NanoMountTM Switches For Microwave
This MicroNOTE was written to help switch designers Introduction To PIN Diodes As Microwave
interface Microsemi’s NanoMountTM PIN diode assemblies Power Control Elements
with the bias (control) circuits and to understand the DC
power requirements of the bias circuits used to turn the PIN A microwave PIN diode is a semiconductor diode that
diodes on and off. operates as a current controlled variable resistor at
RF/Microwave frequencies. If the forward bias current is
The first section of this MicroNOTE is an introduction to varied continuously, it can amplitude modulate or attenuate an
silicon PIN diode RF control circuits with an emphasis on the RF signal. When the control current is switched on and off,
RF power switching function. the PIN diode can be used for switching, for pulse modulation,
or for phase shifting an RF signal. We are concerned here
The discussion starts with the basic PIN diode equivalent with the switching function only.
circuit parameters that relate to the PIN diode switch
performance characteristics in VHF band. Both the control The use of a semiconductor diode as a switching element in
(DC bias) and RF power aspects are discussed in detail. RF/Microwave circuits is based on the difference between the
diode's forward and reverse bias impedance characteristics.
Next we discuss how these switch circuits and circuit elements At low microwave frequencies, the diode (including its
change as operating frequencies increase from VHF to 2GHz package parasitics) appears as a very small impedance (Rs)
and finally, beyond 2GHz to the present WLAN bands under forward bias and as a very large impedance (Xc) under
(2.4GHz to 5.8GHz). reverse bias. PIN diode switches rely on a difference in power
reflection, rather than power dissipation, to perform the
We then present three types of NanoMount assemblies that switching function
Microsemi Corproation offers the microwave industry. The
major RF and biasing features of these assemblies are Semiconductor diode switches rely on a difference in power
emphasized and performance is provided. reflection, rather than power dissipation, to obtain switch
performance. Both the low and the high impedance diode
Finally, we look at some methods to broadband the RF and states can reflect RF power in a 50W line, depending on the
bias circuits, the treatment of resistive loss in switch circuit circuit configuration.
elements, and the importance of reliable electrical grounding
for all shunt elements in the switch design. The basic PIN diode switch is reflective since the low and
high impedance diode states reflect power in a 50 line.
For economy of space, we will be referring to Reference Matched PIN diode switches are structured to provide a
throughout the remainder of this paper. The concepts we need separate path for the reflected power (usually absorbed in a
are developed there in detail. We ask to you download termination) so the signal source does not see the reflected
chapters 1 & 2 and appendix A from www.microsemi.com so power. Circulators and hybrid circuits are used for this
as to have them available as you read AN-710. If you would purpose.
like a copy of the PIN Diode Handbook, Version 2, please
send your request to email@example.com or PIN diodes are current driven devices. This must be
firstname.lastname@example.org and thanks! considered when forward and reverse bias switch driver
supplies are selected. In the forward bias state, the bias
supply must be able to deliver 5mA, 10mA or more, from a
fairly high impedance current source. This bias current causes
DC induced charge to be stored in the diode’s I-region causing
the diode to exhibit low impedance to the RF circuit. When
the reverse bias is applied to the PIN diode, the diode is turned PIN diodes to the electrical characteristics that are used to
off (non-conducting, high impedance state). But the diode is specify switch performance. The equivalent circuit parameters
not really off (high impedance) until all the stored charge is are unique to the physical structure of the thick I-region PIN
removed from the I-region. Thus the reverse bias supply must diodes that allow large values of microwave power to be
exhibit low enough impedance to the diode that the stored controlled by relatively small amounts of bias control power.
charge can flow quickly to ground. There is a later section on
the design of bias injection circuits that separate DC bias from
the RF circuits.
PIN Diode Equivalent Circuits As Related To
PIN Diode Equivalent Circuits figure 1a figure 1b
The concept of an equivalent circuit is fundamental to
electrical circuit analysis. To explain briefly the meaning of The PIN diode equivalent circuit parameters, shown in figure
this concept, let’s consider the following situation – Circuits A 1a and figure 1b, are the series inductance (LS), the forward
and B are connected at a pair of terminals a,b. Circuit A may biased series resistance (RSF), the reverse biased diode
be a voltage generator with its internal resistance and circuit B capacitance (CD), the reverse biased resistance (RSR), and the
may be a complex load circuit containing a Silicon PIN diode. minority carrier lifetime Tau (t). t is related to the charge
“Complex” means that the PIN diode may only be represented stored in the I-region, QS.
