MEMS SPEAKER
Technical Field
[001] The present invention relates to the field of MEMS
acoustic transducers, more specifically to MEMS microspeakers, with a device silicon layer having a bi-directional vibrating region that is actuated by a peripheral excitation zone.
Background of the invention
[002] Microspeakers with high acoustic output and power
efficiency are important requirements for the rapidly growing mobile device sector. When used, the conventional microspeakers (based on the voice coil) consume up to a quarter of the total energy in mobile phones. Further, the acoustic output from these microspeakers is very poor in quality as well as loudness compared to the conventional speakers. The size of these microspeakers is also big (around 15 mm) compared to the other components present in mobile phones.
[003] When the microspeakers based on electromagnetic
actuation are scaled down to make miniature speakers used in mobile phones and similar other applications, the power required to drive the speaker does not scale favourably because the electromagnetic force is scale invariant unlike electrostatic force that scales favourably (i.e., the force increases when the size of the device shrinks). This fact makes electrostatic actuation more favourable for miniaturized actuators. In other words, presently, in field of micro-electromechanical systems (MEMS), the manufacture of micro-speakers is generally based on the central electromagnetic actuation of a conical membrane attached to the

actuator. However, one of the significant areas of concern in designing MEMS speakers with a central electrostatic actuation is the need for small separation between the top and bottom electrodes for efficient driving.
[004] For MEMS microspeakers, with vibrating structures, a
constraint that limits the operation of the device is the maximum amplitude of oscillation, which is governed by the gap between the vibrating structure and the bottom substrate. In devices, where electric potential is applied between the vibrating structure and the bottom substrate, this limit is even much lower. Due to the nature of electrostatic force, the microspeakers become unstable as the amplitude of oscillation of the vibrating structure exceeds one third of the gap and the microspeakers gets pulled-in (called dynamic pull-in phenomenon) and sticks to the bottom substrate due to forces of adhesion. Since, the acoustic pressure radiated from the vibrating structure is directly proportional to the velocity of vibration, the maximum acoustic pressure that the device can radiate is limited by the gap between the structure and the substrate. But, one cannot also have a very large gap, since the actuation force is inversely related to square of the gap.
[005] The electrostatic actuation based MEMS structures
are suspended over a fixed substrate with a small gap. When such structures vibrate, the air in the gap gets squeezed and contributes to the added damping and stiffness of the structure. This added damping and stiffness are functions of the frequency of vibration and hence affect the dynamic response of MEMS devices over a range of frequencies. Hence, it is necessary for

the successful realization of MEMS devices to be able to account for this added damping and stiffness in their design. Squeeze film damping is a common phenomenon having its influence on the MEMS structures.
Objects of the present invention
[006] The primary object of the present invention is to
provide a MEMS speaker with a peripheral electrostatic actuation.
[007] An object of the present invention is to provide a
MEMS speaker with a peripheral electrostatic actuationand a device silicon layer with a bi-directional vibrating region.
[008] Another object of the present invention is to provide
a MEMS speaker with a peripheral electrostatic actuation to prevent the influence of squeeze film damping effect on the MEMS speaker.
[009] Yet another object of the present invention is to
provide a method for fabricating MEMS speaker with a peripheral electrostatic actuation.
Summary of the invention
[010] Accordingly, the present invention provides a MEMS
speaker 100, comprising the handle silicon layer 101 with the primary amplitude enhancer. The insulator 102 with the extended amplitude enhancer 108, is arranged on the handle silicon layer 101. The peripheral excitation zone 109 is disposed in the insulator 102 and positioned adjacent to the extended amplitude enhancer 108. The device silicon layer 103 is disposed to overlay the insulator 102 and the peripheral

excitation zone 109. The bi-directional vibrating region 104, is formed in the device silicon layer 103 and positioned to overlay the primary amplitude enhancer 106 and the extended amplitude enhancer 108. The electrical source 105 is connected to the bi-directional vibrating region 104 through the handle silicon layer 101 and the device silicon layer 103. The present invention also provides a method of fabrication of the MEMS speaker.
Brief description of the drawings
[011] FIG.1 is an exemplary perspective view of MEMS
speaker of the present invention.
[012] FIG.2 is an exemplary exploded perspective view of
MEMS speaker of the present invention.
[013] FIG.3 is an exemplary partial sectional view of MEMS
speaker as shown in FIG.1.
[014] FIG.4 is an exemplary cross-sectional view of MEMS
speaker depicting a primary amplitude enhancer, an extended amplitude enhancer and a peripheral excitation zone.
[015] FIG.5 is an exemplary cross-sectional view of MEMS
speaker depicting a peripheral excitation zone with a curved rectangle profile.
[016] FIG.6 is an exemplary cross-sectional view of MEMS
speaker depicting two primary amplitude enhancers, an extended amplitude enhancer and a peripheral excitation zone.

