Design of analogue filters using Cypress PSoC

Abstract

In this paper, we are going to learn a new technique to implement the analogue filters using PSoC. It is a mixed-signal array IC chip called Programmable System-on-Chip, employed from Cypress Semiconductors. Previously to design any filter we need to implement the circuit using softwares like Cadence, Mentor Graphics etc and we have to check the functionality of that circuit. Instead of going for a long process, we can use PSoC software which is a simplest method to implement the filters. PSoC makes use of the switched capacitor technology with topology to build second order filters. We can save design time, board space, power consumption by using this technique. Microcontroller, Analog and Digital components are integrated in it. PSoC is also known as Software Configurable Silicon. This paper consists of filters, types of filters, their functions and the new method to implement the analog low-pass and analog band-pass filters.

Full text

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
144
DESIGN OF ANALOGUE FILTERS USING CYPRESS PSOC
1
DHANABAL R
1
, BHARATHI V
2
,POLA SAI KUMAR
3
,PRANEETH MADHIRA SASI RAMA
4
1
Assistant Professor (Senior Grade) ,VLSI division,SENSE, VIT University,
2
Assistant Professor,GGR College of Engineering ,Vellore,
3,4
Btech ECE 2010 batch student ,VIT University, Vellore- 632014,Tamil Nadu, India
E-mail: rdhanabal@vit.ac.in, bharathiveerappan@yahoo.co.in, saikumar1163@yahoo.com,
praneeth.msr2010@vit.ac.in
ABSTRACT
In this paper, we are going to learn a new technique to implement the analogue filters using PSoC. It is a
mixed-signal array IC chip called Programmable System-on-Chip, employed from Cypress
Semiconductors. Previously to design any filter we need to implement the circuit using softwares like
Cadence, Mentor Graphics etc and we have to check the functionality of that circuit. Instead of going for a
long process, we can use PSoC software which is a simplest method to implement the filters. PSoC makes
use of the switched capacitor technology with topology to build second order filters. We can save design
time, board space, power consumption by using this technique. Microcontroller, Analog and Digital
components are integrated in it. PSoC is also known as Software Configurable Silicon.
This paper consists of filters, types of filters , their functions and the new method to implement the analog
low-pass and analog band-pass filters.
Keywords: Bandpass Filter, Butterworth Filter, Chebyshev Filter,Bessel Filter
1.
INTRODUCTION
Filter is a common term which is used to pass the
desired portion of the particle and stops the unwanted
portion of the particle. Filters are of many types. For
example, a fuel filter in a vehicle can pass the desired
gasoline and stops the undesired dust particles. But
we are not discussing but normal filters. Here we are
discussing about the filters which pass the signals of
desired frequency and rejects the unwanted frequency
signals. These filters are known as Electronic Filters.
These filters are designed to separate the desired
frequency signals from the undesired frequency
signals.
2.
FILTER TYPES
Electronic Filters are many types and they are
classified into many ways. The common most
classification method is based on selectivity of filter
frequency. Filters are mainly two types: Analog and
Digital filters. Between the two, digital filters are
most efficient. But to design digital filters, we need
analog filters as basic building blocks. It plays an
important role in development of electronics. Digital
filters do not work well if we do not use analog filter
modules first. In other words, we can say analog
filters are mandatory.
Passive Filters are made up of passive components
like resistors, capacitors and inductors.
Active Filters are made up of operational amplifiers.
Passive Filters are built with (a) Resistor Capacitor
and (b)Inductor Capacitor
(a)Resistive Capacitors are RC Filters. These are
easier and cheaper. Therefore these are most used
ones.
(b)Inductor Capacitors are LC Filters. The
performance of this filters are better than other filters.
There are some limitations which effect these filters.
Inductors are more expensive, to tune the exact value
is very difficult and electromagnetic shield may be
required.
We obtain nth order filters by cascading both the
passive and active filters. Depending on their
functionality both passive and active filters are
classified as (i)Low pass Filter, (ii)High pass Filter,
(iii)Band pass Filter, (iv)Band stop Filter
These filters specifications are based on frequency
range of the signals. The frequency range is classified
into three regions.

