Previous Page  52 / 60 Next Page
Information
Show Menu
Previous Page 52 / 60 Next Page
Page Background

50

Chip Scale Review May • June • 2018

[ChipScaleReview.com]

Challenges of flip-chip MEMS microphones: from

state-of-the-art to small housing sizes

By Sebastian Walser, Anton Leidl, Wolfgang Pahl

[EPCOS, a TDK Group Company]

oday, microelectromechanical

systems (MEMS) are commonly

used as sensors in numerous

electronic consumer products. The main

reason for this is an accurately controlled

silicon micromachining technology

in combination with low cost at mass

production. Currently, in mass production,

tight sensitivity distributions down to ±1dB

or beyond can be reached [1]. Especially for

high-volume consumer applications (e.g.,

mobile phone communication), capacitive

silicon MEMS microphones have become

state-of-the-art and reach high signal-

to-noise ratios (SNR) of approximately

65dB(A) with sensitivities of -38dBV/Pa

and component sizes of approximately 3.5

x 2.6 x 1.0mm

3

[2,3]. With the increasing

requirements of smartphone technology, a

demand of the MEMS microphone market

is a continuous reduction in size with the

current acoustic performances. It will be a

challenge for the next few years to minimize

the MEMS microphone package size without

downgrading the electroacoustic microphone

characteristics.

Classification of MEMS microphones

Today the most popular commercial

MEMS microphone principle is the capacitive

transducer. Such microphones consist of

two chips: a sensor chip and an application-

specific integrated circuit (ASIC) chip. Both

chips are integrated in a surface-mount

device (SMD) package. In general, the sensor

chip has a movable membrane and a rigid

perforated back-plate electrode. Within this

structure, an incoming sound wave results in

a capacitive change and will be converted by

an ASIC into an electrical audio signal.

During the last few years, different

package variants established themselves in the

commercial market. A classification of these

MEMS microphones can be done on the basis

of the sound port (

Figure 1

). If the sound

pressure gets through a hole in the metal lid to

the sensor, the microphone is called top-port

(

Figure 1a

). Otherwise, if the sound pressure

gets through a hole in the substrate, the

microphone is called bottom-port (

Figure 1b

).

In general, both classifications reach a similar

electroacoustic performance parameter with

similar package sizes.

In the case of bottom-port MEMS

microphones, there are two different

packaging technology approaches on the

market. Depending on the internal connection,

the MEMS microphone can be built up by

wire bond or flip-chip (

Figure 2

). Both cross

sections show a carrier substrate with electric

contact pads and a sound hole. In each case,

a sensor chip and ASIC chip are mounted on

the carrier substrate. For flip-chip assembly,

the mechanical and electrical connection

of both chips is done by soldering (

Figure

2a

). An example of a f lip-chip MEMS

microphone is presented by G. Feiertag, et. al

in [4]. In a wire bond assembly (

Figure 2b

)

the mechanical connection of both chips to

the carrier substrate is done by adhesive. Due

to the electrical contact pads on the sensor

and ASIC, both chips are mounted reversed

to the flip-chip assembly, and the electrical

connection is done by wire bonding. An

example of a wire-bond MEMS microphone

is presented by A. Dehe, et. al in [5].

The main advantage of a flip-chip MEMS

microphone package is the space-saving

design [4]. In comparison to the flip-chip

package, the wire bond assembly needs

additional spaces on the carrier substrate for

the wire bonds (see

Figure 2

). Assuming

a package size of 3.35 x 2.5 x 1.0mm

3

, a

sensor chip size of 1.45 x 1.45 x 0.45mm

3

and an ASIC chip size of 1.00 x 1.45 x

0.30mm

3

, a flip-chip MEMS microphone

has a front volume of around 0.2mm

3

and a

back volume of around 3.4mm

3

. The front

volume includes everything from the sound

hole to the membrane. The back volume

includes everything behind the membrane.

In comparison to a wire bond assembly, the

sensor cavity increases the back volume

of around 0.5mm

3

. Assuming the same

sensor, ASIC and package size — and

without taking the polymer foil thickness

into account — the back volume of a wire

bond package can be calculated as follows:

3.4mm

3

– 0.5mm

3

+ 0.2mm

3

= 3.1mm

3

. This

corresponds to a decrease in the back volume

of around 8.8%. A smaller back volume leads

to a larger restoring force to the membrane.

This results in a smaller deflection of the

T

Figure 1:

Classification of MEMS microphones on the basis of sound port: a) top-port, and b) bottom-port (by

using image source [6]).

Figure 2:

Classification of bottom-port MEMS microphones on the basis of packaging technology: a) flip-chip

package, and b) wire bond package (by using image sources [3] and [6]).