[017] FIG.7 is an exemplary top view of of MEMS speaker
depicting a bi-directional vibrating region with a square profile.
[018] FIG.8 is an exemplary cross sectional view of MEMS
speaker depicting vibration of the bi-directional vibrating region of device silicon layer.
[019] FIG.9 is a schematic depiction of propagation of
Omni-directional sound from the MEMS speaker of the present invention.
[020] FIG.10(a) is an optical image of the MEMS speaker
of the present invention.
[021] FIG.10(b) is a depiction of an experimental shape of
the MEMS speaker of the present invention.
[022] FIG.11 is a graphical plot showing the displacement
response (centre) and the pressure radiated (at 10 cm) from a single MEMS device of the present invention as a function of actuation frequency.
[023] FIG.12 is a graphical plot showing the measured
radiated pressure from a single MEMS speaker of the present invention and the pressure computed based on the displacement response.
[024] FIG.13 is a schematic depiction of the broad steps of
micro fabrication of the MEMS speaker of the present invention.

Detailed description of the invention
[025] The embodiments of the invention will now be
described in further detail. It is to be understood that limiting the description to the preferred exemplary of the invention is to only facilitate description of the present invention and it is envisaged that those skilled in the art may incorporate modifications and equivalents, without departing from the scope of the defined claims.
[026] The embodiments of the MEMS speaker, which is a
reverse transducer, will now be described, with reference to FIGs.1-9, either individually or in combination thereof.
[027] FIG.1 illustrates a broad structural elements of the
microspeaker 100 that is provided with a handle silicon layer 101, to act as a base for incorporating microelectromechanical functional elements of the microspeaker 100. Functional elements include bi-directional vibrating region 104 and peripheral excitation zone 109. An insulator 102 that is adhered to the handle silicon layer 101 to provide a desired electrical insulation to the other elements of the MEMS speaker 100 VIS-À-VIS the handle silicon layer 101. A device silicon layer 103 is adhered to the insulator 102. A bi-directional vibrating region 104 is formed on the device silicon layer 103. An electrical source 105, is connected to the device silicon layer 103 and the handle silicon layer 101, so that these layers act as electrodes to provide an electric potential between the handle silicon layer 101 and the device silicon layer 103. The bi-directional vibrating region 104 is arranged to receive electric potential from the electric source 105.

[028] The handle silicon layer 101 is advantageously a
mono-crystalline or a single crystal silicon. The diameter of the handle silicon layer 101 is preferably in the range of 1 to 18 inches and whereas the thickness of the handle silicon layer 101 is in the range of 100 to 800 µm. The inherent electrical resistivity of the handle silicon layer 101 is preferably less than 0.01-0.001 ohms-cm. The electronic properties of the handle silicon layer 101 is modified with the addition of a suitable dopant of the type n-type or p-type. Exemplary p-type dopants include boron, aluminium, nitrogen, gallium, and indium and n-type dopants include phosphorus, arsenic, antimony, bismuth, and lithium.
[029] The primary amplitude enhancer 106 is formed in
the handle silicon layer 101, as particularly shown in FIGs.2,3 and 4. The primary amplitude enhancer 106 is a cavity or groove that is formed by etching of the handle silicon layer 101 to form an inner periphery 107. The primary amplitude enhancer 106 is formed in the inner periphery 107 of thehandle silicon layer 101, as particularly shown in FIG.3. The primary amplitude enhancer106 extends through the handle silicon layer 101 and the inner periphery 107 defines the extent of the primary amplitude enhancer 106. In other words, the volume of the primary amplitude enhancer 106 is reciprocal to the thickness of thehandle silicon layer 101. In this exemplary aspect, the profile of the primary amplitude enhancer 106 is exemplarily shown as a circular cavity with an anisotropic profile. It is understood here thatthe primary amplitude enhancer 106 can also be suitably formed with other non-limiting profiles with elliptical or non-circular cross-sections. It is to be noted here

