Result

1. Design a minimum order, stable1, low-pass Butterworth filter with a passband

frequency of 1 kHz and a stopband frequency of 1.4 kHz. Make the attenuation 1 dB
at the passband frequency and 80 dB at the stopband frequency.
2. Design a minimum order, stable, lowpass Chebyshev Type I filter with the same
specifications as the Butterworth filter.
3. Design a lowpass FIR filter using the Blackman Window with a cut-off frequency of
1 kHz. Specify the order of the filter such that the first minimum in the stopband
(preceding the first lobe) is as close to 1.4 kHz as possible without exceeding it.
Discussion

A digital filter includes an analogue-to-digital converter (ADC) that samples the input signal,
a microprocessor, and additional components for storing filter coefficients and data. A
digital-to-analogue converter is also provided shortly before the output step. The
microprocessor program builds a digital filter by operating on an ADC number and executing
mathematical calculations. It may apply various effects to the sample signal, such as
amplification and delay.

A digital filter's behaviour is also significant. To comprehend the responses, many

mathematical methodologies are employed. The easiest method is to examine the response
when a basic input, such as an impulse, is fed into the filter. Then, based on the outcome,
complicated inputs may be examined.

The Butterworth Filter approximation function's frequency response is also known as

"maximally flat" (no ripples) response because the pass band is meant to have a frequency
response that is as flat as theoretically feasible from 0Hz (DC) to the cut-off frequency at -
3dB with no ripples. Higher frequencies beyond the cut-off point roll-off to zero in the stop
band at a rate of 20dB per decade or 6dB per octave. This is due to its low-quality factor, Q
of 0.707.

The Butterworth filter's fundamental shortcoming is that it achieves pass band flatness at the
price of a large transition band as the filter transitions from the pass band to the stop band. It
also has bad phase characteristics. For various filter orders, the ideal frequency response,
referred to as a "brick wall" filter. It's worth noting that the higher the Butterworth filter
order, the more cascaded stages there are in the filter design, and the closer the filter gets to
the ideal "brick wall" response. However, in practise, Butterworth's ideal frequency response
is impossible due to significant passband ripple. The frequency response of a "nth" Order
Butterworth filter is described by the generalised equation:

1
H ( jω)=

√ ( )
2n
2 ω
1+ ε
ωp

Where: n represents the filter order, Omega ω is equal to 2πƒ and Epsilon ε is the maximum
pass band gain, (Amax). If Amax is defined at a frequency equal to the cut-off -3dB corner point
(ƒc), ε will then be equal to one and therefore ε2 will also be one. However, if you now wish
to define Amax at a different voltage gain value, for example 1dB, or 1.1220 (1dB =
20*logAmax) then the new value of epsilon, ε is found by:

H0
H 1=
√ 1+ ε2
Where:

H0 = the Maximum Pass band Gain, Amax.

H1 = the Minimum Pass band Gain.

Transpose the equation to give:

H0
=1.1220=√ 1+ε 2  gives ε =0.5088
H1

A filter's Frequency Response may be theoretically characterised by its Transfer Function

with the standard Voltage Transfer Function H ( jω) represented as:

H ( jω)= [ Vout ⁡( jω)

Vin ⁡( jω) ]
Where:

Vout = the output signal voltage.

Vin = the input signal voltage.

j = to the square root of -1 (√-1)

ω = the radian frequency (2πƒ)

To signify the S-domain, (j) might likewise be written as (s). Thus, the transfer function for a
second-order low pass filter is provided as follows:

 Vout  1
H ( s)= = 2
 Vin  S + S+1

Chebyshev filters are used to separate frequencies in one band from those in another. They
cannot compete with the performance of the windows-sink filter, but they are suited for a
wide range of applications. The major advantage of Chebyshev filters is their speed, which is
usually quicker than that of windowed-sinc. Because these filters are implemented by
recursion rather than convolution. The Chebyshev and Windowed-Sinc filters are designed
using a mathematical approach known as the Z-transform.

Some of the key features of the Chebyshev RF filter can be summarised as below:

The Chebyshev filter has a steep roll-off, which is one of its key characteristics. It achieves
its final roll-off rate faster than other types of filters. As a result, it is commonly utilised in
RF applications where a sharp transition between passband and stop-band is necessary to
eliminate undesirable products such as harmonic intermodulation.

Although the Chebyshev filter has a steep roll-off, it does so at the expense of ripple.
Although there are several varieties of Chebyshev filters, this characteristic of their
performance may exclude their usage.

Because of the in-band ripple, the standard definition of the cut-off frequency as the point at
which the response decreases to -3 dB does not apply to Chebyshev filters. Instead, the cut-
off is defined as the moment at which the gain finally falls to the value of the ripple. This
may be seen in the diagram of a typical Chebyshev filter response.

The term Chebyshev filter derives from the fact that the format and calculations for the filter
are based on Chebyshev polynomials.

The gain (or amplitude) response, Gn as a function of the angular frequency, ω for an n-th
order Chebyshev filter can be expressed in the form of the function below:

1
Gn ( ω ) =|H n ( ω )|=

√ 1+ε 2 T 2n
( )
ω
ωc

Where:

ε = ripple factor

ωc is the cut-off frequency

Tn is the Chebyshev polynomial of the nth order

The Chebyshev filter's pass-band exhibits equi-ripple behaviour. The ripple factor, ε
determines the in-band ripple. The Chebyshev polynomial alternates between -1 and 1 in the
passband. This means that the actual reaction / gain alternates between a maximum of unity
and a minimum level set by the following formula:
 1
G=
√ 1+ε2
IIR filters are difficult to manage and have no specific phase, whereas FIR filters always
allow for a linear phase. IIR has the potential to be unstable, whereas FIR is always stable.
When compared to FIR, IIR might have restricted cycles, but FIR does not. IIR is analog-
derived, whereas FIR has no analogue background. Polyphase implementation is conceivable
with IIR filters, but FIR can always be made casual.

FIR filters rely on linear-phase features, whereas IIR filters are utilised for nonlinear
applications. The latency characteristics of FIR are far superior, but they demand more
memory. IIR filters, on the other hand, are reliant on both i/p and o/p, whereas FIR is just
based on i/p. IIR filters have zeros and poles and need less memory than FIR filters, which
only have zeros. IIR filters may be difficult to build, and tweaks to delay and distort can
change the poles and zeroes, making the filters unstable, whereas FIR filters stay stable. FIR
filters are utilised for higher-order tapping, whereas IIR filters are suitable for lower-order
tapping since IIR filters might become unstable while tapping higher-orders.
Conclusion

Finally, IIR is infinite and is employed in cases where linear properties are not important.
Then there are FIR filters, which are Finite IR filters that are necessary for linear-phase
properties. Aside from that, IIR is better for lower-order tapping while FIR is better for
higher-order tapping. Furthermore, FIR filters are preferable over IIR filters because they are
more stable and do not require feedback. Finally, IIR filters are recursive and are employed
as a backup, but FIR filters have grown too lengthy and pose issues in a variety of
applications.