So next example is electronic switches now
electronic switches we will be using to connect a line or to isolates another line And not
that before electronic switches we did not have this type of application we have some
other alternatives also so let us first discuss what are the basic characteristics of these
different types of techniques which are used to isolate input and output The first example
here based on circulator so in this typical example one single antenna is connected to
the transmitter side and the receiver side and now when this signal from transmitter
side it goes to circulator circulator it directs the signal to antenna and if any signal is
received by antenna then circulator it will direct in the same direction to receiver in
a clockwise direction it is showing so that is how we can isolate receiver and transmitter
so at any time communication is possible or it is a full duplex system Next this is one example of the diplexer so
here diplexer or diplexer it is formed by using filter bank one frequency is used for
transmission and another different frequency is used for reception so again it is a full
duplex system and it can isolate transmitter and receiver at any given time And the third
example is by using switches so in this case we can see that the single antenna it is connected
to transmitter side or receiver side uhh and its connection depends on the state of switch
So 2 important characteristics it should have very good isolation between the input side
and output side or (receive) receiver side and transmitter side because transmitting
power it is very high it can be of the order of a few watt and receiver side typically
receiver it is of the order of nano watt or even less than that so there might be very
good isolation between the transmitter to receiver otherwise there will be leakage And also when the switch it connects one line
let us say transmitter antenna it should provide very low insertion loss so this low insertion
loss and high isolation they are the characteristics of any solid state based switch Also we see
that it is not a full duplex system when we are using the antenna for transmission we
cannot receive any signal because the receiver path is open So here are comparison of the
advantages and disadvantages of these 3 different types of technologies Circulator we have load
pulling protection so right hand side whatever you connect whatever impedance may be it does
not have any problem to left hand side but it has the disadvantage that it is bulky because
it is (d) because it is made of some ferrites device so it needs biasing arrangement it
needs some magnet so that is why it becomes bulky expensive and typically this is also
a narrow band device Next is diplexer its advantage is its linearity
so whatever power we can we use low power or high power it will not give you any (non)
non linear property so this is the uhh main main advantage of diplexer and also it allows
full duplex system But the disadvantage is that it we have to design 2 band pass filters
having 2 different frequencies and now let us say 2 frequency bands they are closely
spaced so the rejection level at these 2 bands for these 2 different filters should be as
high as possible Sometimes the specified value it can be 80 dB or more So from filter theory
already we have seen that then it should we should use very high Q resonator to design
this narrowband filter with very high attenuation rate And if we go for very high Q or low loss
then we have to increase the size of the filter and it would become very bulky and expensive Next switch it is highly scalable in fact
it is possible to fabricate inside any integrated circuit in VLSI technology we can fabricate
hundreds of switches together But main problem is its linearity because in most of the cases
it would associate with some PN junction or even let us say Schottky diode and all of
them are associated with some sort of nonlinearities So typically if we go for a high power applications
in that case it will generate harmonics and signals itself it will be distorted so linearity
is a big problem for this type of switches And not only that we also have to provide
biasing arrangement and the control signals Already we have seen for the PIN diodes sometimes
how problematic it can be so biasing current also it can be very high typically sometimes
tens of mili ampere Now in a system if we have hundreds of switches so you can imagine
how much power it will consume So in this class we will mainly concentrate
on different types of electronic switches so because of their advantages compact size
and since they are highly scalable but already we have seen the disadvantage is that we need
biasing circuit control signal and linearity is a problem They are typically used in MIMO
structure TDMA system phase shifters array antennas voltage controlled oscillator et
cetera Here is a schematic example of a MIMO system we are using two antennas and we have
different receivers and transmitters so we are using multiple multi throw switch here
and that will connect to one antenna to one receiver side or ONe transmitter side depending
on requirement so we are using a switch bank here Another example this is a passive component
butler matrix it is used for phase array antennas Inside the butler matrix we have a combination
of several branch line couplers passive phase shifters crossovers Now the order of the butler
matrix it is 2 to the power n by 2 to the power n so we have 2 to the power n number
of input lines and 2 to the power n number of output lines Now antennas they are connected
to this all output ports and the RF input at a time it is connected to one of these
2 to the power N input by using a switch Now how it operates okay so let me just discuss
a few basic points on phased arrays So let us say we have one 2 antenna system
and this first antenna is connected to a phase shifter and the second antenna we are using
as the reference antenna Now the resultant beam direction it is in broadside direction
when the phase difference between 2 antenna feed is 0 Now if we introduce a phase shift
between these 2 antennas then beam direction will change and this change or new beam direction
it depends on the phase shift value So for an arrays of N antennas we assume it is a
linear array in that case we have to maintain the same progressive phase difference between
successive antenna to direct the beam in a particular direction Now if I go back to butler matrix so for butler
matrix when let us say the input is connected to the first input then we have a typical
phase distribution and because of that we have a progressive phase shift between the
output points If the input is connected to second point then this progressive phase shift
will change and the beam direction will change If we connect to third input point of the
