A Comparison Between RF MEMS Switches and Semiconductor Switches

P.D. Grant,1, ∗ R.R. Mansour,2 and M.W. Denhoff1

1 Institute for Microstructural Sciences, National Research Council, Ottawa, Canada K1A 0R6
2 Electrical and Computer Engineering Department, University of Waterloo, Waterloo, Canada , N2L 3G1
(Dated: 3 November 2001)
This paper addresses the fundamentals of RF switches providing a comparison between semiconductor and
RF MEMS switches. The basis of comparison is introduced by defining a figure of merit that is a function of the
off-state capacitance and the on-state resistance. A simple transmission line model is presented to illustrate the
impact of the switch off-state capacitance on the switch isolation and frequency range of operation. The figure
of merit analysis given in this paper demonstrates that RF MEMS switches have superior insertion loss and
isolation performance in comparison to MESFET and p-i-n diode switches. The paper also addresses several
other design considerations beside insertion loss and isolation for selecting the right RF switch. A discussion is
given on the potential use of RF MEMS switches in satellite and wireless applications.

INTRODUCTION Given the existing state of the art in semiconductor switches

The common microwave switches currently employed in
the microwave industry are mechanical switches (coaxial and
waveguide) and semiconductor switches (p-i-n diode and
FET). Mechanical coaxial and waveguide switches offer the
benefits of low insertion loss, large off-state isolation, and
high power handling capabilities. However, they are bulky,
heavy and slow. On the other hand, semiconductor switches
such as p-i-n diodes and FET provide much faster switching
speed and are smaller in size and weight, but are inferior in in-
sertion loss, DC power consumption, isolation and power han-
dling capabilities than their mechanical counterparts. MEMS
switches promise to combine the advantageous properties of
both mechanical and semiconductor switches. They offer the
high RF performance and low DC power consumption of me-
chanical switches but with the small size, weight and low cost
features of semiconductor switches.
We have worked on improving the performance of mi-
crowave switches for several years [1–3]. Over that time
MEMS devices have become a focus for research in mi-
crowave switches [4–6]. For a review of RF MEMS de-
vices see [7]. While it is generally accepted that mechanical
switches offer superior isolation, we have not seen an analy-
sis of the fundamental limits of semiconductor switches. In
this paper we will present an analysis of fundamental issues
in microwave switching. First we will review the two basic
switch configurations used for microwave switching; series
and shunt. We hope that this review will be of use to those
who have been working in the MEMS field but are not famil-
iar with microwave devices. The principle comparison crite-
rion for devices is the figure of merit which we will relate to
fundamental material physics . We briefly discuss the elec-
tronic models of MESFET, p-i-n, and photo-conductive semi-
conducting switches. The two main types of MEMS switches
will be discussed and then compared with semiconductor de-
vices using the figure of merit criteria. Based on this compar-
ison we will show some fundamental reasons for the advan-
tages that MEMS switches offer for microwave routing. There
are many application areas possible for microwave switches.
and the benefits that MEMS offers we will describe possible
application areas where MEMS devices will be used to advan-
tage.
There is confusion concerning switching times and switch-
ing time requirements. It is often thought that to route a
microwave signal a switching device must switch at the mi-
crowave frequency. That is not the case. Most applications in
telecommunications and radar require switching times of mi-
croseconds to milliseconds. What is important is low loss in
the transmitting state and high isolation in the off-state. We
view a microwave switch as a passive device which can be
changed between two states, an on-state and and off-state.
Before we can make the comparison, we will consider what
properties make an ideal switch.
While each switch has a different circuit model when con-
sidered in detail, the fundamental element that each provides
is a low resistance in one state which can be controllably
changed to a small capacitance in the other state. A switch-
ing circuit is usually made by combining several switching
elements. For example, a MMIC (microwave monolithic in-
tegrated circuit) switch is often a single pole double throw
configuration which in microwave terms is a 3 port device.
A microwave signal can be routed from port 1 to either of
ports 2 or 3. Along each of the possible signal paths there will
be a series switch followed by a shunt switch. The transistor
based series switch element provides limited isolation in the
open state. Hence it is followed by a second switch element
to improve the isolation, but the second stage adds additional
insertion loss. In this paper we want to address fundamentals
of switching elements rather than circuit devices.
MICROWAVE SWITCH
While electronic devices are widely used as switching el-
ements for routing signals there is a surprisingly small body
of literature devoted to microwave routing switches. In many
microwave texts, the general problem of switching receives
less than a chapter. In the area of devices, the requirements of
a good switching element can be as demanding as those for a