that size and the profile of primary amplitude enhancer 106, defines the operating frequency of speakers.
[030] The insulator 102, which is preferably a silicon
dioxide layer is adhered to the handle silicon layer 101, as shown in FIG.1. Thickness of the insulation layer 102 is preferably in the range of 250 nm to 2 µm. However, a suitable thickness for the insulation layer 102 can be selected based on the required excitation voltages and the desired amplitude of the vibrating portion 104. The electrical resistivity of the insulation layer 102 is preferably in the range of 1017 ohm-cm. The insulator 102 is advantageously formed by SIMOX (separation by implantation of oxygen) process or by Smart Cut processes. The insulator 102 provides the desired electrical isolation(insulation) between the handle silicon layer 101 and the device silicon layer 103, which act as electrodes. Since, the drive electrostatic force is inversely proportional to the square of thickness of the insulation layer 102, the high breakdown voltage of the insulator 102 is about 0.5 GV/m, which is advantageous for high amplitude sound production applications.
[031] As particularly shown in FIGs.2, 3 and 4, an
extended amplitude enhancer 108 is formed by further etching of the insulator 102 by forming an inner periphery 110a. The extendedamplitude enhancer 108 is co-axial to the primary amplitude enhancer 106 and extends from the base portion of the insulator 102 on one side andto the device silicon layer 103 on the other side. The extendedamplitude enhancer 108 is a circular cavity having a reciprocal profile as that of the primary amplitude enhancer 106. Therefore, the primary amplitude

enhancing portion 106 and the extendedamplitude enhancer 108 are contiguously and seamlessly connected to each other to present an integrated hollow structure, as particularly shown in FIG.4.
[032] A peripheral excitation zone 109 is formed adjacent
to the extended amplitude enhancer 108, is positioned coaxial to the primary amplitude enhancer 106, as particularly shown in FIG.4. The peripheral excitation zone 109 is formed preferably lateral etching of the insulator 102, so that the peripheral excitation zone 109 extends from the inner periphery 110b of the insulator 102 towards an outer periphery 110a, as particularly shown in FIG.4. The peripheral excitation zone 109 is positioned coaxial to the primary amplitude enhancer 106. The peripheral excitation zone 109 forms a gap between the handle silicon layer 101 and the device silicon layer 103 as shown in FIG.4. During the course of operation of the MEMS speaker 100, the peripheral excitation zone 109 is filled with air so that it acts like a dielectric medium. Accordingly, the peripheral excitation zone 109 renders the desired electrostatic actuation to the vibration zone 104 of the device silicon layer 103,on the application of electric potential to the handle silicon layer 101 and the device silicon layer 103. The peripheral excitation zone 109 is also provided to minimize squeeze film damping effects, since a minimum actuation force is effected in terms of the applied voltage. In this exemplary aspect, as illustratively shown in FIG.4, the outer peripheral end of the peripheral excitation zone 109 is provided with a rectangular-shaped profile, which is etched anisotropically.

[033] In yet another aspect of the present invention, as
shown in FIG.5, the outer peripheral end of the peripheral excitation zone 109 is provided with a curved rectangular-shaped profile, which is etched isotropically. The extent of the peripheral excitation zone 109 determines the effective electrostatic force and squeeze film damping effects of the MEMS speaker 100 of the present invention.
[034] The device silicon layer 103 is adhered to the
insulator 102, as shown in FIG.1. The device silicon layer 103 is preferably a single crystal silicon layer that is arranged on the insulator 102. The device silicon layer 103 is disposed to overlay the insulator 102 and the peripheral excitation zone 109 as particularly shown in FIG.4. In order to enable metal-free contacts, the device silicon layer 103 is suitably doped with either n-type or p-type dopants. Exemplary P-type dopants include boron, aluminium, nitrogen, gallium, and indium and N-type dopants include phosphorus, arsenic, antimony, bismuth, and lithium. Barring the bi-directional vibrating region 104 the device silicon layer 103, as hereinafter described, acts as a support that rests on the insulator 102 as particularly shown in FIG.4. In other words, the ends of the device silicon layer 103 are pivoted on the insulator 102. The thickness of the device silicon layer 103 is preferably in the range of 1µm to 5µm. The thickness of the device silicon layer 103 is based on the required frequency of vibrating portion 104. The electrical resistivity of the device silicon layer 103 is in the range of 0.01-0.001 ohm-cm. The dopant for the device silicon layer 103 can be n-type or p-type. The device silicon layer 103 is preferably formed by Czochralski process.

