.02 dB. The relative frequency response
obtained by electrostatic actuator method was -1.05 dB, and the difference was only 0.03 dB. While
the comparison result of laboratory B2 showed that the repeatability of coupler comparison method
achieved 0.71 dB at 20 kHz with six times measurements.
For coupler comparison method at high frequency, measurements without enough times could
introduce larger uncertainty component, due to the difference between the coupling conditions and
the slight difference in the symmetry of the test microphone and the reference microphone.

6609
 Figure 4- Twenty times measurements of coupler comparison method

4.3 The overall frequency response of microphone

Capacitive microphones have typical frequency response characteristics and the frequency
response curve of the microphone unit should be a smooth curve. In sequential comparison method,
sound leakage could lead some difference in low frequency, while in the high frequency range, the
difference of cavity volume and the asymmetry of the microphone diaphragms, relative to the radial
sound source, could lead more fluctuations. These problems should be investigated from the overall
curve characteristics of the microphone unit. Big fluctuations like the frequency response of
laboratory A8, A9 and B8 should be avoided.

5. CONCLUSIONS
The capability of calibration laboratories was verified by the comparison of the microphone
sensitivity level and its frequency response. It showed good agreement with coupler comparison
method and electrostatic actuator method. For coupler comparison method, the radial sound source
with small cavity size is preferred and simultaneous comparison is better than sequential comparison
especially at high frequency. For electrostatic actuator method, the corrections from electrostatic
actuator response to the sound field response of individual microphone should be known. If typical
correction value is used, the consistency of the microphones in the same type is required to be at
high degree. And the difference in radiation impedance of electrostatic actuators should also be
considered. The typical frequency response characteristics of the capacitive microphones could also
be used to inspect the frequency response measurements.

ACKNOWLEDGEMENTS
Comparison work was supported by civil comparison project from General Administration of
Quality Supervision, Inspection and Quarantine, China. We also acknowledge gratefully for the
supports from National Natural Science Foundation of China (No. 51575502).

REFERENCES
1. International Electrotechnical Commission, Measurement Microphones - Part 4: Specifications for
Working Standard Microphones, IEC 1094-4:1995.
2. International Electrotechnical Commission, Measurement microphones - Part 5: Methods for pressure
calibration of working standard microphones by comparison, IEC 61094-5:2001.
3. International Electrotechnical Commission, Measurement microphones - Part 6: Electrostatic actuators
for determination of frequency response, IEC 61094-5:2004.
4. National Acoustics Metrology technical commission, Verification regulation for working standard
microphones (Coupler comparison method), JJG 1019-2007.
5. National Acoustics Metrology technical commission, Verification regulation for working standard
microphones (Electrostatic actuator method), JJG 175-2015.

6610