Patent Application
20140307889
The described embodiments relate generally to methods and systems for eliminating or reducing acoustic noise in electronic systems, and more particularly to methods and systems for using an adaptive vibration damping mechanism for eliminating or reducing acoustic noise from a printed circuit board (PCB) caused by Multi-Layer Ceramic Capacitors (MLCC) or other similar components containing piezoelectric material.
Electronic systems commonly use Multi-Layer Ceramic Capacitors (MLCC) for such tasks as decoupling power supplies, or filtering signals. The ceramic material in the MLCC has a piezoelectric property which causes it to expand and contract in response to applied electric fields. This expansion and contraction can cause the components to vibrate. These components are very small, so the vibration of an individual part may not be significant. However, when there is an array of these parts vibrating synchronously, the effect is increased. Further, once the parts are fixed to a large, flexible substrate, as is the case when they are soldered down to a printed circuit board (PCB), the vibration is amplified further. What might have been a benign problem becomes a serious problem, particularly when the driving voltage varies at a frequency in the audible range. The problem may be manifested as a high pitched squealing noise coming from the product.
To combat this problem, system engineers and component engineers have focused on MLCC package modifications to minimize the coupling to the PCB, and on placement of the MLCCs, sometimes in pairs, to partially cancel or otherwise reduce the amplification of the vibration. Modifications to the package can lead to degraded performance of the capacitor when the mechanical changes add series impedance. Creative layout solutions only go so far in dampening the acoustic noise, and in many cases require compromises in electrical performance or space allocation to implement them.
Therefore, what is desired is a method or system to eliminate or greatly reduce acoustic noise in electronic systems caused by MLCC or other similar components containing piezoelectric material.
The approach described here is to sense the vibration caused by the excitation of the MLCCs (Multi-Layer Ceramic Capacitors) in response to receiving a varying input voltage, and then to drive a dedicated MLCC, or an array of MLCCs, that acts as a counter actuator to dampen or remove the vibration. In one embodiment, an array can be one or more MLCCs of one or more package sizes and of one or more capacitance values. In one embodiment, the approach can also be applied to other similar, but non-MLCC, components that contain piezoelectric materials. Since both the input voltage signal driving the MLCC and the transfer function that characterizes the conversion from voltage to sound pressure (the audible noise) are varying in time, the approach described here is an adaptive approach. This means that the damping signal is generated using an adaptive filter, which changes dynamically in response to the varying input signal and feedback signal. In this regard, the feedback signal is a proxy for the acoustic noise. The approach described here can involve three MLCC arrays positioned on a PCB. A first MLCC array generates a first vibration responsible for the acoustic noise in response to receiving a varying input voltage. A third MLCC array senses the first vibration, while a second MLCC array generates a second vibration to cancel out or reduce the first vibration. Therefore, the second and third MLCC arrays can be selectively positioned and oriented on the PCB to maximize sensing and cancellation of the first vibration. As an example, the second and third MLCC can be placed near the point of maximum flexure of the PCB.
In one embodiment, a system to eliminate an acoustic noise caused by a first MLCC (Multi-Layer Ceramic Capacitor) array positioned on a PCB (printed circuit board) is disclosed. The system includes a first MLCC array, a third MLCC array, an adaptive filter, and a second MLCC array. The first MLCC array is positioned on the PCB and configured to generate a first vibration in response to receiving a varying input voltage. The first vibration is causing the acoustic noise. The third MLCC array is positioned on the PCB and configured to sense the first vibration and generate a feedback signal. The adaptive filter is configured to use the varying input voltage and the feedback signal to generate an output signal. The second MLCC array is positioned on the PCB and configured to use the output signal to generate a second vibration that acts as a counter to dampen the first vibration. In one embodiment, the adaptive filter continually samples the varying input voltage and the feedback signal to generate the output signal that minimizes the acoustic noise. In one embodiment, the third MLCC array is positioned near a point of maximum flexure of the PCB. In one embodiment, the third MLCC array is oriented to measure all meaningful modes of the first vibration. In one embodiment, the third MLCC array includes more than one independent MLCC sensors, and the more than one independent MLCC sensors are placed at different locations, or in different orientations, or both. In one embodiment, the second MLCC array is positioned near a point of maximum flexure of the PCB. In one embodiment, the second MLCC array includes more than one independent MLCCs, and the more than one independent MLCCs are placed at different locations, or in different orientations, or both. In one embodiment, the second MLCC array is placed near the first MLCC array. In one embodiment, the third MLCC array is not a dedicated sensor capacitor array, and the third MLCC array is configured to perform other functions. In one embodiment, the second MLCC array is not dedicated to generating the second vibration that acts as a counter to dampen the first vibration, and the second MLCC array is configured to perform other functions.
