The present invention relates to microphones, in particular MEMS microphones, with a moving membrane and a stationary backplate.
MEMS (micro-electromechanical systems) microphones are constructed using CMOS processes. However, when using such processes to create a mechanical moving membrane for the microphone, there are variables that are not controlled during the fabrication and assembly process. As such, the thickness of the gap between the movable microphone membrane and the stationary backplate varies between microphones made according to the same processes. This variation affects the performance and sensitivity of the microphones as well as the stability of the microphone.
In one embodiment, the invention provides a method for adjusting a bias voltage and gain of the microphone to account for variations in a thickness of a gap between a movable membrane and a stationary backplate in a MEMS microphone. The microphone is exposed to a first sound level and a sensitivity of the microphone is evaluated according to a predetermined sensitivity protocol. The bias voltage of the microphone is adjusted when the microphone does not meet the sensitivity protocol. The microphone is then exposed to a second sound level and a stability of the microphone is evaluated according to a predetermined stability protocol. The amplitude of the second sound level is greater than the amplitude of the first sound level. The channel gain of the microphone is adjusted when the microphone does not meet the stability protocol. In some embodiments, the bias voltage is also adjusted when the microphone does not meet the stability protocol and the microphone is again evaluated according to the predetermined sensitivity protocol and the stability protocol.
In some embodiments, the sensitivity of the microphone is evaluated by comparing the output signal of the microphone to a threshold. The threshold is a percentage (or a value indicative of a percentage) of the possible full scale output signal. In some embodiments, the stability of the microphone is evaluated by determining if the sensitivity of the microphone changes when the second sound level—a sound level with greater amplitude—is applied to the microphone.
In some embodiments, the bias voltage and the channel gain are adjusted using existing pads on the MEMS microphone. A power supply voltage to the MEMS microphone is increased and, in response, the MEMS microphone logic enters a trim mode. A serial binary signal is then provided to the MEMS microphone logic using a first pad. The MEMS microphone logic adjusts the bias voltage and the channel gain based on the serial binary signal. When the power supply voltage to the MEMS microphone is lowered to a normal operating level, the MEMS microphone logic exits the trim mode. When not operating in the trim mode, the MEMS microphone logic receives a second serial binary signal on the first pad and controls a second operation of the MEMS microphone based on the second serial binary signal. The second operation of the MEMS microphone is unrelated to adjusting the bias voltage or the channel gain of the MEMS microphone.
The invention also provides a MEMS microphone including a membrane that moves relative to the MEMS microphone in response to acoustic pressures applied to the microphone, a stationary backplate positioned a distance from the membrane, a bias voltage module applying a bias voltage on the membrane and the stationary backplate, and a trim module. The trim module is configured to evaluate a sensitivity of the MEMS microphone based on a digital output of the MEMS microphone when a first sound level is applied. The bias voltage of the MEMS microphone is adjusted when the sensitivity does not meet a defined sensitivity protocol. The trim module also evaluates a stability of the microphone based on a digital output of the MEMS microphone when a second sound level is applied. The second sound level has greater amplitude than the first sound level. The channel gain and the bias voltage applied to the movable membrane and the stationary backplate are adjusted when the stability does not meet a defined stability protocol.
Other aspects of the invention will become apparent by consideration of the detailed description and accompanying drawings.
Before any embodiments of the invention are explained in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the following drawings. The invention is capable of other embodiments and of being practiced or of being carried out in various ways.
During operation, acoustic waves cause the membrane 101 to move relative to the stationary backplate 109. As the membrane 101 moves, the thickness of the gap 111 changes. A bias voltage is applied to the membrane 101 and the backplate 109 so that changes in the thickness of the air gap 111 and, therefore, the distances between the membrane 101 and the backplate 109, cause a change in a capacitance measured between the membrane 101 and the backplate 109. This change in capacitance is monitored and used to generate a digital signal representing the sound wave.
Due in part to the small scale of a MEMS microphone system and the CMOS processes used to manufacture the MEMS microphone, there are physical variations between microphones manufactured by the same process. These variations include, for example, the thickness of the CMOS layers 105, the interface between various layers, and time-dependent etchings and release etchings in the various silicon layers. As a result, the air gap 111 often has a varying thickness even between microphones manufactured by the same process.
Because the digital signal representing the sound wave is directly related to the thickness of the air gap 111, process variations result in performance variations. A smaller distance between the membrane 101 and the backplate 109 results in a higher sensitivity. However, the smaller distance also makes a “snap in” effect more likely. The “snap in” effect is when an electrical force or acoustic pressure between the membrane 101 and the backplate 109 causes the membrane 101 to physically touch the backplate 109 and not return to its original position. With high sound pressure events (loud noise), the acoustic pressure applied to the membrane 101 is great enough to cause the membrane 101 to come too close to the backplate 109. Conversely, when the air gap 111 is too thick, the microphone is less susceptible to the “snap in” effect, but will also exhibit a lower sensitivity.
The signal measurement module 201 evaluates the digital signal and performs various tests to ensure that the performance of the microphone meets certain predefined requirements or protocols. The signal measurement module 201 communicates a signal to the trim module 207 indicating whether the microphone 205 meets the requirements. In response, the trim modules 207 adjusts the bias voltage provided by the bias voltage module 209 and the gain of the signal channel module 203 accordingly.
