Design iir filter on fpga technology combined with matlab simulink software tools and electronic design automation software (thiết kế bộ lọc iir trên công nghệ fpga kết hợp cá

International Journal of Research In Vocational Studies (IJRVOCAS) VOL 1, NO 4, April 2022, PP 66~70 Print ISSN 2777 0168| Online ISSN 2777 0141| DOI prefix 10 53893 https //journal gpp or id/index ph[.]

Print ISSN 2777-0168| Online ISSN 2777-0141| DOI prefix: 10.53893

https://journal.gpp.or.id/index.php/ijrvocas/index

Design IIR Filter on FPGA Technology Combined with

Matlab-Simulink Software Tools and Electronic Design

Automation Software

Hong Ngan Vo1, Xuan Toai Nguyen1, Thanh Lanh Le1, Nam Thoi Le1

1Dong Nai Technology University, Viet Nam

Email address:

vohongngan@dntu.edu.vn; nguyenxuantoai@dntu.edu.vn; lethanhlanh@dntu.edu.vn; lenamthoi09@gmail.com

To cite this article:

Author’s Name Paper Title International Journal of Research in Vocational Studies (IJRVOCAS), X(X), XX–XX https://doi.org/10.53893/ijrvocas.v1iX.XX

Received: Dec 08, 2021; Accepted: Dec 16, 2021; Published: Feb 19, 2022

Abstract: This paper presents the design of an IIR filter on FPGA technology by combining modeling with automation of digital system design, in order to provide a solution to design and quickly implement the digital signal processing system The IIR filter is designed by the method of object modeling (Model Based Design) abbreviated as MBD with the help of Matlab and EDA, then experimentally implemented the IIR filter on FPGA technology for a specific problem and achieved quite good results Through this method, it has helped designers reduce design effort and time, improve economic efficiency, and more importantly, the design is highly flexible, ready to response the new technical requirements of the technology

Keywords: IIR Filter, EDA, FPGA, Matlab, MBD

1 Introduction

Currently Altera is known for its FPGA, CPLD and ASIC

architectures that allow programming on it With a growing

number of customers, Altera's 2006 sales were 1.29 billion

USD Altera's headquarters is located in San Jonse, California

with a staff of approximately 2,600 employees in companies

in 19 countries

An FPGA consists of an array of discrete elements that

can be interconnected in a common way, and like a PLD, the

connections between the elements are programmable That

leads designers to write programs in functional blocks

Because of this idea, software vendors have begun to allow

their software to convert from hard designs into VHDL languages applied to FPGA and CPLD technologies including Matlab, Multiplex sim, Altium Designer, Modelsim, [1-4]

In the early 2000s, in technical journals, articles about a new design idea for FPGAs and CPLDs began to appear, which were later called "Model Based Design" as the article

by Professor: Uwe Meyer_Baese, Ph.D by the University of Florida or by Stranneby and William Walker etc This design method brings high economic efficiency Up to now, almost all application designs have followed this method [5-8]

Quartus is Altera's software development tool that

provides a comprehensive design environment for SOPC

(system on a programmable chip) designs [9] This is a fully

integrated package software for logic design with Altera's PLD

programmable logic components, including APEX, Cyclone,

FLEX, MAX, Stratix Quartus helps designers almost all the

most necessary tools to be able to complete a design project in

the fastest and most efficient way Quartus integrates the most

advanced technologies with flexibility, user-friendly GUI

interface

During this period, FPGA and CPLD played a very

important role in digital system processing, which was the

basis for digital electronic devices It has a strong impact on

fields from Electronics and Telecommunications to

Information Technology, National Security and Defense, etc

and it will certainly contribute significantly to the

industrialization and modernization of our country

This paper presents an IIR filter design with the desire to

provide a product designed on FPGA technology for

convenience, with the shortest time and the highest economic

efficiency A modeled and measured audio noise filter gives

good results

2 Objects and Methods

We know that the spectrum of human voice has energy

concentrated in the frequency range from 0.8Khz to 1.7Khz to

be able to clearly recognize each person's voice, then the

standard telephone bandwidth is from 0.3Khz to 3.4 Khz

Suppose for some reason that the audio frequency signal

(human voice) during this recording process is penetrated by

high-frequency noise, reducing the sound quality, causing the

terminal device (of the communication system to) no longer

recognizes the original signal form And the noise energy is

scattered in the frequency range from 10Khz to 100Khz The

task is to design a filter to remove the above noise from the

signal source to ensure the best output signal quality

In order to reduce or enhance a certain frequency range

of the signal, we often use a filter Digital filtering is classified

into two types FIR (Finite Impulse Response) and IIR (Infinite

Impulse Response) [10-12]

The IIR filter works recursively, using past values of the

output [13-14] That way, we can implement a filter that needs

only low order but the passband has the same slopes before

and after The IIR filter has low amplitude attenuation, the structure is much simpler than the FIR filter

In this paper, the Chebyshev-type2 IIR filter [15] will be designed with the structure shown in Figure 1 with selected parameters

Figure 1 Schematic diagram of the Chebyshev-type2 filter

First, we will create a filter block Filter_Cheb2 with the following input parameters for the filter: multiplier, input and output data parameters, register structure, and arithmetic algorithm for the filter

Figure 2 Input and output data for the filter and Simulink

simulation block for the filter after creation

In Simulink's environment, we set up the simulation

block as follows: Sin signal and noise generator block, filter

block as well as display block as shown in the following

figure:

Figure 3 Setting up the simulation system in

Simulink

The specifications of the Sin generator and the noise

generator are shown in the following figure:

Figure 4 Setting the sine wave and noise block parameters

In the Simulink environment we run the simulation

Simulation results are displayed in Scope, the waveform is

shown from the bottom up as follows:

Figure 5 Waveform of the test points in the system

Waveform 1: Sine signal; Waveform 2: Noise signal; Waveform 3: Sum of the Sin signal + noise; Waveform 4:

Sine signal after filter

Thus, we see that the waveform after passing the filter meets the necessary requirements such as less signal attenuation, the signal after filtering compared to the signal before filtering is equivalent Although the Chebyshev_2 filter

is phase-delayed, it is still acceptable

Looking at the output signal form compared to the input signal form of the filter, we see that the output signal completely responses the technical requirements that have been set out for the filter Thus, this filter can be introduced to realize in FPGA

At this point, you can load it directly on the DE2 board for testing or use the simulation program available on Quartus

II Looking at the test results, we can see that the filter can be realized by the diagram of the FPGA just created Switching from the VHDL language to the principle circuit from the Altium Designer Winter 09 software tool, we get the schematic diagram of the filter principle as follows:

Figure 6 Principle circuit diagram created from Altium

software tool Below is a block diagram of the process of measuring and checking IIR filters on FPGA technology Audio and noise

signals are generated at the transmitter It is then converted to

a digital signal by an ADC converter At the output filter_out

(FPGA) we connect to the DAC converter, the signal after the

DAC is checked by an oscilloscope

Figure 7 Digital filter implementation block diagram

3 Results and Discussion

The IIR filter on FPGA technology is modeled as shown in

Figure 8 Using the computer output to generate a sine wave

(fluctuates in the range of 100Khz or less), then using

Generatosaur software to generate white noise as shown in

Figure 9 Input selection Output white noise (White noise) is

left (LEFT), output right (RIGHT) is sine wave (Sinus) by

selecting Active frequency f(Hz) change according to

requirements

Figure 8 Filter test model

Figure 9 Generatosaur software interface

Use a circuit to combine 2 signals and noise Then put

it into the 12bit ADC of the dsPIC33FJ128GP802 microcontroller, then output the port to the FPGA, the FPGA filters the output signal with the IIR filter designed above, then give the port of the FPGA From here, put into another port of the PIC, use the DAC (16bit) to generate analog signals

The actual experimental circuit has the following parameters: 12-bit ADC with a maximum sampling frequency

of 500Khz equal to 5 times the frequency of the signal The signal wave to be filtered is limited to about 100Khz or less The 16-bit output DAC has a maximum output sample rate of 100Khz, just enough for the input 100Khz signal Mixed waves with amplitudes from 0 to 3V3

The signal obtained after passing through the Chebyshev-type2 IIR filter as shown in Figure 10 has eliminated noise quite well However, the signal is not smooth and has a small bump

Figure 10 Waveform at filter output

4 Conclusion

An IIR filter design on FPGA technology was modeled with the help of Matlab and EDA for a specific problem and achieved quite good results With the desire to offer a product designed on FPGA technology so that it is convenient, with the shortest time and the highest economic efficiency A problem of IIR filters in general as well as digital filters in particular is that the better the quality of the signal filter, the more signal degradation will be However, the signal quality

of the digital filter is much better than that of the analog filter

In order to have good digital filter quality and the least signal

loss, we can use an amplifier, in addition, both signal and noise

must be accurately determined in terms of amplitude and

frequency for the signal selection to be effective

References

[1] Trimberger, Steven, et al "A time-multiplexed FPGA."

Proceedings The 5th Annual IEEE Symposium on

Field-Programmable Custom Computing Machines Cat No

97TB100186) IEEE, 1997

[2] Obaid, Zeyad Assi, Nasri Sulaiman, and M N Hamidon

"FPGA-based implementation of digital logic design

using Altera DE2 board." International journal of

computer science and network security 9.8 (2009):

186-194

[3] Hachour, Ouarda, and Nikos Mastorakis "IAV: A VHDL

methodology for FPGA implementation." WSEAS

Transactions on Circuits and Systems 3.5 (2004):

1091-1096

[4] Navabi, Zainalabedin "PLD Based Design." Digital

Design and Implementation with Field Programmable

Devices Springer, Boston, MA, 2005 3-16 Wilson,

Peter

[5] Design recipes for FPGAs: using Verilog and VHDL

Newnes, 2015

[6] Paterno, Fabio Model-based design and evaluation of

interactive applications Springer Science & Business

Media, 1999

[7] Franceschini, Gaia, and Sandro Macchietto

"Model-based design of experiments for parameter precision:

State of the art." Chemical Engineering Science 63.19

(2008): 4846-4872

[8] Erkkinen, Tom, and Mirko Conrad "Verification,

validation, and test with model-based design." SAE

Technical Paper 2008-01 (2008): 2709

[9] Kleitz, William Digital Electronics with VHDL, Quartus

II Version Pearson Prentice Hall, 2006

[10] Litwin, Louis "FIR and IIR digital filters." IEEE

potentials 19.4 (2000): 28-31

[11] Rabiner, L R., et al "Some comparisons between FIR and

IIR digital filters." Bell System Technical Journal 53.2

(1974): 305-331

[12] Pun, Carson KS, et al "On the design and

implementation of FIR and IIR digital filters with

variable frequency characteristics." IEEE Transactions

on Circuits and Systems II: Analog and Digital Signal Processing 49.11 (2002): 689-703

[13] Yang, Jar-Ferr, and Fu-Kun Chen "Recursive discrete Fourier transform with unified IIR filter structures."

Signal processing 82.1 (2002): 31-41

[14] Montazeri, Allahyar, and Javad Poshtan "A new adaptive recursive RLS-based fast-array IIR filter for

active noise and vibration control systems." Signal

Processing 91.1 (2011): 98-113

[15] Filter, Low-Pass "Analog Filters." Signals and Systems

(2018): 227