© NRC publication no. 44453: Published in: Can. J. Elect. Comput. Eng., Vol. 27, No. 1, pp. 33-39, Jan. 2002
2
Vi
Z0 Z0
1 2
Vr Vt
FIG. 1: An ideal series switch.
good amplifier. In the following section we offer a basic in-
troduction to microwave switches as reflectors of quasi-TEM
waves propagating on transmission lines. All of the switch
models discussed in this paper are based on the assumptions
that the switch element itself is very small in comparison with
the wavelength of the signal that is routed. This is true for
both semiconductor switches and for MEMS at frequencies
up to the range of 100 GHz, i.e a free space wavelength of
3 mm. At that frequency, the guided wavelength is typically
1.5 mm while the dimensions of the switch are much smaller.
Transmission line
The switch element is considered to be a part of a cir-
cuit that includes an incoming and outgoing transmission line.
This is one of the significant differences between microwave
switching and low frequency switching. While we can con-
sider the switch element using lumped element models, we
will not be able to include other elements without exceeding
some fraction of a wavelength. In the ideal models that follow
we do not show either the phase shift or losses due to the trans-
mission lines. In general we must consider the impedance of
the transmission line that connects to the switch. Frequently,
we must also consider the mode of propagation for the signal
on the line, because the electric field orientation influences the
value of the off state capacitance.
Series switch
The series reflecting microwave switch is introduced
schematically in Fig. 1. It looks much like any representa-
tion of an electrical switch. To switch microwaves it must
Vi Coff
Z0 Z0
1 2
Vr Vt
FIG. 2: A series switch model in the off state.
NRC publication no. 44453
be possible to transmit microwave signals to and from the
switch, hence it is connected at input and output by trans-
mission lines or waveguides. A further distinction is that the
switch is viewed as a transmitting device with some reflection
when in the closed state. In the open state state, the switch is
viewed as fully reflecting with some un-wanted transmission.
The concept of reflected and transmitted waves may be more
familiar in optical switching. In fact, the microwave frequency
range that we consider in this paper, 1 GHz to 100 GHz, in op-
tical terms is 0.03 to 3 cm−1 wavenumbers.
In the off-state case, a signal arriving at port 1 of the open
switch is reflected with a voltage reflection coefficient of +1
as is shown schematically in Fig. 2. A small part of the wave
is transmitted through the switch to port 2. In the figures, we
show the electrical signal as a short pulse, which emphasizes
the sign of the reflection coefficient which for the series switch
is +1 and for the shunt switch is –1. We also chose a pulse
to emphasize the very broadband nature of switching that is
possible using MEMS. In terms of microwave scattering pa-
rameters, in the frequency domain, the reflected signal is S11
and the transmitted signal is S21 , the forward transmission is
given by the following equations:
1
S11 = , (1)
1 + jωCoff 2Z0
jωCoff 2Z0
S12 = . (2)
1 + jωCoff 2Z0
If jωCoff is much smaller than 1, the denominators of both
equations is approximately 1. This gives unity reflection,
while in transmission the circuit is basically a differentiator.
Usually, the forward transmission under off-state conditions
is termed the isolation of the switch.
In the on-state, sketched in Fig. 3, a signal is mostly trans-
mitted through the switch with some small reflection and some
absorption. The insertion loss is the ratio of the transmit-
ted power to the difference between the incident and reflected
power. If the reflected power is low then S21 is the insertion
loss. The reflected power under these conditions is the return
loss and in the case sketched is equal to S11 . The on-state
scattering parameters are given by
Ron
S11 = , (3)
Ron + 2Z0
Vi Ron
Z0 Z0
1 2
Vr Vt
FIG. 3: A series switch model in the on state.

