Mems by Mama

8/7/2019 Mems by Mama

1/20

1.RF MEMS SWITCH

Introduction:-

The Holy Grail of MEMS devices is the micromechanical switch. For more than a

decade, researchers have endeavored to perfect the development of microminiature relays

using micromachining techniques. With the recent boom in wireless communications,

research has intensified in the quest to develop low cost, ultra-low loss switches. The

goal is to have these switches replace traditional FETS for reduced loss and improved

linearity in key components. There are fundamentally two types of switch contact

mechanisms ohmic contact and capacitive contact. With ohmic switches, two metal

electrodes are brought into contact to create a low-resistance connection. In capacitiveswitches, a metal membrane is pulled down onto a dielectric layer, usually by

electrostatic means, to form a capacitive sandwich. At high frequencies, the capacitive

suseptance of this sandwich acts like a short circuit. In either case, the mechanical action

of the switch causes the switch to efficiently change from high impedance to short circuit.

Micromechanical switches can utilize one of many actuation mechanisms,

including magnetic, piezoelectric, thermal, and most commonly electrostatic forces.

Switches that operate electrostatic ally require very little energy, usually on the order of

tens of nanojoules per switch cycle. The Achilles Heel of all MEMS switches is their

switching speed, which is determined by their mechanical resonant frequency. Actuation

is typically accomplished in microseconds to 10s of microseconds for electrostatic ally

operated devices, and 100s of microseconds to milliseconds for thermal actuators.

Despite the fact that MEMS switches operate slower than their electronic counterparts,

they are still useful in many applications. One important advantage of MEMS switches is

their linearity. Unlike electronic switches made with metal-semiconductor or p-n

junctions, the contact area for MEMS switches is perfectly linear. This means that well-

designed MEMS switches do not create nonlinearities or distortion such as harmonics or

intermediation products. In many cases, these nonlinearities are immeasurable.


2/20

2. History

1948 Invention of the Germanium transistor at Bell Labs (William Shockley)

1954 Piezoresistive effect in Germanium and Silicon (C.S. Smith)

1958 First integrated circuit (IC) (J.S. Kilby 1958 / Robert Noyce 1959)

1959 "Theres Plenty of Room at the Bottom" (R. Feynman)

1959 First silicon pressure sensor demonstrated (Kulite)

1967 Anisotropic deep silicon etching (H.A. Waggener et al.)

1968 Resonant Gate Transistor Patented (Surface Micromachining Process) (H.Nathanson, et.al.)

1970s Bulk etched silicon wafers used as pressure sensors (Bulk Micromaching

Process)

1971 The microprocessor is invented

1979 HP micromachined ink-jet nozzle

1982 "Silicon as a Structural Material," K. Petersen

1982 LIGA process (KfK, Germany)

1982 Disposable blood pressure transducer (Honeywell)

1983 Integrated pressure sensor (Honeywell)

1983 "Infinitesimal Machinery," R. Feynman

1985 Sensonor Crash sensor (Airbag)

1985 The "Buckyball" is discovered

1986 The atomic force microscope is invented

1986 Silicon wafer bonding (M. Shimbo)


3/20

1988 Batch fabricated pressure sensors via wafer bonding (Nova Sensor)

1988 Rotary electrostatic side drive motors (Fan, Tai, Muller)

1991 Polysilicon hinge (Pister, Judy, Burgett, Fearing)

1991 The carbon nanotube is discovered

1992 Grating light modulator (Solgaard, Sandejas, Bloom)

1992 Bulk micromachining (SCREAM process, Cornell)

1993 Digital mirror display (Texas Instruments)

1993 MCNC creates MUMPS foundry service

1993 First surface micromachined accelerometer in high volume production (Analog

Devices)

1994 Bosch process for Deep Reactive Ion Etching is patented

1996 Richard Smalley develops a technique for producing carbon nanotubes of uniform

diameter

1999 Optical network switch (Lucent)

2000s Optical MEMS boom

2000s BioMEMS proliferate

2000s The number of MEMS devices and applications continually increases

2000s NEMS applications and technology grows

Earlier bi-MEMS switch


4/20


5/20

3. MEMS Switch characteristics

3.1Actuation Mechanisms:

The actuation forces required for the mechanical movement can be obtained using

electrostatic, magneto-static, piezoelectric or thermal designs. To date, only electrostatic-

typeswitches have been demonstrated at 0.1-100GHz with high reliability at low RF

powers for metal contact and medium power levels for capacitive contacts (100Million to

50 Billion cycles depending on the manufacturer) and wafer-scale manufacturing

techniques.Other switches which have demonstrated excellent performance are the

Microlab Latching switch (up to 100 Million cycles) using magnetic actuation, and the

thermal switches developed independently by Cronos Microsystems and the Univ. of

California, Davis. It is hard to test thermal switches for long cycle times due to their slow

switching response (1-3ms).