In one embodiment, a system to eliminate an acoustic noise caused by a first electronic component containing piezoelectric material is disclosed. The system includes a first electronic component containing piezoelectric material, a third electronic component, an adaptive filter, and a second electronic component containing piezoelectric material. The first component is configured to generate a first vibration in response to receiving a varying input voltage. The first vibration is causing the acoustic noise. The third component is configured to sense the first vibration and generate a feedback signal. The adaptive filter is configured to use the varying input voltage and the feedback signal to generate an output signal. The second component is configured to use the output signal to generate a second vibration that acts as a counter to dampen the first vibration. In one embodiment, the adaptive filter continually samples the varying input voltage and the feedback signal to generate the output signal that minimizes the acoustic noise. In one embodiment, the third electronic component is selected from the group consisting of a strain gauge, a microphone, and an electronic device containing piezoelectric material. In one embodiment, the first electronic component is positioned on a PCB (printed circuit board). In one embodiment, the second electronic component is a part of the PCB (printed circuit board).
In one embodiment, a method to eliminate an acoustic noise caused by a first MLCC (Multi-Layer Ceramic Capacitor) positioned on a PCB (printed circuit board) is disclosed. The method includes sensing, with a third MLCC, a first vibration and generating a feedback signal. The first vibration is caused by an excitation of the first MLCC in response to receiving a varying input voltage. The first vibration is causing the acoustic noise. The method also includes generating, with an adaptive filter using the feedback signal, an output signal that is used by a second MLCC to generate a second vibration. The method further includes generating, with the second MLCC using the output signal, the second vibration that acts as a counter to dampen the first vibration. In one embodiment, the adaptive filter continually samples the varying input voltage and the feedback signal to generate the output signal that minimizes the acoustic noise. In one embodiment, the adaptive filter periodically samples the varying input voltage and the feedback signal at a fixed time interval. In one embodiment, the varying input voltage changes over time in frequency, phase, and amplitude. In one embodiment, the adaptive filter is a digital filter. In one embodiment, the adaptive filter is a Finite Impulse Response (FIR) linear digital filter. In other embodiments, the adaptive filter can be an Infinite Impulse Response (IIR) filter, an adaptive analog filter, a non-linear adaptive filter, or a Kalman filter.
The described embodiments and the advantages thereof may best be understood by reference to the following description taken in conjunction with the accompanying drawings. These drawings in no way limit any changes in form and detail that may be made to the described embodiments by one skilled in the art without departing from the spirit and scope of the described embodiments.
Representative applications of methods and apparatus according to the present application are described in this section. These examples are being provided solely to add context and aid in the understanding of the described embodiments. It will thus be apparent to one skilled in the art that the described embodiments may be practiced without some or all of these specific details. In other instances, well known process steps have not been described in detail in order to avoid unnecessarily obscuring the described embodiments. Other applications are possible, such that the following examples should not be taken as limiting.
In the following detailed description, references are made to the accompanying drawings, which form a part of the description and in which are shown, by way of illustration, specific embodiments in accordance with the described embodiments. Although these embodiments are described in sufficient detail to enable one skilled in the art to practice the described embodiments, it is understood that these examples are not limiting; such that other embodiments may be used, and changes may be made without departing from the spirit and scope of the described embodiments.
The approach described here is to sense the vibration caused by the excitation of the MLCCs, and then to drive a dedicated MLCC, or an array of MLCCs, that acts as a counter actuator to dampen or remove the vibration. In one embodiment, the vibration can be caused by the excitation of other similar, but non-MLCC, components. In another embodiment, other similar, but non-MLCC, components can act as a counter actuator to dampen or remove the vibration. The ceramic material in the MLCC has a piezoelectric property which is inverse-dual, meaning that a changing electric field can cause them to vibrate, and conversely, a physical vibration can cause an electric field to be generated in them. Therefore, in one embodiment, these similar, but non-MLCC, components can include piezoelectric materials. In one embodiment, MLCC sensors can be used for the feedback. In another embodiment, non-MLCC sensors (e.g. a strain gauge or a microphone) can be used for the feedback.
Since neither the characteristics of the driving signal (the excitation voltage on the MLCC), nor the transfer function from voltage to sound pressure (the audible noise) is fixed, but rather both vary with time, the approach described here is an adaptive approach. This means that the damping signal is generated with the aid of an adaptive filter which changes dynamically in response to the varying input signal and feedback signal, which is a proxy for the acoustic noise.
