4
Applications Information
PIN Diodes
In RF and microwave networks,
mechanical switches and attenua-
tors are bulky, often unreliable,
and difficult to manufacture.
Switch ICs, while convenient to
use and low in cost in small
quantities, suffer from poor
distortion performance and are
not as cost effective as PIN diode
switches and attenuators in very
large quantities. For over 30 years,
designers have looked to the PIN
diode for high performance/low
cost solutions to their switching
and level control needs.
In the RF and microwave ranges,
the switch serves the simple
purpose which is implied by its
name; it operates between one of
two modes, ON or OFF. In the ON
state, the switch is designed to
have the least possible loss. In the
OFF state, the switch must exhibit
a very high loss (isolation) to the
input signal, typically from 20 to
60 dB. The attenuator, however,
serves a more complex function.
It provides for the
“
soft
”
or
controlled variation in the power
level of a RF or microwave signal.
At the same time as it attenuates
the input signal to some predeter-
mined value, it must also present a
matched input impedance (low
VSWR) to the source. Every
microwave network which uses
PIN diodes (phase shifter, modula-
tor, etc.) is a variation on one of
these two basic circuits.
One can see that the switch and
the attenuator are quite different
in their function, and will there-
fore often require different
characteristics in their PIN diodes.
These properties are easily
controlled through the way in
which a PIN diode is fabricated.
See Figure 9.
Bulk Attenuator Diode
Epi Switching Diode
P+ Diffusion
Bulk
I-Layer
N+ Diffusion
Metal Contact
Contact Over
P+ Diffusion
N+ Substrate
Epi
I-Layer
Figure 9. PIN Diode Construction.
RF COMMON
RF 2
RF 1
BIAS
2
3
4
1
2
3
4
1
3
4
1
2
Figure 8. Very High Isolation SPDT Switch, Dual Bias.
Diode Construction
At Agilent Technologies, two basic
methods of diode fabrication are
used. In the case of bulk diodes, a
wafer of very pure (intrinsic)
silicon is heavily doped on the top
and bottom faces to form P and N
regions. The result is a diode with
a very thick, very pure I region.
The epitaxial layer (or EPI) diode
starts as a wafer of heavily doped
silicon (the P or N layer), onto
which a thin I layer is grown.
After the epitaxial growth, diffu-
sion is used to add a heavily doped
(N or P) layer on the top of the epi,
creating a diode with a very thin I
layer populated by a relatively
large number of imperfections.
These two different methods of
design result in two classes of
diode with distinctly different
characteristics, as shown in
Table 1.
As we shall see in the following
paragraphs, the bulk diode is
almost always used for attenuator
applications and sometimes as a
switch, while the epi diode (such
as the HMPP-3890) is generally
used as a switching element.
Diode Lifetime and Its Implications
The resistance of a PIN diode is
controlled by the conductivity (or
resistivity) of the I layer. This
conductivity is controlled by the
density of the cloud of carriers
(charges) in the I layer (which is, in
turn, controlled by the DC bias).
Minority carrier lifetime, indicated
by the Greek symbol
τ
, is a
Table 1. Bulk and EPI Diode Characteristics.
Characteristic
EPI Diode
Bulk Diode
Lifetime
Short
Long
Distortion
High
Low
Current Required
Low
High
I Region Thickness
Very Thin
Thick