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
145
PASS-BAND: Signals in this range of frequencies
are passed with little attenuation. This region extends
from zero frequency (dc) to pass band edge
frequency fpass.
STOP-BAND: Signals in this range of frequencies
are attenuated by at least a specified amount. This
region extends from stop band edge frequency (fstop)
to infinity.
TRANSITION BAND: Signals in this range of
frequencies are present in the region between pass
band and stop band. This region extends from fpass
to fstop. The width of the transition band is
determined by the response type chosen and the order
of the filter.
These three regions determine the filter frequency
range which is used for the filter specifications.
regions. Here fc indicates cut-off frequency, Avs
indicates Attenuation in Stop band region, Avp
indicates attenuation in Pass band region and Hpp
indicates High Power Point of value 0.707 volts.
Fig 1: Lowpass Filter Circuit
(II)HIGHPASS FILTER: High pass Filters are
required whenever low frequencies are needed to be
eliminate from the signal. We can say high pass filter
is the inverse of the low pass filter. In this filter,
signals below the cut off frequency are attenuated
and signals at higher frequencies are passed.
Below Figure 1.2 shows the High pass Filter Circuit
with three regions
Figure 1.2: Highpass Filter Circuit
(III)BANDPASS FILTER: Band pass Filter pass
the band of signal frequencies while attenuating the
frequencies above or below that band. In general we
can say signals between a lower cut off frequency
and an upper cut off frequency are attenuated and
signals at higher or lower frequencies are passed.
Below Figure 1.3 shows the Band pass Filter Circuit
with three regions.
Figure 1.3: Band Pass Filter Circuit
(IV)BANDSTOP FILTER: In this type of filters,
the rejected band frequencies present between the
two pass bands is located. Simply we can say that the
inverse of the band pass filter is known as band stop
filter. Signals at the higher cut off frequencies or
lower cut off frequencies are passed and signals
between the higher and lower frequencies are
attenuated.
Below Figure 1.4 shows the Band stop Filter Circuit
with three regions.
Figure 1.4: Bandstop Filter Circuit
3. Classification of Filters based on Filter
Response
RESPONSE TYPE: Filters can be implemented in
different ways. Each way of implementation has its
own advantages and disadvantages. We have two
kinds of transition bands are there, Narrow transition
band and Wide transition band.
In Narrow transition band, the advantages are it is
very frequency selective in frequency domain and it
requires fewer components to achieve the desired
specification. But the disadvantage is, it has
undesirable overshoot is present and it will ring in
their time domain responses.
When coming to the Wide band transitions, it has a
little or no overshoot in their time domain responses.
But it requires the maximum number of components
to achieve the desired specifications.
In this filter literature, we come across a new concept
called „Brick Wall Filter‟. This is a theoretical
concept which has zero transition band width and
infinite order.

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
146
In this paper, we are going to discuss about some
more common filter response types like Butterworth
response, Chebyshev response, Bessel response,
Elliptical response.
(a)BUTTERWORTH RESPONSE FILTER:
It is also known as Maximally Flat Response because
it has smooth roll-off with no ripple. The width of the
transition band is medium. The filter cutoff frequency
is not too sharp and for the low order filters, the
phase response is not too bad. To compute the pole
and zero locations, it is the easiest filter type as it fall
exactly on the circle whose radius is cutoff frequency
in s-plane.
Below Figure 2, shows the Butterworth filter
response which is most popular filter and it is a
general purpose filter.
Figure 2: Butterworth Filter Response
(b)CHEBYSHEV FILTER RESPONSE:
It is based on Chebyshev polynomial which has equal
ripple about the desired pass-band response. The
desired pass-band response is more accurate and the
width of the transition band is much narrower than
the butter worth response. The order of the filter is
generally less than the butter worth response for a
given specifications. The frequency that the
magnitude response drops below the specified pass-
band ripple is known as cut off frequency of the
Chebyshev Filter. The range of the pass-band ripple
is specified between 0.01dB to 1 dB. There is a
significant overshoot and ringing to a step function.
The phase-response is poor. Whenever the time
domain response is not important, then the
Chebyshev filter is useful.
Below Figure 2.1 shows the Chebyshev Filter
response whose transition slope attenuation is steeper
than the butterworth filter response.
Figure 2.1: Chebyshev Filter Response
(c)BESSEL FILTER RESPONSE:
It is also known as Gaussian Response. It emphasises
linear phase response which is known as constant
time delay at the expense of magnitude response i.e.,
sharpness of cut-off frequency. It makes the filter
very attractive for filtering pulse waveforms by
causing the time-domain response to have a little or
no overshoot to a step function. The transition band
width is wide even when the order is high and it is
very poor at the frequency selectivity.
Below Figure 2.2 shows the Bessel filter response
which is an excellent filter for pulse generator
circuits since it minimizes the ringing and
overshooting responses.
Figure 2.2: Bessel Filter Response
(d)ELLIPTICAL FILTER RESPONSE:
It is also known as Cauer response filter. It is the best
response filter of any other response filters discussed
so far. No other filter will be able to provide a low
order filter for a specifications provided. Elliptical
overcomes this limitation by combining the ripple in
the pass band and stop band. But the elliptical filter is
so difficult to design when compare to any other
response filters.
Below figure 2.3 shoes the Elliptical Filter Response
which is the best analog filter
Figure 2.3: Elliptical Filter Response
4. IMPLEMENTATION OF ANALOG LOW-
PASS AND BAND-PASS USING PSOC
4.1: PSoC: System-on-Chip is one which uses the
configurable hardware surrounding a soft or hard
process core. The Programmable-System-on-Chip is
a mixed signal arrays that integrates microcontroller,
configurable analog and digital components and
programmable interconnections. PSoC is also known
as software configurable silicon. It can integrate as
many as peripheral functions with microcontroller