3.2 Switching Time:

Electrostatic switches can be made small and with a very fast switching time (2-30 s)

while thermal/magnetic actuation requires around 100-2, 000 s of switching time. An

excellent metal-contact switch developed by LETI using thermalactuation but with an

electrostatic hold, thereby requiring very little switching energy and virtually zero hold-

down power. However, its switching time is still relatively slow (300 s). The LETIswitch has been tested to more than 100 million cycles.


6/20

Fig. An SPDT switch packaged using a gold-to-gold seal ring (courtesy of Microassembly, Inc.). The

topcover is taken off so as to show the seal ring .

3.3 Contact Type:

There are two different contacts in RF MEMS switches, a capacitive contact and ametalto-metal (or DC) contact. The capacitive contact is characterized by the capacitance

ratio between the up-state (open circuit) and down-state (short-circuit)positions, and this

is typically 80-160 depending on the design. The down-state capacitance is typically 2-3

pF, and is suitable for 8-100GHz applications. In general, it is hard to obtain a large

down-state capacitance using nitride or oxide layers, and this limits the low-frequency

operation of the device. On the other hand, DC-contact switches with small up-state

capacitances (open circuit) can operate from 0.01 to 40GHz, and in some cases, to

60GHz (for example, the Rockwell Scientific switch has an up-state capacitance of only

1.75 fF and an isolation of 23 dB at60GHz). In the down-state position (short-circuit), the

DC-contact switch becomes a series resistor with a resistance of 0.5-2 , depending on

the contact metal used.


7/20

3.4 Circuit and Substrate Configurations:

As is the case with all two-terminal devices, the switches can be placed in series or in

shunt across a transmission line. Typically, capacitive switches have been used in a shunt

configuration, while DC-contact switches are placed in series. The reason is that it is

easier to get a good isolation with a limited impedance ratio (such as the capacitive

switch) in a shunt-circuit than in a series circuit. Also, MEMS switches are compatible

with both microstrip and CPW lines on glass, silicon and GaAs substrates, and have been

used in these configurations all the way to 100GHz. For low loss applications at

microwave frequencies, it is important to use high-resistivity substrates.


8/20

4. Circuits with RF MEMS switches

The near-ideal electrical response of RF MEMS switches (both metal-contact and

capacitive) have allowed many designers to build state-of-the-art switching circuits from

0.1GHz all the way to 120GHz.In the past 4 years, these applications concentrated on the

replacement of GaAs phase shifters which are commonly used in phased arrays by the

thousands of units. A comparison between 3-bit GaAs phase shifters and MEMS phase

shifters is shown in Table I and it is seen that MEMS switches provide an immense

performance benefit especially at Ka-Band to W-band applications

Average on-wafer lo ss for RF MEMS and GaAs-FET 3-bit phase shifters.

Freq. (GHz) Loss RF MEMS (dB) Loss GaAs FET (dB)

X-Band (10)

Ku-Band(20)

Ka-Band (35)

V-Band (60)

W-Band (94)

0.3/bit

0.45/bit

0.6/bit

0.8/bit

0.9/bit

1.2/bit

1.6/bit

2.3/bit

2.8/bit

3.3/bit

Fig. 4 presents a 4-bit miniature RF MEMS phase shifter developed jointly by the Univ.

of Michigan and Rockwell Scientific. It is based on the Rockwell metal contact switch

and on CLC delay lines for miniaturization. The phase shifter results in an average loss of

1.4dB at 10GHz, a 3 phase error, and is matched to 13 dB at the input and output

ports from 6-16GHz. This phase shifter represents the smaller design using RF MEMS

to-date, and with excellent response. an 885-986MHz 5-pole tunable filter using

switched MEMS capacitors developed by Raytheon Systems Co. In this case, capacitive

switches are used to switch fixed-value metal insulator- metal capacitors in thetransmission line. The filter employs 18 switches and is a very complicated circuit with

variable resonators and impedance inverters. Its measured response is nearly ideal, with

excellent frequency tuning capabilities, very high linearity (in terms of measured IIP3)

and a loss of 5-6 dB due to the finite Q of the planar inductors used (Q = 30 at 0.9GHz).


9/20

Fig. 4. The 4-bit miniature X-band phase shifter developed

by the Univ. of Michigan and Rockwell Scientific. The

size is 3.2 2.1mm2.

Fig. 5. An 885-986 MHz 5-pole tunable filter using switched MEMS capacitors

developed by Raytheon Systems Co. The size is 3.5 14mm2.

