Lecture 33 - Stability and Amplifier Design - II

IIT Roorkee July 201
61 followers
Feb 27, 2024
Concepts Covered: Condition for Stability in an Amplifier, Amplifier Gain, Power Gain, Transducer Gain, Design Targets.
welcome once again to ARF transer design
course we are discussing about about
stability and amplifier design so for uh
stability uh what we have done in the
last lecture if a two Port network is
available and If there is a s parameter
that we know after the biasing of the
circuit if we know the S parameters
based on the S parameter we can find the
uh input and output stability circles
and that Circle will tell us that where
should we uh uh how should we do the
input or output matching so that so that
we could we could not have a uh unstable
so your amplifier will not be a
potentially unstable so that is called a
uh potentially unstable and when you
have a stability it is a uh stable that
uh our amplifier is a
stable now there is another uh method
also considering let's say there is a
circle uh we have drawn stability Circle
that is a one method by using equation
also you can find out that there is a uh
your system or amplifier is
unconditionally stable or uh potentially
unstable so for that there is a k Delta
test or mu stability test so there are
two uh equation that is K and Delta so
if you know the S parameter and uh if
you put in this equation that is a k and
this is a Delta so if K is greater than
1 uh and Delta is less than one so that
is a uh your system is unconditionally
stable if this is less than one and if
if this is uh greater than one then it
is a uh not unconditionally stable it is
a unstable uh system in mu stability
test there is a here in K Delta you
required a two equation
however in mu you required only one
equation this one equation if your mu is
greater than one for Entre frequency
range in which you want to design your
amplifier then it is a unconditionally
stable so the advantage of m test is
that only a single parameter needs to be
evaluated there are no auxiliary
condition like K test the derivation
earlier and then Delta test so this is
a uh methodology by which you can also
find the your system is stable or
unstable and uh when when you draw the
SMI chart because uh in the SMI chart
you have advantage that uh you are also
going to use a SM chart for impedance
matching you can easily cor correlate
the SM chart so this is the two ways of
doing impedance
matching or finding the stability
now we will move to the power gain
equation and this is useful in the any
amplifier design micro amplifier design
or high frequency amplifier design if
you're designing on chip also only the
here the uh there is a two Port black
box is there now this two Port Black Box
can be replaced with any transistor uh
it can be Gan or it can be gas based
transistor we already have discussed the
multiple Technologies by which you can
uh make the Active Components it can be
a transistor Network Also let's say for
example you have a cost
code
so so this is a some bias you have given
there is a
input and this is
a two transistor Network there is a
output is also there so this output can
also have a bias T so there is a biasing
unit to apply here and uh you can also
get the S parameter of this transistor
Network it can be differential amplifier
also but you should prior uh simulated
or you should know the S parameter of
your network uh if you know the
parameter of your network you can find
out the power gain equation so what is
the aim of amplifier design let's say
you have identified the transistor
Network or a transistor then what is the
aim how to select gamma s and gamma L
reflection coefficient at the source
side that is this and reflection
coefficient at the load side that is
that GMA L how you have selected gamma s
and gamma L at a given frequency to
design the input matching Network and
output matching Network so this is
called imn that is input matching
Network and omn that is a output
matching Network Network so this is your
transistor or transistor
Network this is a uh gamma
s uh there is the gamma in so uh the
here the this is a gamma s on this side
so it will be something like this and
then gamma
in so gamma in that you can you look
from this side that is a gamma in you
look from this side that is a gamma out
so what what will happen when you find
out the gamma in we already have find
out in the last class by using the M
centic rule that how to find get the
gamma in and Gamma out uh once you
because gamma in is depend upon the
gamma L and Gamma out is depend upon the
gamma s so our aim is to design the
input matching Network and output
matching Network and there are different
Power gain equation based on what you
are considering you are considering the
completely input side considering the
source or you are considering the uh
completely output uh side so that by
that we can Define the power gain
equation so so these are the amplifier
gains so this is a figure of different
figure of Merit to assist an amplified
design so the first gain is a GP that is
is a power delivered to the load divide
by power delivered to the network so
that is given by
PL divide by P in so here you will have
a you should consider only the this
much portion that is a load
site and the network then uh GA that is
a power available from Network and power
available from the source so that is a
uh you will consider this side of
the
amplifier and GT it is the ultimate
figure of Merit or it is a power
delivered to the load you are
considering the entire circuit and power
available from the source so this PL by
PS will give the complete amplifier
where the source is also considered and
the load is also considered now we'll
try to divide try to derive the each