accurately by a passive R, L, C network over a finite range of
power, temperature, and frequency. Power transfer between
circuits A and B is accurately described by the terminal
voltage Vab and the current flowing into “a” and out of “b” or Bias
We now attach a simpler ensemble of circuit parameters, RF
called circuit C, to circuit A, at terminals a,b. Circuits B and Choke DC
C are said to be equivalent if circuit C produces the same V, I- Block
response when connected to circuit A that circuit B produced DC
when connected to circuit A, over the prescribed range of Block RF
frequency, time, and power. There is no uniqueness implied in Z0 Z0
defining circuit C. There are obviously a number of
equivalent circuit configurations that could satisfy the cited
A PIN diode equivalent circuit is shown in with, the reverse
bias (OFF) circuit in figure 1a, and the forward bias (ON)
circuit in figure 1b. Figure 1a shows the package inductance
(LP), the package capacitance (CP), and the PIN chip Bias
inductance (LS), junction capacitance (CJ), the forward biased
series resistance (RS) and the reverse biased parallel resistance
(RP). The simplified reverse and forward biased equivalent RF
circuits are adequate to represent the electrical behavior of the
packaged PIN diode over the prescribed range of dynamic
circuit variables. DC
PIN Diode Switch Performance Parameters
To understand how PIN diodes control microwave power in a
switch circuit, we relate the equivalent circuit parameters of figure 2b
The switch performance parameters are the switch isolation, filter. Values of L and C can be calculated using reactance
insertion loss, VSWR, power dissipation, and switching times, formulas in. The pass band of RF/DC isolation circuits may
which are related to QS and t. These are related to the switch span several octaves if properly constructed.
performance parameters via the equations accompanying
figure 2a for the series SPST switch and figure 2b for the Figure 1 shows two other reactive elements appear in the RF
shunt SPST switch. Signal distortion and dynamic range are portion of the series switch circuit. An additional coil is used
related to t as well and will be discussed in a later section. to complete the DC path through the PIN diode to ground and
two capacitors are used to block the DC current from flowing
through the RF generator and the RF load. The DC
Bias Current Injection Circuit Elements “blocking” capacitors and the DC “return” coils are in the path
of the RF fields. The blocking capacitor is in series with the
To function in RF power switch circuits, PIN diodes need to transmission line and the DC return shunts the transmission
be biased “on” (low impedance state) and “off” (high line. The choice of L and C is not trivial since they are
impedance state). To turn the PIN diode on, forward bias is perturbations in the RF circuit and affect the transmission line
injected at the appropriate diode terminal. To turn the PIN VSWR. Reference gives plots of VSWR as a function of
diode off, the reverse bias voltage is applied across the diode frequency for values of L and C that are chosen for VSWR <
terminals. In figure 1, when the diode is “on”, the series 1.5 for 0.1 < f < 20GHz.
switch is ON, whereas, in figure 2, the shunt switch is OFF,
when the diode is “on”. Yet another approach is to use an APC-7 Hewlett Packard bias
tee to inject bias current into a series mounted PIN diode and
The bias circuits, that provide the current path and reverse bias another tee for the ground return. These tees exhibit excellent
access to the PIN diode terminals, must be isolated from the broadband performance from 10MHz to 18GHz (depending on
RF circuit, and yet be introduced across the same diode the model chosen). Test fixtures are available that can be
terminals as the RF signal. RF/DC isolation is needed to altered to accommodate series or shunt mounted diodes. The
inhibit RF current from flowing into the DC bias circuit, shunt diode requires one bias tee to feed the bias current and a
which would cause the loss of some RF power from reaching DC block in series with the load impedance. This approach is
the load and would possibly detune the RF circuit, as well. recommended for device evaluation and gives the designer a
feel for how the PIN diodes respond to various bias conditions
There are several ways of approaching the RF/DC isolation and RF power levels.
circuit design. The classic method is to employ filter theory.