Fig. 6. The 3-bit true-time delay distributed MEMS phase


10/20

shifter at 77-100GHz. The size is 1.9 5mm2.Fig. 6 presents a W-band 3-bit phase

shifter developed at the Univ. of Michigan using MEMS capacitive switches. This is the

highest frequency MEMS phase shifter to-date and results in an average loss of 2 .7-2.9

dB at 77-94GHz with an associated phase error of3. The results are about 8 dB better

than GaAs designs. Other circuits, which are not shown due to space constraints, are very

wideband SP4T switches, high isolation series/shunt switches covering 0.1-50GHz,

double-pole double-throw transfer switches, and a whole range of phase shifters from

8GHz to 120GHz.Also, tunable filters covering 200MHz to 23GHz have been developed

by various groups. In general, RF MEMS circuits outperform GaAs FET and PIN diode

circuits by a large margin at all frequencies of interest to the RF and microwave

communities. Most ofthe circuits developed in the world can be found in.


11/20

5. Design and Fabrication of the RF Switch

5.1 Design Optimization

The main purpose of this optimization scheme is to maximize the deflection for a constant applied

voltage to the actuator (5 volts). ANSYS software includes a parametric solver that was used

to

perform the optimization based on the following criteria:

Design Variables:

o Length of beam (150 m


12/20

Figure 7: Cantilever Deflection

Figure 8: Magnitudes of Deflection ANSYS simulation


13/20

5.2 Switch fabrication

The summary of the steps proposed for the fabrication are as follows (figure 8) [7]:

Begin with aSilicon substrate -1-

Deposit thesilicon nitride (SixNy) as an insulating layer using chemical vapor

Deposition (CVD)

Create asilicon dioxide (SiO2) sacrificial layer by CVD

Using positive photo-resist, the sacrificial layer is exposed to ultra violet rays

through a mask

The whole substrate is developed in a developer solution (H2SO4) to remove area of

SiO2 exposed to UV

The first layer of metal (Au) is deposited using sputter deposition -2-

Pattern the layer of Au by deep reactive ion etching (DRIE)

Repeat above processes to deposit the heat sink metal -3-

Deposit an SiO2 layer

Use lithography to pattern the SiO2 then deposit the first Nitinol TiNialloy metal by

sputter deposition and repeat same process to deposit the second Au metal contact

see 5-

Pattern the SiO2 layer to open a window in the sacrificial layer


14/20

Deposit the polyimide to form the cantilever beam -4-

_ The last sacrificial layer is deposited by CVD

_ Pattern the SiO2 layer to deposit the second TiNi alloy metal

_ Use DRIE to pattern the TiNi in the desired form

Selective etch the SiO2 layer in hydrofluoric acid (HF) leaving a free standing micro

structure SiO2 SixNy


15/20

SiO2 SixNy

(a)

Au

(B)

Heat Sink

(C)

Polyimide

(D)


16/20

(E)

Figure 9: Fabrication process steps


17/20

fig. fabrication process


18/20

FIG.Process flow showing the key steps in the fabrication of the RF MEMS switch. (a) Etching

the device layer to form a cavity and fabrication of contact dimple. (b) Patterning of glass wafer with

transmission lines and trenching to half its thickness. (c) Anodic bonding of glass and SOI wafers.

(d) Removal of handle layer and buried oxide. (e) Patterning of upper electrode contact pads and

actuator structure on device layer. (f) DRIE etching to form the actuator, ashing of photoresist and

device singulation


19/20

6. ELECTROMECHANICAL MODELING OF RF MEMS

SWITCHES

A typical capacitive RF MEMS switch consists of a fixed-fixed thin metallic

membrane which is suspended over a bottom electrode insulated by a dielectric film.

When the switch is not actuated, there is low capacitance between the membrane and the

bottom electrode, and the device is in the OFF state. When voltage is applied between the

movable structure and the fixed bottom electrode, electrostatic charges are induced on

both the movable structure and the bottom electrode. The electrostatic charges cause a

distributed electrostatic force, which deforms the movable structure. In turn, such

deformation leads to storage of elastic energy, which tries to restore the structure to its

original shape. The structure deformation also results in the reorganization of all surface

charges on the device. This reorganization of charges causes further

Structural deformation; hence, the device exhibits a highly nonlinear, coupled

electromechanical behavior. Until a certain voltage is applied, the so-called pull-in

voltage or actuation voltage, an equilibrium position exists through a balance between the

elastic restoring force and electrostatic force. After pull-in, the device is in the ON state

and its capacitance is much larger than that in the OFF state.

The switch actuation is therefore a coupled-field problem of electrostatics and structural

response. In order to accurately describe the switch deformation and predict the pull-in

voltage, an effort to realize modeling has to be made. In the following section, we will

discuss a simple 1D parallel-plate actuator model [7, 8], a 2D distributed model [9, 10],

and a 3D fully coupled model [11, 12]. The analyses and simulations are dedicated to

capacitive MEMS switches, although they are also applicable to other types of

electrostatic devices.


20/20

Figure . (a) Z-displacement contour of the type A membrane with four corrugations; (b) magnified z-

displacement along the center line in (a) (only half is plotted due to symmetry); (c) Cauchy stress

distribution in theX-direction (the average stress in the central flat region is 6.5 MPa

Mems by Mama

Documents