In some embodiments, as described in detail below, the time module 207, the bias voltage module 209, and the signal channel module 203 are implemented in the logic layer 105 of the MEMS microphone. The signal measurement module 201 is an external device that is connected to the output of the MEMS microphone and returns a trim code to the trim module after evaluating the signal. In other embodiments, the signal measurement module 201 is also implemented in the logic layer 105 so that the MEMS microphone does not need to be connected to other external equipment when the trim process is being performed.
The trim process is initiated (step 301) and a first sound level is applied to the microphone 205 by an external speaker (step 303). The first sound level is selected to test the sensitivity of the microphone 205. In some embodiments, the first sound level is from 94-104 dB and 1 KHz. The signal measurement module 201 evaluates the digital signal received from the microphone 205 and determines whether the microphone 205 meets a predefined sensitivity protocol (step 305).
The sensitivity of the microphone 205 is evaluated by comparing the output signal of the microphone to a threshold. The threshold is selected based on a percentage of a full, saturated signal. A signal is saturated when the magnitude of the signal is higher than the maximum signal amplitude that can be output by the system. For example, if the maximum output signal is 100 db, the system will output 100 db even if the signal should be 104 db or 110 db. In some embodiments, the threshold for evaluating the sensitivity of the microphone is set at 75% of the maximum output signal (also referred to as −25 db full-scale when the maximum output signal is 100 db).
If the output signal when the microphone 205 is exposed to the first sound level is less than the threshold, the microphone does not pass the sensitivity test. The signal measurement module 201 sends a trim code to the trim module 207, which then adjusts the bias voltage of the bias voltage module 209 accordingly (step 307). The microphone 205 is again exposed to the first sound (step 303) and the bias voltage is again adjusted (step 307) until the microphone 205 successfully passes the sensitivity test.
When the microphone 205 passes the sensitivity test, a second sound level is applied to the microphone 205 (step 309). The second sound level is selected to test the stability of the microphone 205 and has a greater amplitude than the first sound level. In some embodiments, the second sound level is 130-135 dB and 1 KHz. The signal measurement module 201 evaluates the digital signal received from the microphone 205 and determines whether the microphone 205 meets a predefined stability protocol (step 311). In some embodiments, the stability test evaluates the signal to determine if the sensitivity of the microphone changes in response to the higher amplitude sound. If the sensitivity has changed, the microphone does not pass the stability test.
If the microphone 205 does not pass the stability test, the trim module 207 adjusts the channel gain of the signal channel module 203 (step 313) and the bias voltage of the bias voltage module 209 (step 307). The microphone 205 must then again be exposed to the first sound level (step 303) to repeat the sensitivity test (step 305).
The sensitivity test (step 305) and the stability test (step 311) are repeated until the gain and the bias voltage are adjusted to levels where the microphone 205 successfully passes both the sensitivity test and the stability test. When the microphone 205 passes both tests, the trim process is complete (step 315). The microphone can then be packaged and shipped to consumers or installed in an end product.
If the membrane and the backplate are too close together, the microphone 205 will likely exceed the sensitivity threshold and pass the sensitivity test. However, the microphone 205 would then fail the stability test. The system 200 would decrease the bias voltage to reduce the likelihood of the “snap in” effect and increase the channel gain to account for losses in sensitivity caused by the lower bias voltage.
If the membrane and the backplate are too far apart, the microphone 205 will initially fail the sensitivity test. However, the bias voltage and, possibly the channel gain, will be increased to bring it within the acceptable range of both the sensitivity protocol and the stability protocol.
As described above, in some embodiments, the signal measurement module 201 is an external component temporarily connected to the microphone during the trim process. The signal measurement module evaluates the output signal from the microphone and sends a trim code to the trim module, which is implemented in the logic layer of the MEMS microphone. The trim module then adjusts the bias voltage or the channel gain based on the trim code., communication can be implemented through existing pads on the microphone system that serve other functions during normal operation of the microphone system.
To avoid the need for additional pads on the logic layer to communicate with the microphone logic, the signal measurement module transmits the trim code to the trim module through a pad that serves a different function during normal operation of microphone system. The power supply voltage provided to the microphone system is usually between 1 V and 3V. To enter the trim mode, this voltage is raised to a value above 3.5 V. When the power supply voltage has been raised, a serial binary trim code through a specific pad that usually serves another purpose unrelated to the trim method. The binary trim code is sent by toggling the input to the pad between two voltage levels. In other embodiments, the trim mode can be entered by mechanisms other than raising the power supply voltage.
In some embodiments, the trim code is three-digit binary number signaling new bias voltage and channel gain settings to be applied to the microphone based on the evaluation of the output signal performed by the signal measurement module. In other embodiments, the signal measurement module sends two separate trim codes to the trim module through the trim pad—the first indicating a new bias voltage setting and the second indicating a new channel gain setting. In still other embodiments, the trim code simply indicates whether the bias voltage or the channel gain should be increased, decreased, or left the same.
Thus, the invention provides, among other things, a system for adjusting a MEMS microphone to account for manufacturing variations. Various features and advantages of the invention are set forth in the following claims.