Comparison of Butterworth

and Chebyshev Prototype of Bandpass

Filter for MRI Receiver Front End

Shraddha Ajay Joshi, Thyagarajan Jayavignesh and Rajesh Harsh

Abstract Background/Objectives: Filter in MRI front-end receiver chain plays an

important role in rejecting the undesired image frequencies. These filters are
designed for RF and IF stages in the front-end chain. Methods/Statistical analysis:
Bandpass filters are designed for obtaining particular frequency band. The MRI
frequency lies in this frequency band selected by the bandpass filter. Different types
of prototypes for design of bandpass filter are discussed here. The major two
prototypes of filter are designed and simulated. Findings: There are various prac-
tical filter prototypes in which the major two filter prototypes: Butterworth and
Chebyshev are analyzed. These filters give better response and results for the MRI
frequency. They are designed at novel frequency of 63.87 MHz which is the fre-
quency or the MRI receiver chain. The filter is designed and simulated at different
orders for obtaining better performance with respect to the insertion loss and return
loss. The novelty of the results obtained is that the insertion loss is minimum at the
frequency of the MRI receiver front-end chain. This will reduce the loss in the MRI
system, increases gain of system, and thus image obtained will be more clear and
accurate. Application/Improvements: Better filter can be designed by reducing the
passband bandwidth which will provide more accuracy to the system.

Keywords Butterworth prototype
Chebyshev prototype
Insertion loss

Return loss MRI receiver chain

S. A. Joshi
T. Jayavignesh (&)

School of Electronics Engineering, VIT University, Chennai, India
e-mail: [email protected]
S. A. Joshi
e-mail: [email protected]
R. Harsh
Scientist at SAMEER, Mumbai, India
e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2018 395

A. K. Nandi et al. (eds.), Computational Signal Processing
and Analysis, Lecture Notes in Electrical Engineering 490,
https://doi.org/10.1007/978-981-10-8354-9_36
396 S. A. Joshi et al.

1 Introduction

The MRI is used for the noninvasive diagnosis of a body. The microwave fre-
quencies are used in MRI system for detection uneven tissues in our body. MRI
technique is specially used for medical diagnosis where the body is not directly
exposed to any harmful radiation. As magnetic resonance imaging (MRI) is a
technique used to produce the images of the soft tissues present in our body, it is
essential to design a receiver which can yield clear visible images where the defects
hidden in our body can be detected. Therefore, the bandpass filter plays a key role
in filtering the frequency of required signals which will in turn help in discarding
the image frequencies.
The following block diagram (Fig. 1) illustrates the MRI receiver front-end
chain. The bandpass filter stage next to the LNA is used to let on the required band
of frequencies to the succeeding stage.
For various frequencies, the microwave interdigital bandpass filter is designed. It
is designed to allow band of required frequencies by which the frequency response
of the filter gets monitored. Using ADS software, the design and simulation of
equi-ripple, i.e., Chebyshev filter and interdigital bandpass filter, is described in
paper [1].
To obtain the desired frequencies, the parameters such as length, spacing as well
as width and L, C values are varied for interdigital and lumped bandpass filter,
respectively.
A microstrip filter with Chebyshev response having 0.5 dB ripple is designed for
the frequency of 2.4 GHz. First, the low-pass filter prototype is designed and
transformed into bandpass filter by applying scaling transformations.
Further MCLIN and MCFIL microstrip lines are used to design the respective
filter and theoretical and simulated results are compared using ADS software. The
observations for this paper are made in favor of microstrip bandpass filter as
compared to that of the lumped component filter design [2].
A parallel-coupled microstrip BPF is designed at frequency of 2.44 GHz having
fractional bandwidth of 3.42%. Filter parameter, insertion loss was considered to be
a major factor for optimizing results [3].

Fig. 1 Block diagram of

MRI receiver chain
Comparison of Butterworth and Chebyshev Prototype of Bandpass … 397

The comprehensive treatment of RF and microwave circuit elements used for the
designing the circuits using the lumped elements is described. The topics discussed
in the book are related to the materials, fabrication, analyses, design, modeling, and
physical, electrical, and thermal practical considerations of the design. The entire
design of the bandpass filter can be calculated theoretically using the formulas [4].
The passband of ideal filter has zero insertion loss and linear phase response in
the passband and infinite attenuation in stopband. There is no any clear way to
improve the design in image parameter method as it yields unstable filter response.
Therefore, insertion loss method is used and it allows the high degree of control
over the passband and stopband amplitude and phase characteristics [5–8].