The bias network is a low pass filter having a specified pass Nothing has been said, so far, about the band rejection
band, a transition band roll off characteristic, and some value response (leakage) of the bias filter. The out of band response
of out-of-band isolation. If this RF/DC isolation is not large of the bias filters should meet the minimum DC/RF isolation
enough, another filter section may be needed. The network requirement across the entire RF pass band. If there are non-
isolates the RF current from the DC bias supply provided the linear RF devices in the RF signal path, the filter responses at
RF pass band does not include the isolation network’s pass any significant harmonics must be accounted for.
band. Reference, chapter 5, is an excellent introduction to
Microwave and RF filter design principles. Filter elements (L, C) are shown as lumped circuit elements in
figures 1 & 2 but they may be realized in the microwave
Another approach is more intuitive and perhaps more bands by open or short-circuited sections of transmission line,
instructional. Referring to figures 1&2, we see that bias built in coax or stripline.
current is injected into the diode through an inductor (L) in
series with the bias supply terminal and attached to one diode NanoMount Switches Products
terminal. L is a low resistance path to the DC bias current but
L is chosen so that its inductive reactance is large through the Introduction
RF signal pass band (blocking the RF current). L is called an
“RF choke coil”. The NanoMount series of PIN diode assemblies or matrices
offers three unique circuit configuration consisting of four
Isolation is improved by the addition of a by-pass capacitor MPP4203 PIN diodes (MMSMTM products) and an
(C) from the bias supply terminal to ground. C is chosen to interconnect pattern on an Enhanced Performance Surface
be nearly a short circuit to RF current and an open circuit to Mount (EPSMTM) circuit board. These configurations are the
the bias current. This configuration has the appearance of an MNM4200 (series/shunt switch), the MNM4210 (ring circuit),
L-section filter (as it is). If the isolation is not adequate, and the MNM4220 (star circuit). Each has particular
additional L-sections can be added to improve it. Two L- characteristics that are optimum for specific switch
sections can be designed to be a p-section or a Tee-section applications, as discussed farther.
MNM4200 TRANSFER/ DUAL DIVERSITY SWITCH
SINGLE POLE DOUBLE THROW SWITCH RF2
DC1 D1 D4 DC2
D4 D3 RF3 RF4
GND DC3 DC4
figure 3a figure 3b
MNM4200 - Series/Shunt Switch (Figure 3a). MNM4210 – Ring Configured Transfer / Diversity Switch
The usual application for this series/shunt configuration is the
Single Pole Double Throw Switch (SPDT). A single input The MNM4210 is a ring configuration of four MPP4202 PIN
port can be connected to either of the two output ports. This diodes. This circuit is less dependent on the adequacy of the
switch functions reciprocally, as well. Either of two inputs can RF ground than the MNM4200 series/shunt circuit. We will
be switched to a single output. The series/shunt circuit cite a number of terminal conditions that define various switch
provides the best broadband performance for SPDT switches. applications that can be satisfied by the PIN diode ring circuit.
The only drawback is that the DC bias circuit is configured in
a ring. This may cause negative effects both on switching Four Port Usage - Transfer Switch
speed and on power handling.
This configuration has the maximum circuit flexibility. A
Since each switch arm is independently biased, two switch transfer switch is obtained if the switch drivers (that supply
drivers are needed to operate the MNM4200. Each driver the diode bias voltages and currents) are sequenced to activate
must be capable of supplying the required forward bias current the PIN diodes such that port 1 connects to port 4 and port 2
in either polarity, because each switch arm contains both a connects to port 3. Thus, D2 and D3 are biased ON and D1
series and a shunt connected diode. The series diode draws and D4 are OFF. Conversely, with D1 and D4 biased ON and
negative current when the switch is OFF and the shunt diode D2 and D 3 OFF, the transfer switch attains its alternate state.
draws positive current when the switch is ON. A true transfer switch allows a receiver and transmitter to
operate simultaneously (full duplex) assuming the switch
The MNM4200 has potentially the highest broadband isolation structure provides the required isolation between receiver and
of the three NanoMount switch configurations. The transmitter.
series/shunt also offers improved harmonic performance if
biased properly. The shunt diode in each arm provides Three Port Usage (SPDT Switch)
improved isolation if the RF ground provided for the switch
assembly is adequate. A three port switch configuration is obtained if one port is
terminated. If port 4 is terminated, port 3 can be fed as the
The MNM4200 can also serve as a non-reflective SPST switch input port with ports 1 and 2 serving as the switched output
if one of the switched arms is terminated. The isolation in this ports. Since the ring circuit has geometric symmetry, port 3
configuration is much higher that than achievable with a could be terminated and port 4 could act as the input port, and
single shunt diode SPST switch. so on.