[035] The bi-directional vibrating region or portion 104 is
formed in the device silicon layer 103 and positioned to overlay the primary amplitude enhancer 106 and the extended amplitude enhancer 108, on one side, as particularly shown in FIG.4. The constructional arrangement of the bi-directional vibrating region 104 enables the bi-directional vibration of the bi-directional vibrating region 104 both towards and away from the primary amplitude enhancer 106, as shown in FIG.8.
[036] The electrical source 105 is connected to the bi-
directional vibrating region 104 through the handle silicon layer 101 and device silicon layer 103. The handle silicon layer 101 and device silicon layer 103 are provided with a low electrical resistivity to render metal-like characteristics to these layers. To connect these layer 101 and 103 to the electrical source 105, metal wires of gold or aluminum are used for wire bonding process. The electrical source 105 is preferably an alternative voltage source with DC offset. The preferred alternating voltages for actuating the vibration of bi-directional vibrating region 104 are in the range of 80 Vac to 110 Vac and with DC voltage in the range of 80 Vdc to 110 Vdc amplitudes.
[037] Hitherto, the MEMS speaker 100 with a single
primary amplitude enhancer is described. In yet another aspect of the present invention, as shown in FIG.6, the MEMS speaker 100 is provided with two primary amplitude enhancers 106a and 106b. A suitable additional primary amplitude enhancer 106 is incorporated in the MEMS speaker 100 by considering parameters such as squeeze film damping effects.

[038] In yet another aspect of the present invention, as
shown in FIG.7, the bi-directional vibrating region 104 of the device silicon layer 103 is with a rectangular profile and is disposed to overlay the primary amplitude enhancer 106 and the extended amplitude enhancer 108 (not shown in FIG.7). The profiles of the primary amplitude enhancer 106 and the extended amplitude enhancer 108 will also have the corresponding rectangular profile etchings to commensurate with the profile of the bi-directional vibrating region 104.
[039] The bi-directional vibrating region 104 of the device
silicon layer 103 acts as a diaphragm that vibrates both towards and away from the primary amplitude enhancer 106, the extended amplitude enhancer 108 and the device silicon layer 103, as particularly shown in FIG.8, to generate acoustic signals on the application of electric potential. On the application of an alternative electrical potential signal, the bi-directional vibrating region 104 vibrates, moves or oscillates towards and away from primary amplitude enhancer 106. The vibrations push the air zone, which is on the either side of the bi-directional vibrating region 104 so as to generate acoustic signals with decay in pressure gradient.
[040] The functional aspects of the MEMS speaker 100 of
the present invention are now described by particularly referring to FIGs. 7, 8 and 9.
[041] FIGs.7 and 8 illustrate the generation of acoustic
signals on either side of the vibrating region 104. On the application desired voltage to the device silicon layer 103 and the handle silicon layer 101, an electrical potential field is

formed in the peripheral excitation zone 109. The electrical potential field generates electrostatic forces between the bi¬directional vibrating region 104 and the handle silicon layer 101 These forces cause movements in the vibrating region 104 both towards and away from primary amplitude enhancer 106 and the extended amplitude enhancer 108. The deformations in vibrating region 104 cause to push the air zone above and below the bi-directional vibrating region 104. Thus, the developed pressure gradients cause motion of the air zone on both sides of the bi-directional vibrating region 104. These movements take the form of acoustic pressure signals, which reach a receiver. Therefore, the bi-directional vibrating region 104 enables generation of the acoustic pressure signals on both sides, with a phase lag.
[042] FIG.8 depicts the advantage of bi-directional
vibrating region 104 as the presence of acoustic pressure with same level of intensity over near spherical distribution. Exemplary receivers on the either side of the vibrating region 104 receive the same acoustic intensity level with tuned phase lag.
[043] Accordingly, the present invention provides a MEMS
speaker 100, comprising the handle silicon layer 101 with the primary amplitude enhancer. The insulator 102 with the extended amplitude enhancer 108, is arranged on the handle silicon layer 101. The peripheral excitation zone 109 is disposed in the insulator 102 and positioned adjacent to the extended amplitude enhancer 108. The device silicon layer 103 is disposed to overlay the insulator 102 and the peripheral excitation zone 109. The bi-directional vibrating region 104, is