Vi
Z0
1 2
Vr Vt
FIG. 4: An ideal shunt switch.
2Z0
S21 = . (4)
Ron + 2Z0
Shunt switch
The schematic of a reflecting shunt switch is shown in
Fig. 4. The switch element is connected across the transmis-
sion line which is an unusual configuration in low frequency
switching. At microwave frequencies it acts in a fashion anal-
ogous to the series switch in that it reflects incoming signals,
but with a reversal in the sign of the voltage. The transmit-
ting or on-state is shown in Fig. 5. The switch appears as a
transmission line directly connecting ports 1 and 2. The in-
fluence of the switching element is to add a small capacitor in
shunt with the line causing a discontinuity. Due to this dis-
continuity, some power is reflected. As frequency increases,
the impedance of this capacitor drops causing an increasing
reflection, as can be seen in the following equations:
− jωCZ0
S11 = , (5)
2 + jωCZ0
2
S21 = . (6)
2 + jωCZ0
A given switching element such as a p-i-n diode can be
used in either a series or shunt connection. The small capaci-
tance that in a series switch allows some signal to pass in the
off-state causes a reflection when the same device is used in
a shunt configuration. Since a reflection is signal not trans-
mitted, this capacitance causes a loss in the shunt circuit and
Vi rline
Z0
1 C 2 C′
Vr Vt
FIG. 5: A shunt switch model in the on state.
NRC publication no. 44453
3
should be minimized. Another loss element in a shunt switch
is the transmission line loss. This loss may be greater than in
the corresponding line for a series switch because the line is
sometimes made thinner to accommodate the switch element.
This is the case for a MEMS shunt switch.
The switch in Fig. 6 is in the fully reflecting state or off-
state. For an ideal switch, the voltage reflection coefficient is
–1 which is shown by the inversion of the reflected voltage
pulse. In a real switch, there is some residual resistance in the
switch which reduces the magnitude of the voltage reflection
coefficient. The power that is not reflected is transmitted to
port 2 leading to a loss of isolation in the switch. Equations
(7) and (8) give the microwave scattering parameters for the
off or reflecting state, assuming an infinitely large C ′:
−Z0
S11 = , (7)
2R + Z0
2R
S21 = . (8)
2R + Z0
In Fig. 6 we show a resistor in series with a capacitor as
the device model. Both contribute to loss of isolation in the
switch. The capacitor is included in many MEMS switches
to prevent contact between the moving element and the trans-
mission line. This capacitor is required in the type of switch
developed by Goldsmith et. al. [6] because the signal lines are
also part of the electrostatic actuator. It has also been found
that a metal-to-insulator mechanical contact can be more re-
liable than a metal-to-metal contact and hence switches are
made that are variable capacitors. In this case, the reactance
of the capacitor decreases at low frequencies which limits the
power reflected and allows transmitted power. For the remain-
der of our paper we ignore this capacitor by considering it to
be infinitely large.
The resistor that appears in Fig. 6 is the same resistive part
of the switching element that appears as a series element in
Fig. 3. While in the series configuration this resistor absorbs
some of the power that flows through the switch, in the shunt
configuration it limits the magnitude of the reflection coeffi-
cient. For a perfect reflection and isolation, a resistance value
of 0 is necessary. Thus, although the function of the resistor
Vi
Z0
1 R 2
Vr Vt
FIG. 6: A shunt switch model in the off state.
4
seems to be the opposite in the series and shunt configura-
tions, the direction of optimum is the same. In both cases,
minimizing the resistance of the switch optimizes the circuit.
Summary of switch models
In the above discussion of series and shunt switch configu-
rations we have presented an abstract model of a “switching
device” that can be characterized by a small capacitance in
one state and a small resistance in the other state. This may
seem like an oversimplification to purists who would like to
use full models for each semiconductor device. However, in
practice, the simple 2 parameter model provides quite accu-
rate prediction of switching circuit performance. It is also
straightforward to extract these simple parameters from the
basic properties of each type switch. We have also noted that
the direction of optimum performance is the same for either
series or shunt configuration. That is, we want lower resis-
tance and lower capacitance which leads to the concept of fig-
ure of merit.
FREQUENCY DEPENDENT CHARACTERISTICS
Figure of Merit
With direct current one uses the off-on resistance ratio to
characterize a switch. Since at microwave frequencies the
off-state is determined by capacitance, we use the ratio of
impedance to compare switches. The impedance ratio for our
simple two parameter device is just
Zon
Zratio = = jωCR. (9)
Zoff
The ratio is frequency-dependent, a trait which we would ex-
pect since one state is determined by a capacitance. The prod-
uct of CR is a characteristic number called the figure of merit
(FOM) of a switch. It has been used by several authors with-
out very much discussion. A smaller FOM is better since it
means a small on impedance relative to the off impedance.
The reciprocal of the FOM is also used as a metric and is
called the cut-off frequency. The FOM matches what is ex-
pected intuitively. We can make a better switch element by
reducing R while C is held constant or by reducing C while
holding R constant. In an ideal situation we can reduce both
R and C.
While we have developed the FOM from an impedance ra-
tio, we cannot assign it a great deal of physical meaning. In
a MEMS switch FOM values are in attoseconds (as), which
corresponds to cut-off frequencies of tens of terahertz (THz).
These frequencies are extrapolations far beyond the range of
validity of the assumptions that we made earlier in developing
our simple switch model. That means that we cannot plot a
graph of frequency response for a circuit and find a break in
the THz region. For example, if we consider any series switch
NRC publication no. 44453
in a circuit, the impedance of the transmission line introduces
a break frequency that is lower than the cut-off frequency by
the ratio of line impedance to Ron . The FOM provides useful
insight into the fundamental behaviour of switching elements
and lets us compare switching element behaviour in isolation
from circuits and from device geometry.
The first case for which we want to extract an FOM is not
one of our switching elements at all but an ideal element of
semiconductor material. In Fig. 7 we show a small volume
of material deep in the semiconductor. We assume that some
external control is exercised to switch the material from an in-
sulating state to a conducting state. The capacitance of this
small element of material is just C = εr ε0 dxdy/dz. The resis-
tance of the same volume of material is R = ρdz/dxdy. The
figure of merit is then
dx · dy dz
FOM = CR = ε0 εr ρ = ε0 εr ρ. (10)
dz dx · dy
Thus, the geometry cancels and the FOM is just dependent on
the basic material properties of dielectric constant and resis-
tivity. We cannot improve on the FOM by changing geom-
etry. For example, if we want to decrease Ron we can make
dz smaller. At the same time however, that would made Coff
larger in the same proportion as we reduced Ron and the FOM
would be constant. From this reasoning, we conclude that for
a solid state device, the FOM has a floor value that is deter-
mined by the dielectric constant and the lowest resistivity that
can be achieved in that same volume.
We write the FOM in terms of semiconductor mobility, µn ,
carrier concentration, N, and electronic charge, q; FOM =
ε0 εr /(µn Nq). Inserting values for GaAs of N = 1017 cm−3 ,
µ = 5000 cm2 /(V·s), and εr = 13.1 [8], we get an FOM of
14.5 fs. We show in the following sections that this value is
substantially less than any experimental values found for de-
vices. The higher device values are due to contact resistance
and fringing fields. We will also show that MEMS devices
already demonstrate a lower FOM.
In principle, the FOM of a solid state device like a MES-
FET can be calculated by integrating eqn. (10) remembering
that in general ρ and ε are functions of position. Also, the
electric field distribution may change when the switch is in
different states reducing the direct application of the simple
model. This has already been done in the form of device mod-
els from which we can extract the effective R and C values.
This we will do in the following section.
dx
dz ρ , εr E
dy
FIG. 7: An element of the active area of a semiconductor.
 5