2 Filter Design

The filter schemed in this paper is designed for the frequency of 63.87 MHz which
is resonant frequency in the MRI signaling. The filter is designed using insertion
loss method with the help of lumped components such as L and C. The low-pass
filter design is executed first and then converted to bandpass filter using scaling
transformation technique. To design a filter, its bandwidth and design frequency
should be known.
Usually, Butterworth and Chebyshev prototypes are availed for microwave
frequencies. The aspects to be taken into consideration while designing the filter is
insertion loss, return loss, passband and stopband bandwidth, quality factor, and
attenuation. The desired specifications for conception of bandpass filter are tabu-
lated in Table 1.

2.1 Low-Pass Filter Design

The low-pass filter is designed for the aforementioned specifications. The input and
output impedance of the filter is matched to standard value of 50 X. The component
values of the circuit given in Fig. 2 are designed using the formulae given below:

R0 Lk Ck
L0k ¼ ; Ck0 ¼ : ð1Þ
xc R0 xc

Table 1 Specifications of Center frequency (fc) 63.87 MHz

bandpass filter
3-dB bandwidth (BW) 2 MHz
Fractional bandwidth (∆) 0.03
398 S. A. Joshi et al.

Fig. 2 Schematic of fifth-order low-pass filter

Table 2 Element values for Order 3 5 7 9

Butterworth low-pass filter
prototype g1 1.0000 0.6180 0.4450 0.3473
g2 2.0000 1.6180 1.2470 1.0000
g3 1.0000 2.0000 1.8019 1.5321
g4 1.0000 1.6180 2.0000 1.8794
g5 0.6180 1.8019 2.0000
g6 1.0000 1.2470 1.8794
g7 0.4450 1.5321
g8 1.0000 1.0000
g9 0.3473
g10 1.0000

The values given in Tables 2 and 3 are element values for Butterworth and
Chebyshev prototype, respectively, substituted in place of Lk and Ck for afore-
mentioned formulae.
The schematic of the low-pass filter is shown in Fig. 2, and the results are shown
in Fig. 3.

2.2 Frequency Scaling of the Filter

The filter is designed for the required resonance frequency and with the standard
impedance for matching. To design a bandpass filter, frequency scaling is to be
done for the low-pass filter circuit. The low-pass filter is transformed to the
bandpass filter with the scaling technique. In the transformation process:
The series inductor of the low-pass filter is converted to the parallel circuit of
L and C with component values as given below:
Comparison of Butterworth and Chebyshev Prototype of Bandpass … 399

Table 3 Element values for 0.5 dB equi-ripple low-pass filter prototype

Order 3 5 7 9
g1 1.5963 1.7058 1.7372 1.7504
g2 1.0967 1.2296 1.2583 1.2690
g3 1.5963 2.5408 2.6381 2.6678
g4 1.0000 1.2296 1.3444 1.3673
g5 1.7058 2.6381 2.7239
g6 1.0000 1.2583 1.3673
g7 1.7372 2.6678
g8 1.0000 1.2690
g9 1.7504
g10 1.0000

Fig. 3 Return loss and insertion loss results for the fifth-order low-pass filter

R0 L BW
L0 ¼ C0 ¼ ð2Þ
xc BW R0 Lxc

The shunt capacitor of low-pass circuit is converted to the series circuit of L and
C with the component values given below.

R0 BW C
L0 ¼ ; C0 ¼ ð3Þ
xc C R0 BWxc

The fraction bandwidth and center frequency of the BPF can be calculated as:
400 S. A. Joshi et al.

Fig. 4 Schematic of fifth-order Butterworth bandpass filter

Table 4 Component values Components Value Components Value

for the Butterworth prototype
L1 2.7 lH L4 2.4 nH
C1 2.4 pF C4 2.7 nF
L2 2.4 nH L5 2.7 lH
C2 2.7 nF C5 2.4 pF
L3 8.4 lH Zin 50 X
C3 0.75 pF Zout 50 X

x2 x1 pffiffiffiffiffiffiffiffiffiffiffi
D¼ and x0 ¼ x1 x2 : ð4Þ
x0

where x2 and x1 are the two cutoff frequencies of bandpass filter. The values of the
components in bandpass filter for both the prototypes are calculated using the
formulae mentioned above. According to the values, the schematic is designed for
the two types of filter prototypes.
1. Design of Butterworth Bandpass filter
The schematic of the Butterworth bandpass filter is shown in Fig. 4. The values of
the components are tabulated in Table 4. The results of return loss and insertion
loss are shown in Figs. 5 and 6, respectively.
2. Design of Chebyshev bandpass filter
The schematic of the Chebyshev bandpass filter is shown in Fig. 7. The values of
the components are calculated and tabulated in Table 5. The results of return loss
and insertion loss of the filter design are shown in Figs. 8 and 9, respectively.
Comparison of Butterworth and Chebyshev Prototype of Bandpass … 401