formed in the device silicon layer 103 and positioned to overlay the primary amplitude enhancer 106 and the extended amplitude enhancer 108. The electrical source 105 is connected to the bi-directional vibrating region 104 through the handle silicon layer 101 and the device silicon layer 103.
[044] In yet another aspect of the present invention, the
primary amplitude enhancer 106 and the extend amplitude enhancer 108 are arranged coaxial to each other.
[045] In a further aspect of the present invention, the
peripheral excitation zone 109 and the primary amplitude enhancer 106 are arranged coaxial to each other.
[046] In further aspect of the present invention,at least
two primary amplitude enhancers are disposed in the handle silicon layer 101.
[047] In yet another aspect of the present invention, the
primary and the extended amplitude enhancers 106 and 108 are with circular or non-circular cross-sections.
[048] In still another aspect of the present invention, the
primary and the extended amplitude enhancers 106, 108 are formed as cavities or grooves.
[049] In yet another aspect of the present invention, the
peripheral excitation zone 109 is with isotropic and anisotropic profiles.
[050] In further aspect of the present invention, the
peripheral excitation zone 109 is a dielectric medium.
[051] In yet another aspect of the present invention, the
thickness of the handle silicon layer 101 is greater than the

combined thickness of the insulator 102 and the device silicon layer 103.
[052] The functional aspects of the MEMS speaker of the
present invention are further described by way of the following non-limiting example.
[053] The dynamic response of the exemplary MEMS
speaker of the present invention is measured using a Laser Doppler Vibrometer (MSA500, Polytec) and the radiated sound output (at 10 cm) using a ½” microphone (Type 4190-L-001, B&K) is measured as shown in FIG.11. The experimentally measured mode shape of bi-directional vibrating region 104 of the MEMS speaker is as shown in FIG.10(a) & (b), which clearly indicates that the MEMS speaker vibrates in its fundamental mode. In the fundamental mode, vibrating region 104 vibrates with high amplitudes. In this mode of vibration the bi-directional vibrating region 104 has maximum area of interaction with air zone, which is on the either side of bi¬directional vibrating region 104. It is also observed that the vibration amplitudes of the MEMS speaker are considerable even though the actuation is only at the peripheral excitation zone 109. The measured acoustic output as shown in FIG.12 indicates that the sound level is higher than the ambient sound generated in a working environment. It is also observed that a single MEMS speaker of 2.9 mm diameter and 0.5 mm overall thickness is able to radiate a maximum pressure of about 66 dB SPL at 10 cm, while consuming an input power of just 4.8 mW. An array of 25 of such MEMS speakers, which will have a size of 15x15x0.5 mm, produces a sound pressure of approximately 94 dB SPL at 10 cm, while consuming an input power of about 120

[054] The preferred embodiments of the method for
fabricating MEMS speaker 200 of the present invention are now described by referring to FIG.13. A silicon-on-insulator (SOI) wafer 200 is preferred as shown in FIG.9, which is subjected a selective etching to form MEMS speaker of the present invention. The selected SOI wafer 200 is deposited with a mask layer 211 on the bottom surface of the handle silicon layer 201. The preferred material for the mask layer 211 is selected from materials such as silicon dioxide or silicon nitride. However, a suitable photo resist polymer (AZ4562, AZ4533) can be used as a soft masking material for the handle silicon layer 201. It is

