4/24/2005 Automatic Gain Control notes 1/1

5. Receiver Gain and AGC

We find that a detector/demodulator likewise has a dynamic

range, a value that has important ramifications in receiver
design.

HO: Instantaneous Dynamic Range

Q: We have calculated the overall gain of the receiver, but

what should this gain be?

A: HO: Receiver Gain

Q: How can we build a receiver with variable gain? What

microwave components do we need?

A: HO: Automatic Gain Control (AGC)

HO: AGC Dynamic Range

Q: How do we implement our AGC design?

A: HO: AGC Implementation

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Instantaneous Dynamic Range 1/3

Instantaneous
Dynamic Range
Q: So, let’s make sure I have the right—any input signal with
power exceeding the receiver sensitivity but below the
saturation point will be adequately demodulated by the
detector, right?

A: Not necessarily! The opposite is true, any signal with

power outside the receiver dynamic range cannot be properly
demodulated. However, signals entering the receiver within
the proper dynamic range will be properly demodulated only if
it exits the receiver with the proper power.

The reason for this is that demodulators, in addition to

requiring a minimum SNR (i.e., SNRmin), likewise require a
certain amount of power.

If the signals enters the receiver with power greater that

the MDS, then the signal will exit the receiver with sufficient
SNR. However, the signal power can be too large or too
small, depending on the overall receiver gain G.

Q: How can the exiting signal power be too large or too small?
What would determine these limits?

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Instantaneous Dynamic Range 2/3

A: Recall that the signal exiting the receiver is the signal

entering the detector/demodulator. This demodulator will
have a dynamic range as well!

Ps in Ps out = PDin

Rx
SNR in
Rx
SNRRxout
= SNRDin

Say the signal power entering the demodulator (i.e., exiting

the receiver) is denoted PDin . The maximum power that a
demodulator can “handle” is thus denoted PDmax , while the
minimum amount of power required for proper demodulation is
denoted as PDmin . I.E.,:

PDmin ≤ PDin ≤ PDmax

Thus, every demodulator has its own dynamic range, which we

call the Instantaneous Dynamic Range (IDR):

PDmax
IDR = min or IDR (dB ) = PDmax (dBm ) − PDmin (dBm )
PD

Typical IDRs range from 30 dB to 60 dB.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Instantaneous Dynamic Range 3/3

To differentiate the Instantaneous Dynamic Range from the

receiver dynamic range, we refer to the receiver dynamic
range as the Total Dynamic Range (TDR):

Pinsat
TDR = or TDR (dB ) = Pinsat (dBm ) − MDS (dBm )
MDS

Q: How do we insure that a signal will exit the receiver within

the dynamic range of the demodulator (i.e., within the IDR)?

A: The relationship between the signal power when entering

the receiver and its power when exiting the receiver is simply
determined by the receiver gain GRx :

PDin = GRx Ps in

We simply need to design the receiver gain such that PD lies

within the IDR for all signals Ps in that lie within the TDR.

Big Problem Æ We find that typically TDR >> IDR. This can
make setting the receiver gain GRx very complicated!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 1/6

Receiver Gain
Let’s consider each element of a basic super-het receiver:

3
Pin Pout = PD
1 2 4 5

1. LNA - Required to make the receiver noise figure F as

small as possible, thus making the receiver very sensitive.

2. Preselector – Required to reject all spurious-signal

creating frequencies, while simultaneously letting the desired
RF bandwidth pass to the mixer.

3. Mixer - Required for down-conversion; often sets the

receiver compression point.

4. IF Filter - Required to suppress all mixer IF output

signals, with the exception of the one desired signal that we
wish to demodulate. Also determines the noise bandwidth B
of the receiver.

5. IF Amp - Q: Why is this device required? What

receiver parameter does it determine?

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 2/6

A: It is true that the IF amplifier does not generally affect

receiver bandwidth, or sensitivity, or saturation point, or
image rejection.

Æ However, the IF amp is the component(s) that we use to

properly set the overall receiver gain.