SEMICONDUCTOR SWITCHES

MESFET Area A

The model of a MESFET switch shown in Fig. 8 was pre- VR W p+ Coff

sented by Ayasli [9] to describe the relationship between the
i
device physics and the RF switching behaviour of MESFETs.
In the conducting state, the MESFET is operated with zero n+
bias and the channel is approximately a linear conductor. In
the insulating state, the gate to channel Schottky diode is re-
verse biased depleting the channel. The off state capacitance FIG. 9: PIN diode in off state.
is determined by the Csd in parallel with the series combi-
nation of the Csg and Cgd or Cg . The Cg is determined by the
note is that the conducting path is short in comparison with
width of the depletion zone and the gate length. We have mea-
the lateral size, so the capacitance can be modeled as a simple
sured commercial MESFETs and found an FOM of 500 fs.
parallel plate,
Published values by Blackwell [10] for a specially designed
switching MESFET give an FOM of 270 fs. These values Aε0 εr
have been achieved after about 20 years of development of Coff = , (11)
W
MESFETs as commercial devices. The FOM is limited by
the conductivity of the AlGaAs channel. The channel is heav- The second point to note is that Ron does not depend on area
ily doped, near the solubility limit, but it must be made thin of the conducting region, but only on length squared:
enough that the Schottky diode can pinch it off before break-
W2
down. There is a further limitation on the FOM because of Ron = , (12)
the planar structure of the electrodes which add fringing fields Q(µn + µ p )
making Coff relatively large. In fact Blackwell’s improvement where Q = IF τ, is the charge in the intrinsic region which
in the FOM was achieved by selectively thinning the substrate equals the injection current times the lifetime. The FOM is
in order to reduce the fringing field. thus
WAε0 εr
FOM = . (13)
P-i-n (µn + µ p )I f τ

The FOM can be reduced by decreasing W or A, i.e. by

The model of a p-i-n diode shown in Fig. 9 and 10 was decreasing the volume of the conducting region. Since
extracted from several sources [8, 11–13]. The first point to WA/(IF τ) is the reciprocal of charge density, eqn. (13) re-
duces to eqn. (10). While increasing charge density does im-
prove the FOM, what the model does not explicitly show is
LS LSD LD the dependance of carrier lifetime on carrier density. As car-
rier concentration in semiconductors increases, carrier-carrier
GATE interaction causes an increased recombination rate. We extract
from commercial literature Ron and Co f f for p-i-n diodes. At
SOURCE DRAIN

IF

n+
Area A
r r n
C C p+
r W Ron
Q
i
C
n+
W S W

FIG. 8: A sketch of a MESFET. FIG. 10: PIN diode in on state.