Two Port Usage (SPST Switch) Single Pole Double Throw Configuration
A SPST switch, with high isolation properties, is obtained If port 4 is grounded, and input signal to port 1 can be routed
with the input at port 1 and the output at port 4. Both ports 2 to either port 2 or port 3 by applying the proper bias
and 3 should be terminated to avoid mismatch reflections conditions to D2 and D3. If D1 and D2 are ON and D3 is
within the switch structure. Since symmetry considerations OFF, the signal is routed from port 1 to port 2 and so forth.
still apply, other combinations are also available. This bias sequencing would be programmed into the switch
Signal Summing Circuit
Single Pole Three Throw Configuration
If D1 and D2 are ON with D3 and D4 OFF, a signal input to
port 1 feeds ports 2 & 3 and a signal input to port 4 also feeds If all the switch arms are biased OFF, the isolation is better
ports 2 & 3. But ports 1 & 4 cannot be treated as a single port than the ring circuit.
without destroying the symmetry of the ring circuit. Note
that the ring circuit has no common junction or neutral node.
Broadband Microwave Switches
MNM4220 The MNM4200, 4210, & 4220 switches are inherently
JUNCTION/ DIVERSITY SWITCH
broadband due not only to symmetry but to the miniature size
of the EPSM circuit board and the MPP4203 PIN diodes. The
bandwidth of these switches, used in conjunction with other
circuit functions, is also dependent on surrounding
D1 D3 components that are much more band limited than the switch
DC1 DC2 structure themselves.
The bandwidth of the embedded RF switch is limited to the
circuit designer’s ability to design broadband bias tees for bias
injection to the PIN diodes being switched ON and OFF. The
Radio Frequency Choke (RFC) coils in Figures 3a,3b,3c, are a
major problem. These RFC’s and their associated by-pass
capacitors are low-pass filter structures that are characterized
by in-band insertion loss and cut-off frequency.
There are a number of techniques for broad-banding bias tee
coils. One calls for the substitution of the coil by a 500W
resistor (for a Z0 = 50W circuit) thus resistively de-coupling
figure 3c the diode from the bias supply with a by-passed resistor
instead of de-coupling with a resonant structure. More bias
supply voltage is needed to account for the voltage drop
(I^2R) across the de-coupling resistor. Alternative approaches
MNM4220 – PIN Diode Star (Figure 3c) are to combine a resistor and a RFC or to use coils with a
distributed resistive loss.
The star configuration of PIN diodes has a node common to
the four switched diode arms. This feature provides many PIN Diode Switch Drivers
unique features that differ substantially from the ring
configuration. Forward bias current must be supplied by a stiff current source
– usually not a TTL Logic Source. This means that there is a
Single Pole Single Throw Configuration minimum of supply voltage droop with increase of bias
current supplied to the diode.
If ports 2 & 3 are grounded, port 1 is the input and port 4, the
output. This is a high isolation configuration since the two Reverse bias voltage sources must be capable of supporting
ON diodes are in series. From symmetry considerations, ports the efflux of current from the PIN diode’s I-region as the
1 & 4 could be grounded and ports 2 & 3 could be the active diode turns off (becomes an open circuit). The diode has not
ports. completely turned off until all the charge carriers are removed.
The reverse bias supply must make available a DC path for
this I-region discharge current.
In this MicroNOTE, we have explored most of the technical
details associated with the use of Silicon PIN diodes as
microwave power control elements. It is particularly
important to understand that both bias voltage polarities are
necessary to turn on and completely turn off, a PIN diode
switch. The PIN diode is a current driven device in the
forward bias state and must be biased from a DC current
source. To reverse bias the PIN diode completely off, a
relatively low resistance DC path must be provided by the bias
source, so that the charge stored in the I-region during the on
state can be completely removed. This is the seminal issue to
re-emphasize at the end of this paper.
The NanoMount Switch Products (MNM4200 series) require
switch drivers and RF/DC isolation circuits to function as
described in the prior sections of this paper. If biased
properly, the NanoMount switch provides extremely low
signal distortion performance and can control much higher
power levels than competitive MESFET switches.
 W.E. Doherty, Jr and Ronald Joos, PIN Diode Circuit
Designers Handbook, Version 2, 1999, Chapters 1 & 2
and Appendix A
 David M. Pozar, Microwave and RF Design of Wireless
Systems, Wiley, 2001, Chapter 5
 Joseph F. White, Microwave Semiconductor Engineering,
JFWP, E. Orleans, MA, 1995
 Kenneth R. Philpot, “High Performance Surface Mount
Diode Technology for Microwave Wireless Circuit
Applications”, MicroCurrents, Winter 2000, pp 8-9