understood here that the selection of suitable material for the mask layer 211 is based on the selectivity between mask and silicon etch, and thickness of the handle silicon layer 201. In the method of fabrication of MEMS speaker of the present invention, silicon dioxide (SiO2) is preferred as a masking material for etching silicon through a deep reactive ion etching process (DRIE). In the present invention DRIE is preferred, since DRIE has the ability to microfabricate deep trenches in silicon layers, while maintaining a high selectivity to the selected masking material, good profile control and low non-uniformity across the SOI wafer. The mask layer 211 is deposited preferably by means o fplasma enhance chemical vapour deposition (PECVD) or the mask layer 211 is grown by thermal oxidation. The thickness of mask layer 211 is determined by the selectivity between silicon and SiO2 in DRIE process.
[055] The mask layer 211is etched for incorporating
microelectromechanical functional elements as primary amplitude enhancer 206 and bi-directional vibrating region 204 of the MEMS speaker 200. A predefined pattern 212 is etched into the mask layer 211 and the etching of the mask layer 211 is preferably performed by a reactive ion etching (dry etching). The etching of the mask layer 211 can also be performed by wet etching in the presence of hydrofluoric acid. The dry etching is preferred to obtain a desired anisotropic profile of the etched portion and whereas wet chemical etching is preferred to obtain an isotropic profile of the etched portion. Accordingly, in the method of the present invention the etching of the mask layer 211 is performed by wet etching to obtain anisotropic profile.

[056] A deep reactive ion etching (DRIE) of the handle
silicon layer 201 is performed to form functional elements in handle silicon layer 201. In this step high aspect ratio structures with greater depth and with high anisotropic in nature are formed. The anisotropic etching of the handle silicon layer 201 is performed to obtain a primary amplitude enhancer 106. The functional elements are the primary amplitude enhancer 206 and the bi-directional vibrating region 204, which are in the mask layer 211 are transferred on to handle silicon layer 201.
[057] Etching of handle silicon layer 201 is further
continued before reaching the insulator 202. The insulator 202 acts as an etch stopper in DRIE of silicon etch process. The depth of the etching and time of the etching can be minimized by choosing a thin handle silicon layer 201.
[058] The insulator 202 is etched, preferably in the
presence of hydrofluoric acid vapour. The hydrofluoric acid (HF) etching of the insulator 202 is performed anisotropically to form the extended amplitude enhancer 208. Thereafter, with an increase in the etching time, the insulator 202 is etched laterally to form a peripheral excitation zone 209.The lateral etching is advantageously performed isotropically. The extent of the laterally etched portion the insulator 202 determines the area of peripheral excitation zone 209.
[059] The primary amplitude enhancer 206 and the
extended amplitude enhancer 208 are formed co-axially to establish an integrated portion, to facilitate vibration of the bi¬directional vibration region204 of the device silicon layer 203,

into and away from the primary amplitude enhancer 206 and the extended amplitude enhancer 208.
[060] The etching of the insulator is performed up to the
device silicon layer 203, so as to define the diameter of the bi-directional vibrating region 204 and the extent of width of the primary amplitude enhancer 206.The lateral etching particularly defines the peripheral excitation zone 209. It is understood here that considering circular profile of the MEMS speaker the bi¬directional vibrating region 204 is formed with a circular profile. Therefore, the profile of the bi-directional vibrating region 204 can be suitably changed to non-circular profiles including elliptical, square or rectangular and the corresponding reciprocal profiles for the primary amplitude enhancer 206 and the extended amplitude enhancer 208 can be selected. In this method of microfabrication, HF vapour has very low etch rate of silicon compared to silicon dioxide and hence the overtime etch does not affect the bi-directional vibrating region 204. In addition, the potential problem of stiction is also eliminated, which is the most reported problem in surface micro-machined MEMS structures.
[061] Electrical source 205 is connected to the bi-
directional vibrating region 204 of the handle silicon layer 201 and the device silicon layer 203. The electrical source 205 provides an electric potential between the handle silicon layer 101 and the device silicon layer 103, which also act as electrodes. The bi-directional vibrating region 104 is arranged to receive electric potential from the electric source 105.

[062] Accordingly, the present invention provides a method
of fabrication of MEMS speaker by initially forming the primary amplitude enhancer, through the mask layer and the handle silicon layer. Subsequently, the extended amplitude enhancer is formed through the insulator and up to the base of the device silicon layer. The peripheral excitation zone is formed adjacent to the extended amplitude enhancer by laterally etching the insulator. Thereafter bi-directional vibrating region is formed from the device silicon layer, where the bi-directional vibrating region is overlaid on the extended amplitude enhancer and the primary amplitude enhancer. In order to render an electrical connectivity to the MEMS speaker the electrical source is enabled to the bi-directional vibrating region through the handle silicon layer and the device silicon layer.
[063] In the method of the present invention, the mask
layer is fabricated by photo printing on a plastic base.
[064] In another aspect of the method of the present
invention, the formation of primary and extended amplitude enhancers is performed isotropically and anisotropically.
[065] In yet another aspect of the method of the present
invention, the peripheral excitation zone is performed by lateral etching and the lateral etching is isotropic and anisotropic.
Advantages of the present invention
[066] The MEMS speaker of the present invention
consumes very less power and offers high efficiency compared to conventional speakers due to its compactness and minimal heat losses as compared to condenser speakers.