NRC publication no. 44453

6
5 mWatts, an FOM of 220 fs can be achieved and at 25 mWatt
that can be reduced to 110 fs. There are several reasons for the
slow reduction in FOM with increasing power. The increased
recombination rate as noted above is one reason, but there is
also the effect of contact resistance as values of Ron reach the
1 Ω level.
Photoconductive
Photoconductive switches have been used for very high-
speed optically controlled switching of microwave sig-
nals [14] and there has been an interest in making efficient
optically controlled switches [1]. A photoconductive switch
has the geometry of an interdigital capacitor. Under illumi-
nation photo-carriers generated in the semiconductor regions
between the fingers provide an Ron which closes the switch.
The structure of this switch is very similar to that of a MES-
FET in that the electrodes are planar. However, the isolation is
somewhat better because the gap is usually made longer than
the short channel length of a MESFET. The Ron is dependent
on the optical power, the carrier mobility and lifetime in a
manner similar to a PIN diode when optical power is substi-
tuted for bias current. Starting with eqn. (10), we can write
the FOM for the photoconductive region as follows:
1 h̄ω
FOM = ε0 εr Γ, (14)
µn qτ P
where µn is the electron mobility, q is the electronic charge, τ
is the carrier lifetime, P is the optical power density, ω is the
optical frequency, h̄ is Planck’s constant, and Γ is the quantum
efficiency. While we could could use this theoretical value for
FOM, for our comparison in Table I we take experimentally
determined values of Ron and Coff .
MEMS SWITCHES
There are both resistive MEMS switches that make metal
to metal contact and variable capacitor MEMS switches that
make metal to insulator contact. The MEMS switch that we
consider is the metal to metal contact type shown in Fig. 11.
The microwave structure is based on a gap in a coplanar
waveguide. A top contact suspended from an insulating beam
can be lowered to make electrical contact across the gap thus
closing the switch. Two large capacitors act as the actuator to
bend the bridge and lower the contact. This structure provides
a small area and large gap which results in a small capacitance
in the off state.
The simple material model shown in Fig. 7 is not an ap-
propriate starting place to derive a MEMS FOM, because ge-
ometry changes in a MEMS switch are used to effect control.
We will examine capacitance and resistance separately. The
contact resistance in a MEMS switch is much greater than
would be expected based on resisitivity of metals. It is be-
lieved that the higher resistance arises because contact actu-
NRC publication no. 44453
ally takes place at only a small number of “high” spots lead-
ing to a reduced effective surface area. For example, gold-
gold contacts with an area of 20 × 20 µm, a conducting length
of 0.4 µm and resistivity of ρ = 2.5x10−6 Ω · cm would give a
calculated resistance of 2.5x10−5 Ω. In practise, a resistance
of 0.22 Ω was measured. This is 8800 times greater than ex-
pected which we interpret as an effective area factor of about
10−4 . We calculate the capacitance of a single gap in Fig. 11
as C = ε0 A/G. For the on state resistance, we use an effective
area factor ae , a conducting length l equal to the thickness of
the contact films and a resistivity of ρ to give R = ρl/(A · ae ).
This leads to an FOM of
ε0 ρl
FOM = CR = . (15)
G · ae
Comparing eqn. (15) with eqn. (10), we see that εr has been
reduced to 1 giving an FOM reduction of 10−1 . The resistivity
of the metal is 10−4 less than the semiconductor giving a re-
duction of FOM of 10−5 . The effective area factor is approxi-
mately 10−4 meaning that much of the benefit of the material
change is largely lost due to the effective area factor. How-
ever, in contrast to eqn. (10), this FOM includes 2 geometry
factors, the gap and the conducting length. The conducting
length is the thickness of the films that make up the contact
which we cannot reduce without introducing loss in the signal
path to the contact. Thus, the principle degree of freedom that
is introduced by MEMS is the gap G. It accounts for a factor
of less than 10−1 in the MEMS switch of Fig. 11 if we con-
sider a 2 µm gap to represent a MEMS switch and an 0.4 µm
channel length to represent a MESFET. For our comparison
in Table I, we take experimental values from the literature to
determine FOM. Ron of 0.22 Ω has been reported for a contact
A
A
signal
GND GND
PECVD SiN
G metal
L
quartz substrate εr = 3.8
FIG. 11: Top is a sketch of a MEMS switch. A section across AA is
drawn in the bottom part of the figure.
 7

area of 20 × 20 µm. Off state capacitance values of 1 to 10 fF 0

have been reported for those same structures.
-10
There are three important conclusions to draw from the 1
FOM analysis of MEMS switches. First, there is 1 degree -20 2

off on
/S12
of freedom available to the switch designer that does not ex- -30 3
ist in a semiconductor switch. That is the ability to set the 4 and 5