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
147
which is used to save the time to design board space
and power consumption. The analogue peripherals
like amplifiers, adc/dac, filters, comparators etc and
the digital peripherals like timers, counters, PWM,
SPI etc can be easily configured.
Below figure shows the logic diagram of PSoC
architecture. It comprises of four main areas. They
are PSoC Core, Digital System, Analog System and
System Resources. It also contains the M8C CPU
Core, memories like flash, sram. It has an inbuilt
internal oscillator of 24MHZ. The PSoC CY8C29x66
family have five IO ports which connect to the global
digital and analog interconnects and it provides
access to 8 digital blocks and 12 analog blocks.
Figure 3: Basic Logic Block of PSoC Architecture
PSoC Core: It contains CPU, memory, clocks and
configurable General Purpose IO (GPIO). The CPU
core is a powerful processor with speed 24MHZ,
providing four MIPS 8-bit Harvard architecture
Memory provides 16K of Flash for storing the
program, 256 bytes of SRAM for storing data, and up
to 2K of EEPROM using the Flash. Program Flash
contains four protection levels on blocks of 64 bytes,
allowing customized software IP protection.
The PSoC device provides flexible internal clock
generators which includes a 24 MHz IMO (internal
main oscillator) accurate to 2.5% over temperature
and voltage. The 24 MHz IMO will be doubled to 48
MHz by using the digital system.
PSoC GPIOs provides connection to the CPU, digital
and analog resources of the device. Each pin drive
mode will be selected from eight options, allowing
great flexibility in external interfacing.
Digital System: The digital system contains 8 digital
PSoC blocks. Digital peripheral contains 8 to 32 bits
of PWMS, Dead band zone, Counters and Timers.
Through series of global buses, the digital blocks can
be connected to any of the GPIO that can route any
signal to any pin.
Analog System: The analog system contains 12
configurable blocks. Each block allows complex
analog signal flows by comprising an opamp circuit.
It contains upto 4 bit Analog-to-digital converters
with 6- to 14-bit resolution as Incremental, Delta
Sigma, and SAR), 2,4,6 and 8 pole band-pass, low-
pass and notch filters, upto 4-bit amplifiers with gain
of 48x, comparators of 4 bit with selectable
thresholds, upto 2 bit instrumentation amplifiers with
gain of 93x, DACs with 6 to 9 bit resolution.
4.1: Design Of Analogue Low-pass and Band-pass
Filters:
Switched Capacitor technology is used in the PSoC
with topology to build second order filters. To get 4th
or higher order filters, we can cascade two or more
order filters. The column frequency can control the
cutoff frequency or resonant frequency as well as
capacitor values in the switched capacitor
technology. The filter coefficients can be calculated
by using Filter Design Wizard in the PSoC Designer.
In this paper, we are designing the analog low pass
and band pass circuits with butterworth response. The
design of those filters is described in the following
steps. The cutoff frequency of the low-pass filter is
14KHZ and the resonant frequency of the bandpass
filter is 14KHZ with 1KHZ bandwidth frequency.
We have to choose both the LPF2 and BPF2 from
filters in user modules and we have to set the global
resources as well as module parameters. Then select
the input and output terminal pins and assign them to
low-pass and band-pass filters accordingly.
As mentioned earlier, LPF2 and BPF2 are based on
switched capacitor technology. Therefore column
frequency and capacitor values are the key elements
for the proper filter design. From global resources, a
4MHZ clock is sent to the first analogue column for
LPF2 and 800KHZ is assigned to the 4th analogue
column of the BPF2 to meet the desired filter
requirements.
Below figure shows the block diagram of the Low-
pass and Band-pass filters design using PSoC.
Figure 4.1: Design Of Low-Pass And Band-Pass Filters
Using Psoc