Say that we have designed a receiver with some specific TDR

(i.e., MDS and Pinsat ). This receiver will be connected to a
demodulator with a specific IDR (i.e., PDmin and PDmax ). All we
have left to do is determine the proper gain of the IF
amplifier to give us the required gain of the overall receiver.

This gain must satisfy two requirements:

Requirement 1 -We know that the overall receiver gain GRx

must be sufficiently large such that the smallest possible
receiver input signal ( Ps in = MDS ) is boosted at least to the
level of the smallest required demodulator signal ( PDmin ). Thus,
the absolute smallest value that the receiver gain should be is
GRxmin :

PDmin
G min
Rx  or Gmin (dB )  PDmin (dBm ) − MDS (dBm )
MDS

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 3/6

Requirement 2 - Likewise, the overall receiver gain GRx must

be sufficiently small to insure that the largest possible
receiver input signal (i.e., Ps in = Pinsat ) arrives at the
demodulator with a power less than to the maximum level PDmin .
Thus, the absolute largest value that the receiver gain should
be is GRxmax :

PDmax
G max
Rx  sat or GRxmax (dB )  PDmax (dBm ) − Pinsat (dBm )
Pin

Q: Seems simple enough! Just select an IF amplifier so that

the overall receiver gain lies between these two limits:

GRxmin < GRx < GRxmax

Right?

A: Not exactly. We are typically faced with a big problem at

this point in our receiver design. To illustrate this problem,
let’s do an example.

Say our receiver has these typical values:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 4/6

Pinsat = −10dBm

PDmax = −20dBm

PDmin = −60dBm

MDS = −90dBm

Note then that TDR = 80dB and IDR = 40dB .

Thus, this receiver must have a gain of at least:

GRxmin (dB ) = PDmin (dBm ) − MDS (dBm )

= −60 − ( −90 )
= 30 dB

But likewise have a gain of no more than:

GRxmax (dB ) = PDmax (dBm ) − Pinsat (dBm )

= −20 − ( −10 )
= −10 dB

So here’s our solution! The receiver gain must be any value

greater than 30 dB, as long as it is simultaneously less than
-10dB:
30dB < GRx (dB ) < −10dB

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 5/6

Hopefully, it is evident that there are no solutions to the

equation above!!

Q: Yikes! Is this receiver impossible to build?

A: Note that the values used in this example is are very

typical, and thus the problem that we have encountered is
likewise very typical.

We almost always find that GRxmin > GRxmax , making the solution
GRx to the equation GRxmin < GRx < GRxmax non-existent!

To see why, consider the ratio GRxmax GRxmin :

GRxmax PDmax Pinsat PDmax PDmin IDR

= = =
GRxmin PDmin MDS Pinsat MDS TDR

In other words, for Gmax to be larger than Gmin (i.e., for

Gmax Gmin > 1 ), then the IDR must be larger than the TDR (i.e.,
IDR TDR > 1 ).

But, we find that almost always the demodulator dynamic

range (IDR) is much less than the receiver dynamic range
(TDR), thus Gmax is almost never larger than Gmin .

Typically, TDR >> IDR

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 Receiver Gain 6/6

Big Solution Æ However, there is one fact that leads to a

solution to this seemingly intractable problem.

The one desired input signal power can be as small

as MDS or as large as Pinsat , but it cannot have both
values at the same time!

Thus, the receiver gain GRx may need to be larger than GRxmin
(i.e., when Ps in = MDS ) or smaller as GRxmax (i.e., when
Ps in = Pinsat ), but it does not need to be to be both at the same
time!

In other words, we can make the gain of a receiver adjustable

(i.e., adaptive), such that:

1. the gain becomes large enough (GRx > GRxmin ) when

Ps in GRx the input signal power Ps in is small, but:

2. the gain becomes small enough (GRx < GRxmax ) when

the input signal power Ps in is large.
Ps in
GRx

Q: Change the gain of the receiver, how can we possibly do

that?

A: We can make the gain of the IF amplifier adjustable, thus

making the overall receiver gain adjustable. This gain is
automatically adjusted in response to the signal power, and we
call this process Automatic Gain Control (AGC).