S12
gap independently of the materials. The FOM can be reduced -40
7
by as much as the gap, or mechanical travel distance, can be -50 6
increased. Secondly, the change in dielectric from semicon-
ductor to air gives an order of magnitude lower value for ca- -60
pacitance. The third conclusion is that the effective area fac- -70
tor is the dominant “material” property. It is not known if the
-80
factor of 10−4 is a fundamental limit or if improvements can 1 10 100
be made. The study of the physics of metal to metal contact Frequency (GHz)
in small areas may be productive in future improvements in
FOM. FIG. 12: Graph showing the comparison of the FOM for a number
of RF switches. The labelled curves are for 1 - measured S21 NRC
opto, 2 - Opto 40 Ω, 80 fF, 3 - PIN 1 Ω, 110 fF, 4 - Opto 100 Ω,
30 fF, 5 - FET 5 Ω, 100fF, 6 - Rockwell MEMS switch, and 7 - 60 µm
COMPARISON coplanar waveguide gap on quartz measured at NRC. Gray boxes are
the switching ratio for mechanical coaxial switches.
A number of sources were used to compile the data that
is summarized in the following Table I. The capacitance and
resistance data was extracted from commercial information,
published papers or determined experimentally by the authors. first generation devices. This is predicted by our theoretical
The last entry was taken from ARPA’s website. analysis presented above and represents a general advantage
of MEMS devices over semiconductor devices.
A further comparison of the switching elements is pre-
off /Son
sented in Fig. 12. In this graph we plot the ratio of S12 12
for a switching element connected in a series configuration.
For the optical switch and the MEMS switches, this is a con-
figuration that might be be used. The curves for individual
p-i-n diodes and MESFETS were produced by simulation us- APPLICATION OF THE FIGURE OF MERIT
ing Touchstone models. The data for the models was extracted
from commercial data sheets or measured from sample tran-
sistors. The FOM may be of most use when trying to develop a
A point that this graph emphasizes is that semiconductor new switch device. Our use of the FOM was developed af-
devices have similar off-on ratios even though they do not ter attempting to invent a new type of semiconductor switch
function by the same principles. The MEMS devices shown without initially considering the high frequency AC behavior.
are about 30 dB better in switching ratio even though they are The assumption was that if a large change in DC resistance
could be produced, then it would be possible to find a width
to length ratio that would produce a useful switch. In work
which we have reported earlier [1], we developed an optically
TABLE I: Summary of devices used in the comparison. controlled semiconductor switch. What we found was that
Device Class FOM Power Cap. Res. the switching characteristics were similar to that available in
(fs) (mW) (fF) (Ω) other devices even though those devices functioned by appar-
ently quite different physics. This can be seen in Fig. 12 and
IMS-Small Opto 4000 5 80 50
is in fact predicted by eqn. (10) which we subsequently de-
IMS-large Opto 3000 5 30 100
veloped. Which shows that the limit of microwave switch-
NE3290 FET 500 0 100 5 ing behavior is largely predicted by the dielectric constant of
Blackwell AlGaAs FET 270 0 170 1.6 the material in the off-state and by the conductivity in the on-
MA4GP022 GaAs PIN 4.5 220 5 110 2 state. What eqn. (10) makes quite explicit is that there is no
MA4GP022 GaAs PIN 20 110 25 110 1 geometry that will improve this result. What is further demon-
Raytheon MEMS memb. 12 0 35 0.35 strated by eqn. (15) is that if the switch physically moves when
Rockwell MEMS cant. 2.5 0 11 0.22 changing from the off-state to on-state, we can introduce an
COM DEV Coaxial 0.07 0 0.35 0.2 “engineerable” degree of freedom into the device. So far, we
ARPA-proj. MEMS 0.01 0 0.05 0.2 have not found a similar degree of freedom available by using
conventional semiconductor engineering.

