How to design a miniaturized, ultra-thin, high signal-to-noise ratio electret microphone (ECM)

At present, the market demand for microphone products with a diameter of less than 4mm and a thickness of less than 1.5mm is gradually increasing. While the demand for device miniaturization is increasing, the requirements for sensitivity and signal-to-noise ratio have increased. This has a great impact on the design and production of microphones, and it requires the optimization of various aspects of ECM design to achieve it.

The ECM signal-to-noise ratio consists of two aspects, sensitivity (corresponding to the signal size) and output noise (corresponding to the noise size). Sensitivity consists of two aspects: acoustic sensitivity and circuit gain. Among them, the acoustic sensitivity is related to the diaphragm area, diaphragm tension, plate spacing, polarization potential, and back acoustic cavity volume of the ECM; the circuit gain is related to the lead wire parasitic capacitance, amplifier gain, amplifier parasitic capacitance, etc. The output noise is also mainly composed of two aspects: acoustic noise and electrical noise: acoustic noise is related to the plate spacing, plate tension, plate area, and back acoustic cavity volume; electrical noise is related to amplifier gain and amplifier noise.

This article will mainly explore the impact on the ECM signal-to-noise ratio from two aspects: acoustic design and circuit design; and give specific design examples.

Acoustic Design Affecting ECM Signal-to-Noise Ratio

In microphone design, the design parameters related to acoustic characteristics include diaphragm area, diaphragm tension, plate spacing, polarization potential, back acoustic cavity volume, etc. Among them, in a specific microphone target product, when the gasket width is constant, the diaphragm area is basically fixed; in addition, for products with a certain height, the back acoustic volume is also basically fixed. In this way, in ECM design, the optimization and adjustment of the product can only be carried out mainly from the aspects of diaphragm tension, gasket thickness (plate spacing), polarization potential selection, etc.

In microphone design, the parameters of diaphragm tension, gasket thickness, and polarization potential affect and restrict each other. In order to obtain the optimal microphone design, it is necessary to make appropriate compromises within their selection range. To do this, it is first necessary to clearly understand the relationship between the various parameters.

Figure 1 shows the relationship between the maximum polarization potential of the diaphragm and the distance between the plates under the condition that the diaphragm does not attract. Ideally, the attraction voltage of the plates (the potential difference between the diaphragm and the back plate) is proportional to the 1.5th power of the distance between the plates. In order to ensure that the diaphragm does not attract during the processing and specific applications, it is necessary to ensure that the maximum potential on the diaphragm is less than 2/3 of the attraction voltage. After aging, the polarization potential will further decrease and tend to be stable.

Figure 1 The effect of the distance between the electrodes on the maximum polarization potential of the microphone

Figure 2 shows the effect of the back acoustic cavity volume on microphone noise. In miniaturized microphones, in general, when the back acoustic cavity is large (such as a 4015 microphone), the acoustic noise is much smaller than the circuit noise. Therefore, the acoustic noise accounts for only a very small part of the output noise, and the main output noise is dominated by the circuit noise. However, as the microphone becomes thinner and thinner, the back acoustic cavity volume decreases rapidly, and the proportion of acoustic noise in the microphone output noise also increases rapidly. For example, in a typical 3013 microphone, the back acoustic cavity volume is only 1/6 to 1/4 of that of a 4015 microphone.

Figure 2 Effect of back acoustic cavity volume on microphone noise

Figure 3 shows the relationship between diaphragm tension, resonant frequency and sensitivity. Generally speaking, the resonant frequency is inversely proportional to the diaphragm area and directly proportional to the square root of the diaphragm tension. The sensitivity is inversely proportional to the diaphragm tension. At the same time, when the diaphragm tension increases, the maximum polarization potential also increases.

Figure 3. The influence of diaphragm tension on the resonant frequency and microphone sensitivity

Figure 4 shows the relationship between the plate spacing and sensitivity and harmonic distortion. When the polarization potential of the microphone is the same, the sensitivity is inversely proportional to the plate spacing. However, when the plate spacing decreases, the second harmonic introduced by the nonlinear relationship between capacitance and sound pressure rises rapidly, resulting in a deterioration of the harmonic distortion characteristics.

Figure 4 Effect of diaphragm tension on resonant frequency and microphone sensitivity

Figure 5 shows the optimal range of the distance between the microphone plates when the polarization potential and the diaphragm tension are constant. It can be seen from the figure that, assuming that the polarization potential and the diaphragm tension are constant, the optimal design value of the distance between the plates with the maximum signal-to-noise ratio as the optimization target decreases as the size of the microphone decreases. If the polarization potential and the distance between the plates are assumed to be constant, the optimal design value of the diaphragm tension with the maximum signal-to-noise ratio as the optimization target increases as the size of the microphone decreases.

Figure 5 Optimal range of microphone plate spacing assuming constant polarization potential and diaphragm tension

Since the various acoustic design parameters of microphones affect each other and are limited by various aspects such as the size of the microphone, diaphragm material, reliability, cost, and mass production yield, in actual engineering production, obtaining an optimized design requires a lot of engineering practice and certain theoretical guidance. For a specific product, the possible range of variation of its acoustic parameters is very limited. Therefore, in the design of modern electret microphones, better electrical design is often used to obtain a larger acoustic optimization range, thereby obtaining better product performance.

