DOI: http:/dx.doi.org/10.18180/tecciencia.2015.18.

Development board based on the TMS320F28335 DSP

for applications of power electronics

Placa de desarrollo basado en TMS320F28335 DSP para aplicaciones de

electrónica de potencia

Julian Saavedra1, Jonathan Firacaitve2, Cesar Trujillo3

1Universidad Distrital Francisco Jose de Caldas, Bogotá, Colombia, [email protected]
2
Universidad Distrital Francisco José de Caldas, Bogotá, Colombia, [email protected]
3Universidad Distrital Francisco José de Caldas, Bogotá, Colombia, [email protected]

Received: 11 Aug 2014 Accepted: 14 May 2015 Available Online: 29 May 2015

Abstract

This paper presents the design of a special purpose development board for power applications based on the digital signal
processor (DSP) TMS320F28335. We also propose some considerations for the design of any four-layer printed circuit
line with international recommendations to counteract the EMI (electromagnetic interference) and increase the EMC
(electromagnetic compatibility). Additionally, the main features of each module on the board and their respective
conditioning circuits designed are presented in order for the development board to be useful for any application of power
electronics as motor control, switching power supplies, LED lighting, communications through the electric network, etc.
Finally, a test protocol to verify the performance of the card and the comparison between the final cost per unit and similar
cards on the market were performed.

Key Words: anti-aliasing filter, electromagnetic interference, multilayer printed circuit board, signal conditioning,
sensitivity.

Resumen

En el presente artículo se presenta el diseño de una tarjeta de desarrollo de propósito específico en aplicaciones de potencia
con base en el procesador digital de señales (DSP) TMS320F28335. Se proponen además unas consideraciones para el
36 diseño de cualquier circuito impreso de cuatro capas acordes con las recomendaciones internacionales para contrarrestar
los efectos de interferencia electromagnética (EMI) y aumentar la compatibilidad electromagnética (EMC). Como factor
adicional se presentan las principales características de cada modulo en la placa y sus respetivos circuitos de
acondicionamiento, esto con el fin de que la placa de desarrollo sea útil para cualquier aplicación de electrónica potencia
como control de motores, fuentes conmutadas, iluminación LED, comunicaciones a través de la red eléctrica, etc. A
continuación, se realizó un protocolo de pruebas para verificar el funcionamiento de la tarjeta y la comparación entre el
costo final por unidad y las tarjetas similares presentes en el mercado. Finalmente, se presentan las conclusiones en donde
se resalta que la tarjeta implementada cumple con las características adecuadas para operar en el desarrollo de prototipos
electrónicos de potencia a nivel de innovación y/o investigación.

Palabras clave: Acondicionamiento de señales, circuito impreso multicapa, filtros antialiasing, interferencias
electromagnéticas, sensibilidad.