NRC publication no. 44453

8
FABRICATION OF MEMS SWITCHES
Design considerations will restrict the choices for fabrica-
tion processes for MEMS RF switches. Due to the nature of
microwaves, switch design must be integrated with the mi-
crowave circuit. The simplest and most efficient way to do
this is to build the switch on top of the circuit. The process of
building a MEMS structure on top of a substrate is called sur-
face micromachining. A general review of surface microma-
chining is given in [15]. In order to develop a reliable and low
cost process the techniques of photolithography and thin film
processing are used as much as possible. An important excep-
tion to this is that the MEMS part of the device must be freed
from the substrate. This involves selective etching of a sacrifi-
cial material under the MEMS part. There will also be restric-
tions on parameters such as process temperature and etching
techniques so that the underlying microwave circuit and sub-
strate are not damaged. A large number of design and fabri-
cation methods are being developed for RF MEMS switches.
An overview is given in [7]. There are two basic switching
mechanisms: capacitive coupling where the switching mech-
anism is due to a capacitance change and metal contacting
where opening and closing an ohmic contact is the switching
mechanism. At NRC, we have been developing a metal con-
tacting switch based on SiO2 or SiN fixed-fixed beam MEMS
structure and a polyimide sacrificial layer [2, 3].
Since the structure has free standing moving parts, the me-
chanical properties are of utmost importance. Properties such
as residual stress, Young’s modulus, and defects must be well
characterized and their effects on the mechanics of the device
understood. This is complicated by the extreme aspect ratios
of thin film components and by the relative importance of sur-
face properties due to the small scale of the devices. There is
still a great deal of research needed in these difficult areas of
mechanics of thin films.
APPLICATIONS OF MEMS SWITCHES
Of the many possible applications of MEMS switches, we
consider only 2 that reflect opposite extremes of the switch
market. Spacecraft applications demand the highest switching
performance and benefit by mass/volume reductions. At the
other extreme, wireless handheld phones use low cost semi-
conductor switches.
Satellite Applications
Fig. 13 illustrates a simplified block diagram of a satellite
payload. It consists of receive/transmit antenna, a beam form-
ing network (BFN), input filter assembly (IFA), high gain re-
ceiver (RCV), input and output multiplexers, high power am-
plifiers (HPA) as well as several switch matrices. A satellite
system of this type would typically have 100’s of switches
NRC publication no. 44453
RCV
IFA
LOW
INPUT OUTPUT POWER
RCV INPUT
SWITCH SWITCH SWITCH
MUX MATRIX
MATRIX MATRIX
IFA
RCV
HPA
HIGH
OUTPUT POWER
BFN MUX SWITCH
MATRIX
HPA
ANT
FIG. 13: A simplified block diagram of a satellite payload.
on-board integrated in the form of switch matrices to pro-
vide system redundancy. The receiver input/output and low
power switch matrices are typically implemented using coax-
ial switches while the high power switch matrix is imple-
mented using waveguide switches. The RF-MEMS technol-
ogy could be potentially used to build miniaturized switch
matrices to replace the bulky coaxial technology for the re-
ceiver input/output as well as the low power switch matrix.
One would expect to achieve more than one order of magni-
tude reduction in mass and volume by replacing the coaxial
technology with MEMS technology.
Such mass and volume reduction would have a dramatic
impact on the economics of a satellite program, since launch
costs are related to satellite weight. Alternativley, the mass
saved in the switch matrix could be replaced in other areas to
increase the capability to the satellite. That could be in the
form of increased capacity by adding payload electronics or
extended station keeping life by adding more fuel.
MEMS switches could be also potentially used in beam
forming networks (BFN), particularly, in the design of re-
configurable Butler matrices and phase shifters for the multi-
beam satellite communication systems.
Wireless Applications
MEMS switches can replace the SP2T and SP3T switches,
which are currently used in dual-band, and triple-band cell
phones. These switches are currently implemented using
semiconductor technology. The advantage of using MEMS
technology in this case would be RF-performance improve-
ment, which would in turn reduce dc power consumption.
This is a very low cost high volume application at present
filled by GaAs MESFET switches. A commercial MESFET
provides about 0.9 dB insertion loss which by itself consumes
about 19% of generated RF power. In principle a MEMS

