MEMS Microphone Source Follower Biasing for HD2 Linearization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Single-ended MEMS microphones suffer from prominent 2nd order harmonic distortion (HD2), particularly at low-mid input sound pressure levels, due to the inherent parasitic capacitance in the connection node between the MEMS device and the readout ASIC, which degrades system linearity.
Innovation Solution
An additional HD2 component is introduced on the ASIC, out of phase with the MEMS HD2 component to cancel it out, and a bulk node biasing mechanism is employed to optimize the source follower transistor configuration, allowing for selection of optimal bulk voltages through a multiplexer and look-up table to minimize distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a conventional source follower amplifier is used to amplify the MEMS output signal, then the circuit is simple and easy to implement, but prominent 2nd order harmonic distortion occurs at low-mid input sound pressure levels due to parasitic capacitance
Solution Approach 1:
An intermediate bulk node is introduced between the source follower transistor source and ground, serving as a mediator to apply a specific voltage that counteracts the parasitic capacitance effect. This bulk node acts as a control point to adjust the transistor's electrical characteristics and eliminate the harmful harmonic distortion while maintaining the simple source follower topology
Solution Approach 2:
The voltage at the bulk node is changed from the conventional zero voltage (direct ground connection) to a specifically optimized nonzero voltage value. This parameter change modifies the transistor's operating point and electrical characteristics, transforming the harmful parasitic capacitance effect into a beneficial linearizing effect that reduces harmonic distortion
2Device complexity
If the bulk node is directly connected to ground (zero voltage differential), then the circuit is simple, but the linear operation range is limited with a lower 1% THD cross-point
Solution Approach 1:
The bulk node voltage is made adjustable rather than fixed, allowing dynamic optimization for different operating conditions. A multiplexer is used to select from multiple predetermined bulk voltage values, enabling the system to adapt to different input sound pressure levels and extend the linear operation range while keeping the overall circuit structure relatively simple
3Reliability
If an optimized nonzero bulk voltage is applied to extend the linear operation range, then the 1% THD cross-point increases by 10-15 dB, but additional circuit components (multiplexer, buffer amplifier) are required
Solution Approach 1:
Multiple predetermined bulk voltage values are pre-calculated and stored in look-up tables during the design phase. The multiplexer simply selects from these pre-optimized values based on operating conditions, rather than requiring complex real-time calculation circuits. This preliminary preparation extends the linear operation range while minimizing the complexity of the actual implementation circuit
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A microphone includes a MEMS device having an output node for generating an analog voltage in response to an input sound wave or vibration; a source follower having a control node coupled to the output node of the MEMS device, a current path coupled between a first controlled node and a second controlled node, and a bulk node, wherein the first controlled node is configured for providing a microphone output voltage, and wherein the second controlled node is coupled to a reference voltage; a current source coupled to the first controlled node; and a voltage differential between the first controlled node and the bulk node, wherein a nonzero value of the voltage differential is configured such that a first 1% total harmonic distortion (THD) cross-point of the microphone output voltage is greater than a second 1% THD cross-point of the microphone output voltage using a zero value voltage differential.