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 1/7

Automatic Gain Control

To implement Automatic Gain Control (AGC) we need to make
the gain of the IF amplifier adjustable:

Preselector
GIF IF Filter
Filter

Gain
Control

Q: Are there such things as adjustable gain amplifiers?

A: Yes and no.

Typically, voltage controlled amplifiers work poorly, have

limited gain adjustment, or both.

Instead, receiver designers implement an adjustable gain

amplifier using one or more fixed gain amplifiers and one or
more variable attenuators (e.g., digital attenuators).

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 2/7

Adjustable
G1 G2 Attenuator(s)
AL < A < AH

“IF Amplifier” GIF

Gain
Control

Two amplifiers are used in the design above, although one,

two, three, or even four amplifiers are sometimes used.

The adjustable attenuator can likewise be implemented in a

number of ways. Recall the attenuator can be either digital
or voltage controlled. Likewise, the attenuator can be
implemented using either one attenuator, or with multiple
cascaded attenuator components.

However it is implemented, the gain of the overall “IF

amplifier” is simply the product of the fixed amplifier gains,
divided the total attenuation A. Thus, for the example above:

G1G2
GIF = GIF (dB ) = G1 (dB ) + G2 (dB ) − A (dB )
A

Now, the key point here is that this gain is adjustable, since
the attenuation can be varied from:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 3/7

AL < A < AH

Thus, the “IF amplifier” gain can vary from:

GIFL < GIF < GIFH

Where GIFL is the lowest possible “IF amplifier” gain:

G1G2
GIFL = GIF (dB ) = G1 (dB ) + G2 (dB ) − AH (dB )
AH

And GIFH is the highest possible “IF amplifier” gain:

G1G2
GIFH = GIF (dB ) = G1 (dB ) + G2 (dB ) − AL (dB )
AL

Note the gain is the highest when the attenuation is the

lowest, and vice versa (this should make perfect sense to
you!).
However, recall that the value of the lowest
attenuation value is not equal to one (i.e., AL > 1 )!
Instead AL represents the insertion loss of the
attenuators when in their minimum attenuation
state. The highest attenuation value AH must
likewise reflect this insertion loss!

Recall also that the total receiver gain is the product of the
gains of all the components in the receiver chain. For
example:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 4/7

GRx = GLNA Gpreselector Gmixer GIF GIFfilter

Note, however, that the only adjustable gain in this chain is

the “IF amplifier” gain GIF , thus the remainder of the
receiver gain is fixed, and we can thus define this fixed gain
GRx
fixed
as:
GRx
GRxfixed 
GIF

Thus, GRx
fixed
is simply the gain of the entire receiver, with the
exception of the “IF amplifier”.

Since the gain of the “IF amplifier” is adjustable, the gain of

entire receiver is likewise adjustable, varying over:

GRxL < GRx < GRxH

where:
GRxL = GRxfixed GIFL

and:
GRxH = GRxfixed GIFH

Q: So what should the values of GIFL and GIFH be? How will I
know if my design produces a GIFL that is sufficiently low, or a
GIFH that is sufficiently high?

A: Let’s think about the requirements of each of these two

gain values.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 5/7

1: GIFH

Remember, a receiver designer must design their “IF

Amplifier” such that the largest possible receiver gain GRxH
exceeds the minimum gain requirement (i.e., GRxH > GRxmin )—a
requirement that is necessary when the receiver input signal
is at its smallest (i.e., when Ps in = MDS ).

To accomplish this, we find that:

GRxH > GRxmin

GRxfixed GIFH > GRxmin
GRxmin
G H
IF > fixed
GRx

Thus, since GRxmin = PDmin MDS we can conclude that our “IF
amplifier” must be designed such that its highest possible
gain GIFH exceeds:

PDmin
G H
IF > fixed
GRx MDS

or

GIFH (dB ) > PDmin (dBm ) − GRxfixed (dB ) − MDS (dBm )

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 6/7

2: GIFL

Additionally, a receiver designer must design their “IF

Amplifier” such that the smallest possible receiver gain GIFL is
less that the maximum gain requirement (i.e., GL < Gmax )—a
requirement that is applicable when the receiver input signal
is at its largest (i.e., when Pin = Pinsat ).