The functional electronic system 100 can have an array 112 of MLCC components configured to receive a varying input voltage Vin 130, whose variation is in the acoustic frequency range. In
In one embodiment, a large MLCC package, or MLCC array 132, can be used as a strain sensor. In
A complimentary array 122 of MLCC can be used to generate a second vibration that acts as a counter to dampen the first vibration associated with array 112 of MLCC. In FIGS. 1 and 2A-2E, array 122 is labeled as Array—2. The MLCC Array—2 and Array—3 can be equivalent packages (all 0603, for example), or they can be of different package sizes (0603 and 0805, for example). There can be advantages to either approach.
A useful property of piezoelectric materials is that they are inverse-dual, meaning that a changing electric field can cause them to vibrate, and conversely, a physical vibration can cause an electric field to be generated in them. Electronic system 100 takes advantage of this property with the MLCCs belonging to Array—3, which is also labeled as array 132. An adaptive system senses the voltage created across the Array—3 MLCCs as the flexing PCB puts strain on the MLCC packages. This voltage acts as a feedback signal to an adaptive control system.
In summary, both FIGS. 1 and 2A-2E illustrate three separate MLCC arrays. The first array, Array—1, includes functional MLCCs on PCB, which can cause vibration and acoustic noise in response to voltage changes in Vin. The third array, Array—3, includes MLCCs on PCB, which can be used to measure strain as the PCB flexes. The second array, Array—2, includes MLCCs on PCB, which can be used to flex the PCB in opposition to the vibration and acoustic noise of Array—1.
The adaptive vibration damping mechanism can be implemented in multiple ways. The signals can be analog, and the algorithm can be implemented as either analog or digital. Alternatively, the signals can be digitized, and the algorithm implemented digitally, with the output then being converted back to analog to drive the MLCCs. The following description assumes an all-digital implementation, with a digital filter for block W. This all-digital implementation was previously shown in
The adaptive filter, W, is a structure which adjusts its filter coefficients after every digital sample. The adjustment can be made based on any number of algorithms described in the literature: Least Mean Squares (LMS), or Recursive Least Squares (RLS), for example. There are also nonlinear adaptive filtering approaches which can be applied to this problem (the Volterra adaptive filter for example). The simplest embodiment to describe is the LMS filter, which is a gradient descent approach using an FIR (Finite Impulse Response) linear digital filter.
The FIR digital filter operates by first multiplying successively delayed input samples by weight values called coefficients (or tap weights). In one embodiment, the FIR digital filter can have 10 coefficients. These delayed and scaled samples are then summed to generate an output sample. The nature of the digital filter is determined by its impulse response, which is defined by these tap weights, and can be configured for low-pass, high-pass, etc., or can be configured for phase and magnitude adjustments of the input signals. In principle, if the ideal impulse response to use is known in advance, the vibration caused by an MLCC array can be perfectly canceled. When the required impulse response is not known in advance, an algorithm can be used to adaptively attain it.
As shown in
In
The resulting cancellation may not be perfect, because a nonlinear system (i.e., the transfer function P) is being modeled using a linear filter (i.e., the FIR filter of W). To improve the cancellation, one embodiment of the system can elect to use a more elaborate nonlinear adaptive filter and algorithm for W. However, a nonlinear implementation requires more cost, space, power, and computing resources, as compared to a linear implementation. As such, the simplest linear implementation might be the most attractive option.
In another embodiment, the cancellation can be improved by including an Array—2 and Array—3 for each significant vibration mode in the PCB (selected by physical orientation of the arrays). In the extreme case, there can be an independent W filter and algorithm operating on each mode array set. Alternatively, the sensor inputs can be merged into a single algorithm, which generates multiple outputs.
As shown in
In summary, this disclosure describes a method of actively driving an MLCC array to dampen the vibrations of a PCB in an electronic product. The features include:
The present implementation of an adaptive system is novel, because vibrations (i.e., physical forces) are “added” or “combined” to create the error signal. In particular, the error signal (i.e., strain signal 185 in
The foregoing description, for purposes of explanation, used specific nomenclature to provide a thorough understanding of the described embodiments. However, it will be apparent to one skilled in the art that the specific details are not required in order to practice the described embodiments. Thus, the foregoing descriptions of specific embodiments are presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the described embodiments to the precise forms disclosed. It will be apparent to one of ordinary skill in the art that many modifications and variations are possible in view of the above teachings.