[067] The MEMS speaker of the present invention works
with high impedance and provides an inherent protection from overloading.
[068] The MEMS speaker of the present invention is a
voltage driven system, thereby offering negligible heat generation as compared to the current-driven speakers.
[069] The MEMS speaker of the present invention is
compact for suitable use in mobile phones, tablets and laptops.
[070] The MEMS speaker of the present invention has
negligible squeeze-film damping effects and feasibility of high displacements due to presence of primary amplitude enhancer.
[071] The MEMS speaker of the present invention is
compatible for integration with CMOS (complementary metal oxide semiconductor) circuitry.
[072] The MEMS speaker of the present invention is
fabricated a photo-printed mask.
[073] As many apparently widely different embodiments of
this invention may be made without departing from the spirit and scope thereof, it is to be understood that the invention is not limited to the specific embodiments thereof except as defined in the appended claims.

We claim:
1. A MEMS speaker 100, comprising:
- a handle silicon layer 101 with a primary amplitude
enhancer;
- an insulator 102 with an extended amplitude enhancer
108, is arranged on the handle silicon layer 101;
- a peripheral excitation zone 109 is disposed in the
insulator 102 and positioned adjacent to the extended
amplitude enhancer 108;
- a device silicon layer 103 is disposed to overlay the
insulator 102 and the peripheral excitation zone 109;
- a bi-directional vibrating region 104, is formed in the
device silicon layer 103 and positioned to overlay the
primary amplitude enhancer 106 and the extended
amplitude enhancer 108;and
- an electrical source 105 is connected to the bi¬
directional vibrating region 104 through the handle
silicon layer 101 and the device silicon layer 103.
2. The speaker 100 as claimed in claim 1, wherein the primary
amplitude enhancer 106 and the extend amplitude
enhancer 108 are coaxial.
3. The speaker 100 as claimed in claim 1, wherein the
peripheral excitation zone 109 and the primary amplitude
enhancer 106 are coaxial.

4. The speaker 100 as claimed in claim 1, wherein at least two
primary amplitude enhancers are disposed in the handle
silicon layer 101.
5. The speaker 100 as claimed in claim 1, wherein the primary
and the extended amplitude enhancers 106 and 108 are
with circular or non-circular cross-sections.
6. The speaker 100 as claimed in claim 1, wherein the primary
and the extended amplitude enhancers 106, 108 are
cavities or grooves.
7. The speaker 100 as claimed in claim 1, wherein the
peripheral excitation zone 109 is isotropic and anisotropic.
8. The speaker 100 as claimed in claim 1, wherein the
peripheral excitation zone 109 is a dielectric medium.
9. The speaker 100 as claimed in claim 1, wherein the
thickness of the handle silicon layer 101 is greater than the
combined thickness of the insulator 102 and the device
silicon layer 103.
10. A method for fabricating a MEMS speaker, said method
comprising the steps of:
(a) forming a primary amplitude enhancer, through a
mask layer and a handle silicon layer;
(b) forming an extended amplitude enhancer through an
insulator and up to the base of device silicon layer;

(c) forming a peripheral excitation zone adjacent to the
extended amplitude enhancer by laterally etching the
insulator;
(d) forming a bi-directional vibrating region from the
device silicon layer, where the bi-directional vibrating
region is overlaid on the extended amplitude enhancer
and the primary amplitude enhancer; and
(e) connecting the electrical source to the bi-directional
vibrating region through the handle silicon layer and
the device silicon layer.
11. The method as claimed in claim 10, wherein the mask layer
is fabricated by photo printing on a plastic base.
12. The method as claimed in claim 10, wherein the formation
of primary and extended amplitude enhancers is performed
isotropically and anisotropically.
13. The method as claimed in claim 10, where the peripheral
excitation zone is performed by lateral etching and the
lateral etching is isotropic and anisotropic.