Filter Design1

Whites, EE 481/581
Lecture 29
Page 1 of 8
Lecture 29: Microwave Filter Design

by the Insertion Loss Method.
The next major topic were going to cover in this course is
microwave filter design. Its theoretical basis is exactly the same
as low frequency analog filters, as you saw in your electronics
courses.
For example, youll see Butterworth and Chebyshev filters, but
designed to operate at microwave frequencies. Implementation
of these filters is different, however. For example, we wont use
discrete inductors and capacitors.
Insertion Loss Method

We will begin this process with the design of analog filters, but
perhaps with more detail than youve seen before in other
courses.
There are different methods for systematically designing filters,
but the insertion loss method is probably the most prominent. In
this technique, the relative power loss due to a lossless filter
with reflection coefficient :
2015 Keith W. Whites
Whites, EE 481/581
Lecture 29
Z0
Page 2 of 8
Z0
Z in
is specified in the power loss ratio PLR defined as:

P
Po
PLR inc
Pload P 1 2
o
2
PLR 1
or
(8.49),(1)
If both the load and source ports are matched for this network,
2
then PLR 1 S 21 .
In Section 4.1, the text shows that is an even function of

2
. This implies that can only be expanded in a
polynomial series in 2 .
2
In particular, for a linear and time invariant system, is a

rational function
2
Z Z 0
in
Z in Z 0
meaning it can be expressed as a quotient of (real and

nonnegative) polynomials M 2 and N 2 . For physically
realizable filters, we choose:
M 2
2

(8.51),(2)
2
2
M N
Whites, EE 481/581
Lecture 29
PLR 1
Using (2) in (1)
Page 3 of 8
M 2
N
(8.52),(3)
This is valid for any linear, time invariant system that is an even
function of .
Types of Low Pass Filters

There are four types of low pass filters discussed in the text that
are all based on (3):
1. Maximally Flat, Butterworth, Binomial Filter. For this type
of low pass filter:
2N

PLR 1 k 2
c
where N filter order and c cutoff frequency.
(8.53),(4)
If k 1, then PLR 2 at c , which is the 3-dB frequency:

1
PLR
Flat
1
12
Pass band
Stop band
For large and with k 1, then
Whites, EE 481/581
Lecture 29
2N
Page 4 of 8
2N

PLR k 2 =12
(5)
c
c
From this result we learn that the insertion loss IL, defined as
IL 10log PLR ,
(8.50)
increases by 20N dB/decade in the stop band for the

maximally flat low pass filter.
2. Equal Ripple or Chebyshev Filter. For this type of low pass
filter:

PLR 1 k 2TN2
(8.54),(6)
c
where TN x is the Chebyshev polynomial. A typical plot of
1
PLR
in (6) is
1
PLR
Equal ripple
1
1 k 2
Pass band
Stop band
Generally, N is chosen to be an odd integer when the source

and load impedances are equal (two-sided filters).
For large c and using the large argument form of TN , (6)
becomes
Whites, EE 481/581
Lecture 29
Page 5 of 8
2N
k 2
PLR
4 c
As with the Butterworth filter, (7) also increases at
dB/decade, but with the extra factor
22 N
4
compared to (5). Consequently, there is more so-called
off with the Chebyshev low pass filter. For example,
223
N 3 log10
12.0 dB
4
225
N 5 log10
24.1 dB
4
2
(7)
20N
(8)
roll
This is a very sizeable increase in the stop band attenuation

of the Chebyshev filter over the Butterworth.
3. Elliptic Filter. This type of low pass filter has an equi-ripple
response in both the pass band and the stop band. It has a
faster roll off than the previous two filters.
4. Linear Phase Filter. If its important that there be no signal
distortion, then the phase of the filter must be linear in the
passband.
Why? Its the group delay that contributes to signal distortion
of a transient waveform.
Whites, EE 481/581
Lecture 29
Page 6 of 8
General Procedure for Filter Design