parameters by using the signal flow
graph so before going for the uh find
out the uh each parameters we'll first
find the B2 by A1 so that is a
t21 uh what is a transfer function for
t21 that is a B2 by i1 so how to find
out so in order to get uh t21 which is
equal to B2 by uh A1 so you need to make
the signal flow graph so signal flow
graph once you have made you can find
out how many paths and how many uh Loops
are there
so if if we see the
network if we try to draw that
is this is A1
B1 B2
A2 these are the notes so always signal
will travel from A to
B this is
S21 S12
S11
s
22 and because there is a load you
should consider
the gamma
l so if you if you want to find out the
B2 by A1 how many paths are there so
there is only one path that is a S21 so
there is only One path S21 how many
Loops are there which is not touching to
the path there is no Loop which is not
touching to the path that is why Delta 1
is equal to
1 Delta 1 is equal to 1 here and you
will have a Delta because there is one
Loop s to2 gamma l so 1 - s22 into gamma
l once you put in this equation you will
have B2 by A1 S2 1 by 1 - s22 gamma l so
this is the way you should find out the
uh transfer function by using the
mission semantic rule so now we will use
this B2 by A1 this equation in order to
find the various power gain so the first
is the power gain GP power delivered to
the load divide by power delivered to
the network so power delivered to Lo
load PL is given so this is
a network that we we are going to
consider there is a gamma l so power
delivered to the load is given by this
and
power delivered to the network is given
by P in so you can directly write B2
Square minus A2 Square while discussing
the S parameter we have already
discussed that you can represent the
power in terms of b and a incident and
reflected wave so this is a B2 Square -
A2
squ similarly you can also get the input
power because there is a input that is
A1 Square - B1
squ once you take out B2 and A1 common
there is 1 - A2 by B2 squ 1 - B1 by A1
Square so uh B2 by A1 Square we have
already find out a2x B2 because here
when you consider this comma s
your gamma L will be like this which
will be from B2 to A2 so you can get
your
A2 which is equal to gamma L
B2 that's why B A2 by
B2 is equal to GMA l so that you replace
here and at the input side your input
reflection icient is given by B1 by A1
so that is gamma in square we already
have derived B2 that is S2 1 s by 1 - S2
to gamma l so we just replace
this which is a t21 that we have derived
in
the last slide so this is a
GP power gain equation which is a power
delivered to the load and power
delivered to the network we will see
there are there is a specific
application to consider this power gain
now we will move to the another equation
that we are going to derive that is the
what is the attributes of the power gain
so for M loads you can consider gamma L
which is equal to Z and Gamma in which
is equal to S11 the power gain that is 1
by 1 - G Square S2 1 sare 1 - GMA L
Square here uh s22 into gamma l so if
you make this as a zero this term is a
one uh this is also zero then you will
have S2 1 Square 1 - S11 squ so for
lossless device if you have a GP which
is equal to 1 we have the power
conservation relation S21 squ + S1 1
square which is equal to 1 so this is
also satisfying the power conservation
relation and it can also it is widely
used for power amplifier design where
output impedance is more important so
this GP gain is uh considered when you
have a output impedance matching is more
important and you your gamma L that you
need to find
out then we will go to the transducer
power gain that is a GT which is a
ultimate uh figure of Merit GT that is
power delivered to the load and Power
available from The Source equal to PL by
P
AVS so power delivered to the load RMS
value you can directly write B2 squ - A2
Square you will get B2 sare which is
equal to 1 - comma l squ so how do we
get the power available from The Source
uh same we will try to uh get this is a
JS and then that is conjugate if it is
match then from this there is incident
wave signal A1 and reflected is a B1 you
can directly write A1 and B1 so there is
a gamma s conjugate and Gamma s so BS
that is a source to the signal will
travel to the A1 and then A1 to A1 so
there is only magnitude 1 in order to
find out the BS by A1 uh a A1 in the
relation to the BS so what is the A1
here A1 we can directly write A1 is
equal
to BS by 1
- gamma
s² by using this signal flow graph
because there is only one Loop that is
gamma Square there is uh One path that
is BS so A1 by BS uh that is 1 by 1 -
GMA s
squ so P AVS you can write A1 Square -
B1 Square which is equal to A1 squ 1 -
B1 by A1 Square A1 Square you can
replace with this equation that is 1 -
GMA a squ whole square and there is a 1
- GMA Square B1 by A1 again it is a GMA
a square so this uh will be cancel out
and you will have B squ by 1 - GMA s² so
here GMA s is less than 1 so uh that is
why that is uh we can directly cancel
this so transducer gain then finally
return as a B2 by B is 1 - gamma l
s into 1 - gamma s squ so here gamma L
square
square based on this equation so when
you take B2 square that is 1 - comma l
squ
so this is the GT now we try
to this we will try to find out the B2
by BS so now important term is a B2 by
BS so how to find the B2 by BS we first
draw the signal flow graph so this is a
signal flow graph for the same This is
BS A1 gamma s S11 s22 2 GMA L S12 we we
have made the signal flow graph now how
many paths are there from B2 to BS there
is a only single path if you want to
reach to B2 that is only single path
S21 and how many Loops are there there
are three loops and two are non touching
Loop so one is gamma s gamma L