butler matrix then again we have a phase distribution among the output points so we have a new beam
direction So that means for a 2 to the power n input or 2 to the power n output lines we
have 2 to the power n possible beam directions and at a given time the actual RF input is
connected to one of the input ports of the butler matrix so we need a filter bank at
the input side to control the beam position So these are some examples of applications
of electronic switches Now switch specifications for small signal
case the main characteristics are insertion loss so this is the loss between input and
output reflection loss seen at the input terminal isolation when the switch it is off we are
expecting very good isolation between input and output and then the bandwidth of operation
But when we go for large signal high power case in addition to all these points we have
to consider some extra points one for example what is the highest power handling capability
Next what distortion it produces then what is the 1 dB gain compression point if we keep
on increasing the power then output power at some point can decrease then what is the
effect of intermodulation So since because of the nonlinearities we may have distortion
or intermodulation effect and (be) why it is so important because it can generate some
other frequency components which can interfere at the receiver side so we have to be very
careful about these points when we go for high power applications Now different types of switches according
to the number of points to which this switches (connect) connected and the number of points
where it can connect It can be named as single pole single throw switch in this case the
switch is connected to just 1 point or it has one pole or ONly one switching position
is possible one single ON or OFF or we call a single throw position or simply SPST switch
The next one is single pole double throw switch so switch that is connected to one point and
it can connect either the first line or second line so 2 throw positions are available we
call it SPDT switch In general we can define Alpha pole Beta throw switch or Alpha p Beta
t switch Now small signal model of a switch already
we have seen that a PIN diode or a Schottky diode we can use as switch Typically PIN diode
is used at low frequency operation and switching speed is low and Schottky diode is used for
very high frequency switching Now when the diodes are in forward biased condition in
simplified form we simply represent by them a resister of small value of resister and
when they are in reverse biased condition we replace the diodes by a small value of
capacitor so this capacitor value and the resister value is defined by the uhh by the
companies it depends on fabrication procedure and the materials they are what the materials
they are using Now how to connect the switches in practical applications We can connect them
in series configuration we can also connect them in shunt configuration in the previous
example we use PIN diode in shunt configuration So in shunt configuration when the PIN diode
switch is on that means we will be representing the PIN diode by a small value of resister
then it will be simply RF signal it would be grounded so let me draw a diagram better
So let us say we have a microstrip line I am showing just the top view of the microstrip
line and this left hand side is the RF in point and the right hand side it is RF out
point and now I am using a switch in series configuration sorry in shunt configuration
so it is connected between the microstrip line and ground Now when the diode is ON in
that case we can represent the diode by a small range equivalent resistance r ON typically
r ON is just a few ohms So in this case what will happen since the resistance is just a
few ohms so RF signal whatever coming from left hand side it is grounded through the
register and we have very good isolation between input point and output point since we do not
have any microwave signal or millimetre wave signal present at the output Now consider the second scenario when this
diode is switched OFF In that case we can represent the diode by a small value of capacitor
so C typically a few Picofarads so in that case since it is a small value of capacitor
so equivalent reactance will be very high for the RF signal and whatever RF signal we
have it will simply pass from left hand side to right hand side or we have direct connection
between the input and output so this is in shunt configuration We can use the same diode
in series combination as well so in that case the switching condition will simply change
so how to connect in that case we have the same microstrip line but now we are using
diode in series combination so we have RF in left side a load may be RF out right hand
side when the diode is switched ON we will be representing the diode by a small value
of resistance So when the resistance is connected since
it is a very small value it will connect simply left hand side to right hand side so r ON
in series it will be it will give you some insertion loss and we see in this case left
hand side is connected to right hand side you can compare this situation with the previous
one and we have transmission from left to right just opposite what we have seen in previous
case Now let us say the diode it is in OFF condition so we will represent the diode by
a small value of capacitor here and since we have very small value of capacitor it is
associated with very high reactance and we have high impedance as seen by the RF input
port we have input reflection from this capacitor and we have very good isolation between the
input and output point Now the question is that let us say one type of switch is given
it can be one PIN diode or it can be one Schottky diode then should we use in series configuration
or shunt configuration which one should we prefer Then we have to do some analysis and after
analysis the conclusion is that in shunt configuration if I use a single diode it would provide better
isolation compared to that in series configuration considering that the shunt capacitance typically
a few picofarad and ON resistance is a few ohms So how we do the analysis so let us see
we are considering SPST switch let us first consider series configuration so as I discussed
it is modelled by a resister in ON state and by a capacitor in OFF state So in ON state
then looking at the circuit whatever I have drawn previously we can easily calculate what
is the S 21 so if I go back to my drawing so we can consider it as a 2 port device right
where we have port 1 here port 2 here and this is the section of transmission line this
is another section of transmission line and in between we have a resister connected to
ground So simply by multiplication of ABCD matrices