FIG. 14: A switched filter bank for multi-band receivers.
switch could provide 0.2 dB insertion loss which reduces the
power loss to 4.5%. This improvement is a worthwhile en-
gineering objective, but as a consumer product it would have
to be available at about the same cost as a GaAs MESFET.
Presently GaAs MESFET switches cost less than $1 in quan-
tities of 100,000. There are as yet no similar commercially
available MEMS switches so that we cannot make a cost com-
parison.
A further requirement for this application is that the
switches must be specified for the environment found in a
hand held phone. The temperature range required is -40 °C
to 85 °C. The state of development of MEMS switches seems
to be still at a level of ensuring basic functioning. We have
not found references to RF characterization over the required
temperature range.
It is expected that future wireless receivers would need to
operate at several bands covering a wide range of frequencies
from 900 MHz up to 5 GHz. These receivers will have also
to operate in an environment of increasing interference. Such
requirements could be met with the switched filter bank as
shown in Fig. 14. The entire band to be covered by the re-
ceiver is divided into several channels where the appropriate
narrow-band is selected using a switched tunable-filter bank.
In view of current conventional technologies for RF filters,
a receiver with such capability would be bulky, power con-
suming and very expensive. However, with the use of MEMS
technology, MEMS switches could be integrated with MEMS
tunable filters to build the whole re-configurable receiver on
one single chip. The availability of high performance MEMS
switches and filters is a key to miniaturization of such type of
wireless receivers.
CONCLUSION
In comparison with the dominant semiconductor devices,
MEMS offers a fundamentally superior basis for developing
NRC publication no. 44453
9
microwave routing devices. This will become particularly ad-
vantageous as frequency is increased. There are numerous
unresolved problems of developing suitable fabrication tech-
niques and a need for more reliable switches. However, some
applications of RF MEMS switches are expected in the next
few years.
References
∗ Electronic address: peter.grant@nrc.ca
[1] P.D. Grant, M.W. Denhoff, M. Gao, and R.R. Mansour, “A
Photo-conductive Microwave Switch” in Conference Proceed-
ings of ANTEM98, Ottawa, ON, 1998, pp. 475-482.
[2] M.W. Denhoff, P.D. Grant, M.A. Harry, and M. Yu, “Fabrica-
tion of a microwave MEMS switch,” in Proceedings of ANTEM
2000, Winnipeg, 30 July - 2 August, 2000, pp. 79-82.
[3] P.D. Grant, L. Winchiu, S. Piek, M.W. Denhoff, M. Yu, and
M.A. Harry, “Microwave Performance of a MEMS Switch De-
signed for RF and Microwave Routing,” in Proceedings of AN-
TEM 2000, Winnipeg, 30 July -2 August 2000, pp.83-89.
[4] R.N. Tait, “An IC-compatible process for fabrication of RF
switches and tunable capacitors” in Can. J. Elect. Comput. Eng.,
vol. 25, no. 1, Jan 2000, pp. 26-28.
[5] J.J. Yao and M.F. Chang, “A surface micromachined miniature
switch for telecommunication applications with signal frequen-
cies from DC up to 4 GHz”, in Proc. Transducers 95 – Eurosen-
sors IX, Stockholm, Sweden, 1995, pp. 384-387 vol. 2.
[6] C.L. Goldsmith, Z. Yao, S. Eshelman, and D. Denniston, “Per-
formance of low-loss RF MEMS capacitive switches”, in IEEE
Microwave and Guided Wave Letters, vol. 8, no. 8, Aug. 1998,
p. 269.
[7] J. Jason Yao, “RF MEMS from a device perspective,” in J. Mi-
cromech. Microeng., vol. 10, no. 4, Dec. 2000, pp. R9-R38.
[8] S.M. Sze, Physics of Semiconductor Devices, 2nd. ed., John
Wiley & Sons, New York, 1981.
[9] Y. Ayasli “Microwave switching with GaAs FETs”, in Mi-
crowave Journal, vol. 25, no. 11, Nov. 1982, pp. 61-74.
[10] D.A. Blackwell, D.E. Dawson, and D.C. Buck, “X-band MMIC
switch with 70 dB isolation and 0.5 dB insertion loss”, in IEEE
Microwave and Millimeter-Wave Monolithic Circuits Sympo-
sium, Orlando, FL, USA, 1995, pp. 97-100.
[11] G. Hiller, “Design with PIN diodes,” Alpha Industries Applica-
tion note APN 1002, June 1999.
[12] “Applications of PIN Diodes,” Hewlett-Packard application
note 922.
[13] “MA4GP Series Data Sheet,” M/A-COM Semiconductor Prod-
ucts, Burlington, MA.
[14] D.A.M. Khalil and A.M.E. Safwat, “On the improvement of the
performance of the optically controlled microwave switch”, in
IEEE Trans. MTT, vol. 45, no. 8, Part 2, Aug. 1997, pp. 1358-
1361.
[15] J.M. Bustillo, R.T. Howe, and R.S. Muller, “Surface Microma-
chining for Microelectromechanical Systems,” in Proceedings
of the IEEE, vol. 86, no. 8, Aug. 1998, pp. 1552-1574.