Fig. 5 Return loss results for the fifth-order Butterworth bandpass filter

Fig. 6 Insertion loss results for fifth-order Butterworth bandpass filter

402 S. A. Joshi et al.

Fig. 7 Schematic of fifth-order Chebyshev bandpass filter

Table 5 Component values Components Values Components Values

for the Chebyshev prototype
L1 7.5 lH L4 3 nH
C1 0.8 pF C4 2.04 pF
L2 3 nH L5 7.5 lH
C2 2.04 pF C5 0.8 pF
L3 10 lH Zin 50 X
C3 1 pF Zout 50 X

3 Results and Analysis

The results of two prototypes are tabulated below which shows that the Butterworth
filter gives superior output in terms of return loss and insertion loss. Also, the order
of the filter has an effective impact on these two measuring parameters. So, it is
important to be selective in terms of the order and the prototype for the filter.
The Chebyshev filter produces ripples in the passband as compared to
Butterworth filter, thus affecting the insertion loss parameter resulting to a higher
value in turn declining the return loss. Therefore, the fifth-order Butterworth filter is
selected for designing the bandpass filter in the receiver front-end chain. The values
are shown in Table 6.
Comparison of Butterworth and Chebyshev Prototype of Bandpass … 403

Fig. 8 Return loss results for the fifth-order Chebyshev bandpass filter

Fig. 9 Insertion loss results for fifth-order Chebyshev bandpass filter

404 S. A. Joshi et al.

Table 6 Simulation results (1) Butterworth filter

of BPF for 1 Butterworth filter
Order Maximally flat prototype
prototype and 2 Chebyshev
filter prototype Return loss (dB) Insertion loss (dB)
3 0.1 14.7
5 23.64 0.01
7 1.6e–7 74.2
9 3.85e–15 89.2
(2) Chebyshev filter
Order Chebyshev prototype
Return loss (dB) Insertion loss (dB)
3 0.15 14.75
5 0.122 15.56
7 2.04 4.25
9 0.77 7.82

4 Conclusion

The simulation result divulges that Butterworth filter gives better results as com-
pared to the Chebyshev filter. The novelty of the paper is the filter is designed for
63.87 MHz which is the resonance frequency of the MRI receivers RF stage in the
front-end chain. The return loss of the Butterworth filter is nearly 23.64 dB and
insertion loss of 0.01 dB which yields quite better performance than the Chebyshev
filter, thus making us to select the Butterworth filter for receiver chain. The values
of the circuit component were calculated and the filter was simulated in ADS
software. Further, the circuit can be fabricated on PCB using same component
values. There might be slight variations in the simulated values and practically
measured results.

Acknowledgements I would like to thank Mr. Rajesh Harsh, Head of Technological Innovation
Division (TID) and Mrs. Twisha Pandey, Senior Research Scientist (TID) for their continuous help
and encouragement to work on this ongoing research topic.

References

1. Indira ND, Nalini K, Khan H (2013) Design of interdigital bandpass filter. Int J Eng Adv
Technol (IJEAT) 2(4). ISSN: 2249–8958
2. Khandelwal KS, Kureshi AK (2014) Realization of microstrip band-pass filter design. Int J
Adv Res Comput Eng Technol (IJARCET) 3(12):4242–4247
3. Srivastava S, Manjunath RK, Shanthi P (2014) Design, simulation and fabrication of a
microstrip bandpass filter. Int J Sci Eng Appl (IJSEA) 3(5):154–158
4. Bahl Inder (2003) Lumped elements for RF and microwave circuits. Artech House Inc.,
Bosten, London
Comparison of Butterworth and Chebyshev Prototype of Bandpass … 405

5. Rohde UL, Newkirk DP (2000) RF/microwave circuit design for wireless applications. Wiley,
New York
6. Leenaerts D, Tang J, Vaucher CS (2003) Circuit design for RF transceivers. Wiley, New York
7. Laskar J, Matinpour B, Chakraborty S (2004) Modern receiver front-ends system, circuit and
integration. Wiley, New York
8. Pozar DM (2012) Microwave engineering, 3rd edn. Wiley, New York, pp 396, 401–404