1. Introduction resources in order to achieve energy coverage for the

entire population. The wide array of renewable energy
Today, renewable energies have gained a lot of importance sources includes solar, wind, geothermal, biomass, tidal,
in the sustainable development of nations. Worldwide and wave power, among others [1].
energy consumption of renewable sources exceeds 25%
and is expected to increase by 50% at the middle of this Thus, the Investigation Group LIFAE (the Spanish
century. Societies are increasingly consuming such acronym for Research Laboratory of Alternative Sources
of Energy) of the Universidad Distrital Francisco José de
*Corresponding Author. How to cite: Saavedra, J.; Firacaitve, J.; Trujillo, C. Development
E-mail: [email protected] board based on the TMS320F28335 DSP for applications of power
Tel: + 57 1 3239300 ext 1510 electronics , TECCIENCIA, Vol. 7 No. 18., 36-44, 2015, DOI:
http:/dx.doi.org/10.18180/tecciencia.2015.18.6
Caldas in Bogotá, Colombia investigates such issues, 2.2. Digital Signal Processor TMS320F28335
specifically photovoltaic systems, solar radiation and
ambient temperature characterization, stand-alone and The TMS320F28335 is a 32-bit floating-point DSP
interconnected systems, and their respective applications. belonging to the sub-family of Texas Instruments C2000
These investigations have been based on energy Delfino, which combines the versatility of a general-
processing, understood as the transformation of renewable purpose microcontroller with the high precision and
energies into electricity suitable for isolated loads or grid performance of a digital signal processor. It is made of
feeding and associated power electronics (topologies of technology CMOS, can operate at a frequency of 150 MHz
DC-to-DC converters, inverters, Voltage-Source and has modified Harvard architecture. From a power
Converters (VSC) and High-Voltage Direct Current electronics perspective, the most important module is the
(HVDC) based on VSC) [2]. pulse width modulation. There are six of these modules,
where each consists of two complementary outputs
To accomplish the above, computer systems with high containing a common 16-bit counter. Each time a sample
processing capabilities are required in order to properly parallel is taken, a channel of each module can work at a
develop control algorithms and perform energy higher resolution (24 bits).
management. Hence, the need arises for low-cost devices
which operate at high frequencies, have low consumption, Furthermore, the DPS has 16 channels for analog-to-
the capacity for signal processing in real time, a high level digital converter module with 12-bit resolution at a fast
of integration, high resolution and are adaptable to specific conversion speed up to 80 ns. The improved quadrature
applications of power electronics, which perform the encoder module is considerably useful for obtaining
processing of energy [3]. information on the position, direction and speed of a rotary
machine. It has three 32-bit timers and has a varied group
Consequently, in this study we seek to analyze a purpose of protocols in the communication port [5]: Serial
board specific for applications in power electronics in Peripheral Interface (SPI), Inter-Integrated Circuit (IIC),
order to support the DSP, which is an electronic device that Controller area Network (CAN), Serial Communication
fully meets the above characteristics. Furthermore, we Interface (SCI).
expect the development of a series of circuits which allow
for peripherals to adapt to the function required, as well as 2.3. Modules adapted to the development board
the adaptation of analog to digital signals, the modification
of the Pulse-Width Modulators (PWM), the conditioning The development board is divided into four groups, as
of general-purpose input/output signals, and the shown in Figure 1. In the ADC module there are circuits
implementation and operation of different communication with analog signals and it consists of anti-aliasing filters,
systems (I2C, SPI, etc.) for the transmission of signals of conditioners, circuits of protection and overvoltage, and
interest such as the signals transmitted to the Digital-to- analog-to-digital converter (ADC). The module named in
Analog Converter (DAC), among others. the figure as “PWM, QUEP. CAP, GPIO” goes to the
digital part consisting of PWM generator, module of
2. Methods and materials capture external events (CAP), incremental rotary encoder 37
interface (QEP) and module to the General-Purpose
Input/Output signals (GPIO). Similarly, the module
2.1. DSP Selection appearing as “SPI, CAN, SCI, IIC” consists of the
communication port conformed by the modules (SPI, IIC,
The wide variety of specialized digital signal processors SCI, CAN). In addition, there is a DAC module and a
in different electric and electronic applications, require module including the polarization sources for both the
taking into account specific criteria for the selection of digital part and analog part.
DSP, such as arithmetic format, data bus, speed of
execution and transmission, memory size, number of
PWM channels, sample rate and resolution of the analog-
to-digital converter. Thus, the selection of specific DSP
provides adequate performance to the system in which the
processor is going to be included [4].

The DSP was selected considering it would have a

floating-point format. It is ideal for applications requiring
flexibility, high dynamic range and accuracy. A second
parameter that was considered was the operation Figure 1 Physical distribution of the modules on the
frequency. In addition, the amount and variety of development board.
peripherals were also taken into account. In this way, and The signal generator module PWM has 12 independent
based on the above, it was established that the Texas channels. Signal conditioning was implemented through
Instrument processor TMS320F28335 was the processor the level shifter CD4504b IC (Integrated Circuit), which
which meets the criteria at an appropriate cost.
 allows for three (3) different output voltage values (3.3, 5, has a Quadrature Decoder Unit (QDU), a Position Counter
15) V. It includes four (4) buttons for handling external and Control Unit (PCCU) and a Unit time base for
interrupts, which are responsible for forcing the output speed/frequency Measurement (UTIME).
level of the PWM signal to a level predefined by software.
The GPIO module was conditioning by the IC of the Because the DSP does not have an analog-to-digital
74LVC245ADWR bidirectional data Buffer. This module converter, the SPI for the exchange of data between the
has 27 shared inputs/outputs and six (6) independent DAC and the DSP was defined, since it offers a higher
inputs/outputs with the output voltage level conditioned as transmission speed than the IIC and the tuple data is 16
the PWM module, as shown in Figure 2 bits. Figure 3 shows the electric circuit implemented for
such convertor. DAC module has a 12-bit resolution
The eCap is exclusively dedicated to the accurate capture channel and an output signal between (3-0) V, conditioned
of external events. It presents six (6) channels that can to the board by the IC MCP 4921.
perform as an auxiliary PWM signal generator.

The eOEP module was likewise adapted as an interface

between an incremental rotary encoder and the DSP. It

Figure 2 Conditioning of the PWM and GPIO signals

38

Figure 3 Electric circuit implemented for the DAC.