Electrical design affecting ECM signal-to-noise ratio

Equivalent circuit of ECM

Figure 6. Equivalent circuit inside the microphone and external interface circuit

FIG. 6 shows the circuit equivalent diagram inside the microphone and the interface circuit outside the microphone consisting of the output load resistor RL and the output coupling capacitor Co.

The blue part marks the electrical parameters related to the microphone acoustic design. Among them, Ve is the polarization potential of the microphone, that is, the potential difference between the diaphragm and the back plate after the microphone has passed the polarization and aging process. The diaphragm capacitance is the capacitance between the diaphragm and the back plate. The structural parasitic capacitance Cps refers to the sum of the parasitic capacitance of the conductor part connected to the amplifier input pin, such as the back plate, copper ring, PCB lead, etc., to the ground (tube shell) in the structure of the microphone.

The light grey part marks the electrical parameters related to the microphone interface amplifier (such as JFET). Among them, Cpa is the input capacitance of the amplifier; Cc is the Miller parasitic capacitance, which is composed of the sum of two parts of parasitic capacitance: the parasitic capacitance between the input and output pins of the amplifier, and the capacitance between the conductor part of the microphone connected to the input pin of the amplifier and the conductor part on the output pin of the microphone; Gm is the equivalent transconductance of the amplifier.

Inside the microphone, two RF decoupling capacitors of 10pF and 33pF are often connected in parallel, so that better RF interference suppression characteristics can be obtained in terminals such as mobile phones.

Effect of parasitic capacitance in ECM

In general microphone design, the Miller parasitic capacitance Cc is small. At this time, the input parasitic capacitance of the amplifier and the structural parasitic capacitance have a greater impact on the microphone sensitivity. Assuming that the sound pressure signal input by the ECM causes the diaphragm to displace and causes the capacitance between the diaphragm and the back plate to change, the rms amplitude Vin of the voltage signal at the input pin of the amplifier is:

Figure 7 Effect of parasitic capacitance on microphone sensitivity

In a typical 4015 microphone, the input parasitic capacitance Cpa of the amplifier is about 3.5pF (TF202). At this time, Cpa has little effect on sensitivity. However, when the microphone size is reduced to 3015, due to the reduction of Cm and Cps, Cpa will cause an additional 3~4dB sensitivity drop. Therefore, in a 3015 microphone, the actual voltage gain of TF202 will be reduced from -2dB to about -6dB, resulting in a deterioration in the microphone sensitivity and signal-to-noise ratio. When the microphone size is further reduced to 2.5mm, Cpa will cause a 6~8dB sensitivity deterioration, making it completely unusable.

Even if you choose a microphone amplifier with a smaller input capacitance Cpa, such as the ACT503D electret microphone amplifier provided by Beijing Zhuorui Micro Technology Co., Ltd., whose Cpa is about 0.1pF, you still need to carefully consider the structural design and substrate design of the microphone to make full use of the excellent characteristics of the amplifier. For example, in a 4011 microphone using a copper ring contact, assuming that Cm is 4pF and Cps is 2.2pF, when using TF252, Cpa is 3.1pF. When switching from TF252 to ACT503D, the signal attenuation introduced by the parasitic capacitance of the circuit is improved by 20*log10((4+2.2+3.1)/(4+2.2+0.1)), which is about 3dB. However, if the copper ring contact is further changed to copper wire point contact and the substrate design is optimized, reducing Cps to 1pF, the sensitivity can be increased by about 2dB. Since the DC gain of ACT503D is 6dB, the sensitivity of the final microphone is nearly 10dB higher than that of TF252, which makes it easy to realize a high-sensitivity thin microphone product.

In miniaturized microphones, due to the small volume of the back acoustic cavity and the small diaphragm area, the acoustic sensitivity is low, and a higher gain microphone amplifier is required to obtain the appropriate microphone sensitivity. In such a microphone, the influence of Miller capacitance will be obvious, resulting in the gain attenuation of the amplifier. Since Miller capacitance is related to the Crss of JFET, the Crss of a typical JFET is around 0.7pF~1.1pF, so in miniaturized microphones, the use of high-gain JFETs is greatly limited. Similarly, the actual amplification factor of RS908/RS916 will drop sharply in small microphone applications. Since the ACT503D provided by Zhuorui Micro Technology adopts a ghost current output method, its equivalent Crss is less than 0.05pF, so it has a great advantage in high-gain voltage amplification.

Actual microphone test data

Figure 8 shows the sensitivity distribution of an ultra-thin 4mm microphone product using the ACT503D provided by Beijing Zhuorui Micro Technology Co., Ltd. As can be seen from the figure, the yield rate in the range of 40+-2dB reaches more than 88%. Table 1 shows the typical signal-to-noise ratio test results.

ACT503D sensitivity distribution in 4mm ultra-thin microphones

Table 1 Signal-to-noise ratio of ACT503D in 4mm ultra-thin microphone

Figure 9 shows the sensitivity distribution of a 3mm microphone product using ACT503D. As can be seen from the figure, the yield rate in the range of -42+-2dB reaches more than 80%. Table 2 shows the typical signal-to-noise ratio test results.

Figure 9: Sensitivity distribution of ACT503D in 3mm ultra-thin microphones

Table 2 Sensitivity and signal-to-noise ratio of ACT503D in 3mm microphone