The general procedure for designing a filter using the insertion
loss method can be summarized in three steps (Fig 8.23):
1.
2.
3.
Filter specifications. These include the cutoff frequency, the

stop-band attenuation, the pass-band insertion loss, the
pass-band behavior, etc.
Design a low pass prototype circuit. In such a prototype,
R 1 and c 1 rad/s. Filter tables are used for this step,
or perhaps a computer package.
Scale and conversion. Finally, the filter is scaled to the
proper impedance level and, if desired, to a high pass, band
pass, or band stop topology.
Prototype Circuit for the Low Pass Filter

Well consider this whole process in an example in the next
lecture. Before that, well derive the L and C values for a second
order, low pass prototype filter, as shown in Fig 8.24:
Z in
Whites, EE 481/581
Lecture 29
Page 7 of 8
This will help demystify the origin of filter table coefficients

that well see in the next lecture.
From this above figure
1
R
(9)
|| R j L
1 j RC
jC
The reflection coefficient at the input port is
Z 1
in
(10)
Z in 1
From the definition of PLR in (1) and using (10)
1
2
Z in 1
Z in 1 Z in* 1
*
PLR 1
(8.59),(11)

*
2 Z in Z in
Z in 1 Z in 1
Z in j L
Your text in equation (8.60) gives expressions for the numerator

and denominator of (11) in terms of R, L, and C.
For a maximally flat low pass filter, then from (4) and with k 1
2N

4
PLR 1
(12)
1
c N 21
c
Equating (11) and (12) and after some algebra, the text shows
L C 2 and R 1
(13)
This is second row of Table 8.3.
This circuit is a prototype in that
1. The source and load resistances = 1 , and
2. c 1 rad/s
Whites, EE 481/581
Lecture 29
Page 8 of 8
The two topologies for low pass prototype circuits are shown in
Fig 8.25:
L2=g2
R0=g0=1
C1=g1
C3=g3
RN+1=gN+1
L1=g1
G0=g0=1
L3=g3
C2=g2
GN+1=gN+1
Both give identical responses. The immitance values, g, for a

low pass filter are defined as
inductance for series elements
gk
k 1,..., N
capacitance
for
shunt
elements
Filter tables can be used to determine these parameters in

prototype circuits for maximally flat, equi-ripple, and other
types of filters. A great reference for this and related filter topics
is:
G. L. Matthaei, L. Young and E. M. T. Jones, Microwave
Filters, Impedance-Matching Networks, and Coupling
Structures. Norwood, MA: Artech House, 1980.
The following filter tables are listed in the text:
Table 8.3 Maximally flat low-pass,
Table 8.4 Equi-ripple w/ 0.5 dB and 3 dB ripple,
Table 8.5 Linear phase (maximally flat time delay).

Filter Design1

Uploaded by

Copyright:

Available Formats

Filter Design1

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Filter Design1

Uploaded by

Copyright:

Available Formats

Whites, EE 481/581

Lecture 29: Microwave Filter Design

Insertion Loss Method

2015 Keith W. Whites

is specified in the power loss ratio PLR defined as:

In Section 4.1, the text shows that is an even function of

In particular, for a linear and time invariant system, is a

meaning it can be expressed as a quotient of (real and

Using (2) in (1)

Types of Low Pass Filters

If k 1, then PLR 2 at c , which is the 3-dB frequency:

For large and with k 1, then

increases by 20N dB/decade in the stop band for the

Generally, N is chosen to be an odd integer when the source

This is a very sizeable increase in the stop band attenuation

General Procedure for Filter Design

Filter specifications. These include the cutoff frequency, the

Prototype Circuit for the Low Pass Filter

This will help demystify the origin of filter table coefficients

Your text in equation (8.60) gives expressions for the numerator

Both give identical responses. The immitance values, g, for a

Filter tables can be used to determine these parameters in

You might also like