To accomplish this, we find that:

GRxL < GRxmax

GRxfixed GIFL < GRxmax
GRxmax
G L
IF < fixed
GRx

Thus, since GRxmin = PDmin MDS we can conclude that our “IF
amplifier” must be designed such that its highest possible
gain GIFH exceeds:

Thus, since GRxmax = PDmax Pinsat we can conclude that our “IF
amplifier” must be designed such that its lowest possible
gain GIFL is below:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC 7/7

PDmax
G L
IF < fixed sat
GRx Pin

or

GIFL (dB ) < PDmax (dBm ) − GRxfixed (dB ) − Pinsat (dBm )

Q: I’m still a bit confused. Now what is the difference

between GRxmin ,GRxmax and GRxL ,GRxH ?

A: The values GRxmin and GRxmax are in fact requirements that

are placed on the receiver designer.

* In other words, there must be some IF gain setting that

will result in a receiver gain GRx greater than GRxmin (a
requirement for detecting Ps in = MDS ) , and there must be
some IF gain setting that will result in a receiver gain GRx less
than GRxmax (a requirement for detecting Ps in = Pinsat )

* In contrast, the values GIFL and GIFH are the actual minimum
and maximum values of the receiver gain. They state the
performance of a specific receiver design.

Properly designed, we will find that GRxH > GRxmin , and

GRxL < GRxmax . However, this is true only if we have
properly design our “IF Amplifier”!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC Dynamic Range 1/3

AGC Dynamic Range

Now let’s consider the dynamic range of our AGC, defined as:

GRxH GRxfixed GIFH GIFH

AGC Dynamic Range = L = fixed L = L
GRx GRx GIF GIF
Therefore:

AGC Dynamic Range (dB ) = GRxH (dB ) − GRxL (dB )

= GIFH (dB ) − GIFL (dB )

Q: Just how much dynamic range do we need?

A: Since for a properly designed receiver, GRxH > GRxmin and

GRxL < GRxmax , we can conclude that for a properly designed
receiver:
GRxH GRxmin
AGC Dynamic Range = L > max
GRx GRx

Meaning that, since GRxmin = PDmin MDS and GRxmax = PDmax Pinsat :

PDmin Pinsat
AGC Dynamic Range >
MDS PDmax
PDmin Pinsat
> max
PD MDS
TDR
>
IDR

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC Dynamic Range 2/3

Thus, we conclude that for a properly designed receiver:

TDR
AGC Dynamic Range >
IDR

or

AGC Dynamic Range (dB ) >TDR (dB ) − IDR (dB )

From the standpoint of “IF Amplifier” design, this result has a

specific meaning.

Since the gain of the amplifiers used in the “IF Amplifier”

design is fixed (e.g., G1 G2 ), the ratio of the largest and
smallest IF amplifier gain is simply the ratio of the largest
and smallest attenuator values:

GIFH G1G2 AH A
AGC Dynamic Range = L = = H
GIF AL G1G2 AL
or

AGC Dynamic Range (dB ) = AH (dB ) − AL (dB )

Thus, we can conclude that the variable attenuator(s) in an

“IF amplifier” must be selected such that the range of
attenuation, from AH to AL satisfies:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/3/2006 AGC Dynamic Range 3/3

AH TDR
>
AL IDR

or

AH (dB ) − AL (dB ) >TDR (dB ) − IDR (dB )

Note this is a necessary requirement for proper receiver

operation, but it is not a sufficient one!

In other words, the expression above provides only one “IF

Amplifier” design equation. We must also select the gains of
the amplifiers in the “IF Amplifier” such that:

G1 G2 GRxmin
G H
IF = > fixed
AL GRx

where we recall thatGRx

fixed
represents the gain of all the
receiver components, except those components comprising the
“IF Amplifier”.

Jim Stiles The Univ. of Kansas Dept. of EECS