combination another is gamma l s s22
combination and then third is a gamma s
S21 GMA L S12 so we we tried to consider
all these three Loop in the transfer
function so that is a P1 Delta 1 by
Delta Delta 1 is equal to 1 because
there are no nontouching Loop all the
loops are touching to the path so that
is S2 1 divide by 1 minus uh the first
Loop that is gamma s into S11
then second Loop s22 into gamma L and
third Loop that is S1 to gamma s S21 GMA
L S12 so this is a third Loop so 1 minus
gamma s S11 + s22 gamma L plus gamma s
S21 S12 gamma
L plus because these two are
non-touching Loops so you can multiply
that is gamma s S11 and Gamma s s22 so
here straightforward we have tried to
use the mission semantic Rule and if you
try if you work out this equation B2 by
b² 1 - comma l s into 1 - comma
s² equal to 1 - gamma s whole Square so
only the thing is you need to replace
this if you uh rewrite this and uh it is
a straightforward you can uh and if you
consider gamma in or gamma out equation
you will get 1 - gamma s² by 1 - S11 GMA
is S21 squ 1 - GMA l s by 1 - GMA out
gamma L whole s so this
is this is the way you can find out the
transducer gain that is GT it is a
ultimate figure of Merit for the power
gain so what is the attributes of uh
Power gain here you are if you are
considering the gamma L which is gamma
out conjugate then it is a power
available from Network so this is a GT
PL by P AVS uh B2 by b s² 1 - GMA L
Square 1 - GMA square if you consider
that your
uh transistor is a unilateral where it
is S12 is equal to 0 in that case you
can consider gamma out
which is equal to
s22 so in unilateral that means you do
not have any power
flow so here this uh from A2 to B1 there
is no power flow means you have a very
good isolation between input and output
and S12 is equal to 0 so if you consider
S12 is equal to 0 in this equation you
will have 1
minus gamma s S11
minus
s22 GMA L plus gamma s
S11 gamma l
s22 so you can easily write this as a 1
minus gamma s
S11
1us gamma l s 22 so this gamma out will
be replaced with the s22 if you consider
the S12 is equal to 0 that is a that is
called a unilateral
case so or if you can also consider the
gamma out which is the gamma is equal to
whole Square so what is a property your
GT is a function of as parameter of your
network transistor Network it is a
function of gamma s what is a uh Source
reflection coefficient gamma L what is
the load reflection coefficient and both
source and load imp Ides so it is a
function
of all the parameters available in the
transistor Network so if you have a
gamma s GMA L is equal to Zer that means
if it is a match to the S11 and s22 you
will have a GT which is equal to S21 squ
for gamma s and gamma L is equal to Zer
because if you put gamma s and gamma L
is equal to Z you will have a GT S21
Square Now power available power gain
you can also get GA where power
available from the network P AVN and
power available from The Source P AVS so
only the difference in power power
available from the network if you
consider gamma L is equal to gamma out
conjugate so you can directly replace in
the
GT uh this GMA L uh is GMA out conjugate
so gamma out Square you can directly
write this a available power gain
equation so here GA is s function of
gamma and because there is a source is
more important and we already know that
for low noise amplifier you are going to
connect antenna when we have discussed
the different transceiver architecture
the first component is a Lono amplifier
and its input impedance should be match
with the antenna so input side is is
more important and in that case uh this
GA is a function of gamma and as your
output is already match so it it has a
uh good application this G has
application in Ln design and GP as I
told it is application in the power
amplifier design where the load output
matching network is important because we
are going to connect the power amplifier
with the antenna output of the power
amplifier with the
antenna so uh this is all the gain
formula that we have
considered so if we summarize this is a
transistor Network so there is a two
Port transistor Network and
here uh there is a input reflection
coefficient output reflection
coefficient that is important
GT this is a transducer gain uh GA that
is uh available uh Power gain means uh
and GP that is a power gain so these all
are the uh figure of Merit and uh we can
try to use it uh for Designing the
amplifier now how to select uh gamma s
and gamma L at given frequency so if you
want to uh design an
amplifier uh how you can design an
amplifier so when you want to design the
amplifier our focus should be uh in this
section will be on achieving a specific
gain Target what do you want you want a
maximum gain GT or specific gain GT
Which is less than GT Max or other
possible targets to discuss are the
minimize the noise because that is
related to the gamma when we see the low
noise amplifier we'll discuss more in
detail the gamma s or how you can find
the optimum value of gamma s then
maximum output power P1 DB we need to
have a uh design specific gamma so that
you will get the maximum output power
p1db and it is application in the power
amplifier and we need to check that the
device is a stable for this gamma s or
gamma L selected how do we obtain the
maximum gain so our Target is to find
the maximum gain of the amplifier and uh
you can use all this uh gain formula and
stability criteria in order to find out
the value of GMA s and gamma L required
to achieve the necessary gain or a
maximum gain so in this uh lecture we
have discussed all the gain formula in
the next we will try to uh design the uh
amplifier by using this formula so thank
you we'll discuss about the amp fire
design in the next
lecture