and then converting it to S parameters we can calculate what is the S 11 and S 21 of
this arrangement Obviously when we implement any switch there will not be any one single
PIN diode or Schottky diode in addition to this we should have DC blocking and RF blocking
arrangement and for that we have some additional capacitors and inductors but at RF frequencies
those capacitors or inductors we will be representing by short circuit or inductor by open circuit
and capacitor by short circuit So not only for this case for the second case also we
can calculate what is the S 11 and S 21 So let us say the Port impedance it is given
by Z 0 in that case when the switch is ON we are considering series configuration now So in series configuration when the switch
is ON left hand side is connected to right hand side and we replace the switch by a few
ohms resister which is r ON If you calculate S 21 that is equal to in dB 20 log 10 of 1
plus r ON divided by twice Z 0 and in OFF state that means when the switch is replaced
by a small value of capacitor if I calculate S21 so isolation I series that is 10 log 1
plus 4 Pie f c off Z 0 hold to the power minus 2 in dB so we have actually one half term
1 square term here because of that the 20 log it becomes 10 log so a typical example
for a maximum insertion loss of 0 point 34 db r ON equal to 4 ohms If I look at the expression
insertion loss it depends on r ON isolation it depends on c off and operating frequency
So if I increase the operating frequency for a given switch we will what we expect decrease
isolation because capacitive reactance it decreases with increase in frequency Now let us consider shunt spaced switch in
this case we are using the PIN or Schottky diode in shunt configuration So in ON state
now we have to represent diode by a capacitor that means in this case left hand side would
be connected to right hand side and we call the switch is ON but diode is OFF and S 21
or insertion loss in shunt IL shunt that is equal to 10 log 1 plus pie f c off Z 0 equal
to the s s square in dB And in OFF state isolation I shunt 20 log 1 plus Z 0 by twice r ON So
we see here now insertion loss it becomes a function of frequency whereas the isolation
it is a function of r ON so for very high frequency operation we have to be careful
depending on the values of r ON and c off and we can choose a proper circuit implementation
proper implementation scheme it can be series or it can be shunt So already we I I discuss that for a few (kep)
picofarad capacitance and considering that a few ohms r ON mostly in shunt configuration
these switches provide better isolation So here are some examples we can design SPDT
switch by using a SPST switch here the example is shown here One antenna it is connected
to transmitter side and also receiver side so depending on switching condition we can
direct signal from transmitter to antenna or antenna to receiver and this is the right
hand side it shows the simplified equivalent model We are representing the ON condition
by r ON and OFF condition by c off and transmitter side so at millimetre wave frequencies or
even at microwave frequencies if nothing is specified it is 50 Ohms system Or left hand
side that is why it is represented a 50 Ohms right hand side it is also represented by
a 50 Ohms load antenna input impedance when matched that is also 50 Ohms Now how to improve the isolation for a given
switch because it is a problem at higher frequencies A PIN diode let us say is providing a very
good isolation of 40 dB at 1 gigahertz if I use the same PIN diode above 30 gigahertz
isolation it can degrade just 2 3 dB so isolation improvement is a challenge for any given switch
So next step if we have two PIN diodes or two Schottky diodes we can use them in series
shunt configuration to improve the isolation value so what we will be doing here let us
say the switch we are going to use in series shunt configuration so one PIN diode we will
be connecting in series configuration and another one in shunt configuration okay so
let me draw it So this is the microstrip line top view one
diode will be used in series configuration and the second diode we will use in shunt
configuration and together it behaves as a single switch SPST switch Now if I want to
connect left hand side to right hand side then what should be the conditions of the
diode I want to connect left hand side to right hand side when the switch is ON so in
ON condition this diode it should be replaced by a resister r ON small resistance value
so this diode should be in forward bias condition and this diode it should provide good isolation
so we should represent the diode by a capacitor c off or the second diode it should be in
reverse bias condition Then in OFF state it is just opposite in OFF
state then this first diode it should be in reverse bias condition and the second diode
it should be in forward bias condition so we have isolation due to first diode and further
isolation for the second diode that is how we can improve isolation here but insertion
loss it will degrade But anyway for any practical application we mainly look at the difference
between the insertion loss and isolation so now again if we calculate S 21 for this case
in ON state insertion loss or S 21 is given by this expression it now becomes the function
of both r ON c off and frequency And the isolation or the switch is OFF in that case S 21 is
equal to 10 log of this function so it is also a function of frequency capacitance and
r ON series resistance So for any given switch how good it is to
understand that we define a parameter we call it figure of merit for a given first order
switch r and c so this is simply r ON multiplied by c off just like one r c Circuit the only
the unit is second time constant but r in ON state c in OFF state so smaller is the
figure of merit better is the switch here are some comparisons So insertion loss versus
for different figure of merits so we see that for 25 femtoseconds r ON c OFF at even 140
gigahertz the insertion loss is below 1dB for 200 femtoseconds it degrades to almost
4dB we had considered r ON equal to 4 Ohms Now right hand side figure it shows the variation
of isolation for 200 femtoseconds you see at 30 gigahertz isolation at least we have
7 8dB but at 140 gigahertz it is less than 5dB we do not have any isolation If we go for much better switch 25 femtoseconds
let us say in that case at least we have 15 13 to 15 dB isolation at 140 gigahertz So
isolation insertion loss both depends on figure of merit r ON into c off okay so next day
we will continue we could not finish the switch here so next day we will continue another
10 15 minutes on switch then we will start the millimetre wave propagation path thank

Leave a Reply

Your email address will not be published. Required fields are marked *