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
148
4.2: Analogue Low-pass Filter:
The steps involved in designing an analogue low-
pass filter are described below. The design wizard of
the analogue low pass filter with 14KHZ cutoff
frequeuncy is shown below.
After placing the LPF2 from the user module, right
click the mouse to pop up the design wizard. Fill the
details of corner frequency, sample frequency, gain
and type of response. Normally butterworth response
type is chosen to get the maximally flat magnitude
response.
Figure 4.2: Design Wizard Of Analogue Low-Pass Filter
Check the Calculated Values to determine whether
the calculated damping coefficient, gain are satisfied
or not. Also, make sure that the calculated column
frequency is feasible. Calculate the over again if any
Calculated Values are not satisfied. After everything
is satisfied, hit “Apply” bottom to obtain a set of
coefficients to switched capacitors in the low-pass
filter topology. The final step is to obtain a specified
column frequency, 4 MHz in this case, in order to
accomplish desired low-pass filter.
4.3: Analogue Band-pass Filter:
The step involved in designing the analogue band-
pass filter is same as the designing low-pass filter.
The design wizard of the band-pass filter is shown in
figure below.
After placing the BPF2 from user modeule, right
click the mouse to pop up the design wizard. Enter
the filter Parameters with required centre frequency,
gain and bandwidth. Then, apply a suitable sampling
clock frequency and value of switched capacitor C2
to match green expected curve to the blue nominal
curve.
Check Calculated Values column to check if
calculated frequency selectivity and gain are satisfied
or not. Also, make sure that the calculated column
frequency is satisfied. Calculate over again if any
item of Calculated Values are not satisfied. After all
values are checked, hit “Apply” bottom to apply a set
of coefficients to switched capacitors in the band-
pass filter topology. The final step is to obtain a
specified column frequency, 800 kHz in this case, in
order to meet the desired band-pass filter.
Figure 4.3: Design Wizard Of Analogue Band-Pass Filter
After completing the design wizard level, now the
output should be verified. It can be done by
programming the software in C language. The
compiled software is then programmed into a PSoC
chip to obtain the required low-pass and band-pass
filter outputs.
5. RESULTS
5.1 Low pass Filter:
The output waveform of the Low-pass Filter is shown
in the figure 5.1. The LPF2 shows the output with
respect to 10KHz.All the high frequency inputs
above the 10KHz are attenuated.
Fig 5.1:Lpf2 Waveform
5.2 Band pass Filter:
The output waveform of the Band pass filter is shown
in the figure 5.2. The BPF2 shows the output with
respect to 10 KHz. Only those
inputs around the center frequency (10 kHz) are
allowed to pass through.

Journal of Theoretical and Applied Information Technology
10
th
February 2014. Vol. 60 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved.
ISSN:
1992-8645
www.jatit.org E-ISSN:
1817-3195
149
Fig 5.2:Bpf2 Waveform
CONCLUSION
In this paper, we can conclude that designing of
Analogue Filters using PSoC is an easiest way
instead of going for other software tools like
Cadence, Mentor Graphics etc.. In other tools, we
want to design the circuit, should apply the
parameters values and then we need to write the code
to implement the design.
But in Cypress PSoC designer everything is inbuilt. It
contains both Low-pass and Band-pass filter designs.
Notch filters cannot be designed in this software. For
low-pass and band-pass filters, four design topologies
are present. We can choose the required topology
according to our requirement. The interconnections
can be changed according to our requirement. We
just need to connect the required IO pins. The code
required for the desired filter can be obtained from
the datasheet present in the properties of the filter.
We need to modify the code according to our filter
design.
Finally we can say that, by using the Cypress PSoC
software, we can reduce- the time required to design
the filter, space of the board, power consumption.
Not only filters, we can design ADCs, DACs,
Timers, counters, PWMs etc..
(I) LOWPASS FILTER: Whenever to limit the
high-frequency content of a signal, we need lowpass
filter. It will stop all the frequencies greater than the
cut-off frequency. Simply we can say that signals
from dc to upper cutoff frequency are passed and
signals at higher frequencies are attenuated.
Below Figure 1 shows the low pass filter circuit with
pass band, stop band and transition band
REFRENCES:
[1] Chia-Chang Tong, Wu-Shun Jwo, Jhih-Yu Lin,
Shih-Fan Li, Juin-Yi Li ,“The Firmware Design
Of Analogue and Digital Filters”, Digital Signal
Processing Workshop and IEEE Signal
Processing Education Workshop (DSP/SPE),
2011 IEEE , 4-7 Jan. 2011 ,pg 523 – 528,
[2] PSoC Mixed Signal Array Final Data Sheet,
Document 38-12013 Rev. H, February 15,
2007, available at: http://www.cypress.com/pso.
[3] “Practical Analog and Digital Filter Design”,
Textbook by Les Thede January 2005,
November 1, 2004 | ISBN-10: 1580539157 |
ISBN-13: 978-1580539159.
[4] “Frequency Filters” Textbook by Kenneth A.Kuhn
[5] “Handbook &Manual for Programmable System
on Chip Lab” by Dr MK Deshmukh, Aalap
Tripathy.
[6] R.Dhanabal, V.Bharathi, G.Prithvi Jain, Ganeash
Hariharan, P.Deepan Ramkumar, Sarat Kumar
Sahoo”, Gabor Filter Design for Fingerprint
Application Using Matlab and Verilog HDL”,
International Journal of Engineering and
Technology (IJET),2013.