.
Figure 4 Adaptation Circuit of the Common Protocol - CAN.
 Figure 5 Conditioning of the SPI standard.
The CAN protocol is based on the Carrier Sense Multiple of the filters and the space on the printed circuit board
Access model with a transmission rate up to 1 Mbps and a (PCB).
range of 40m. It discerns between temporary errors and Of the main topologies [7] [8], Sallen-Key Topology
permanent failures of nodes. It is adapted to the ISO 11898 (Figure 6 a) and Multi FeedBack (MFB) Topology
Standard by the IC SN65HVD232 CAN transceiver, as (Figure 6 b) both contain a single operational amplifier
shown in Figure 4 for a second order filter. Among these two, a sensitivity
analysis was performed [7], defined in Equation 1 where
The SCI is an asynchronous serial communication y represents parameters that mathematically model the
protocol with double buffer for transmission and reception filter and x represents the values of resistors and capacitors
functions. It is adapted to IC SN75179B differential driver conforming the filter, thus finding the topology
and receiver pair, and is ideal for applications with representing less percentage changes in the quality factor
balanced transmission lines in the RS-485 and RS-422 𝑄 and in the natural frequency of the filter 𝜔0 , according
standards. to the percentage changes in the market values of the
passive elements composing it. Topology to be selected
The ICC Module has a transmission rate of between 10 was based on the sensitivity calculations, the resistance for
kbps and 400 kbps. It is compatible with the frequencies and the lower value of the synthesis of the
Semiconductors IIC-bus v2.1 and is decoupled through sensitivities.
𝑦 𝜕𝑦 𝑥 (1)
the IC ADuM1250. 𝑆𝑥 =
𝜕𝑥 𝑦
2.5. PCB Design
The conditioning of the peripheral SPI was performed
through the integrated circuit ADuM1401 of four separate
and isolated channels, which are based on Analog Currently available technology has led to have circuits 39
Devices, Inc., iCoupler® technology, as seen in Figure 5. with a higher level of integration, considerably reducing
the sizes of the PCB and thus increasing the capacity of
This device is characterized by its low power consumption interfering one signal into another and, in general,
(a tenth of the conventional optocouplers, approximately) adverse electrical factors affecting signal integrity.
and ease of operation. It was created to be used in
applications with SPI interface for converting data with a
transmission rate up to 90 Mbps.

2.4. Anti-aliasing filter - ADC Module

Signals acquisition was made by the analog-to-digital

converter, which performs a sampling that must satisfy the
condition of the Nyquist–Shannon sampling theorem. In
case this parameter fails to be met, a phenomenon known
as Anti-aliasing occurs [6]. Therefore, by means of a low-
pass filter, any signal outside the range of interest Figure 6 (a) Sallen-Key Topology, (b) Multi FeedBack
(sampled signal) gets eliminated and the presence of such Topology
phenomenon avoided. PCB design for the implementation of the board and in
general for any multilayer PCB, was based on the
Topology selection focused on the searching of the one following considerations so that the effects of EMI and
involving less active elements, thereby decreasing the cost EMC counteract [9] [10] [11] [12] [13] [14]:
  The outward and return path of a signal should be  In mixed PBC, meaning those composed of analog,
close enough so that the inductance of the circuit is digital and power circuits, these must be distributed
reduced and thus the magnetic field flux generated. in different areas and only if necessary, join them into
a single point. If along the PCB different frequency
 The larger the area of the conductor is the reactance bands are presented, components whose signals have
substantially decreases, thus controlling the effects of low spectral components should be located inside the
electric and magnetic fields. plate. Moreover, components with higher frequency
signals must be located in the periphery of the PCB,
 Creating return planes on the signal layer helps to previous to the connectors.
substantially lower the inductance of the paths and
therefore the total inductance of the PCB. This plane  Likewise, the following parameter for the distribution
catches and drains the radiation generated by the layers was considered: signal layers must be as close
signals interacting in the printed circuit board as possible to ground planes in order for EMI to
reduce.
 If the PCB presents return planes in the outer layers,
must be verified that there are not small isolated  Linear polarization must be adjacent to ground/return
copper areas (island) as these behave likely as planes for adding capacitances between planes, which
microstrip antennas, susceptible to radiate helps to reduce polarization noise and
electromagnetic fields of adjacent signals. Moreover, electromagnetic radiation.
if there are large copper segments that are not
For the separation of copper paths, the 3W Rule is
connected to the ground plane on the return plane,
suggested, which states the spacing distance as three times
these must be grounded via routes known as
the width of the path, between traces centers.
“stitching vias”.

 Linear polarization and analog and digital ground  Avoid 90 degrees angles on paths carrying signals in
planes must be separated, as shown in Figure 7. high frequency band or higher, as these increases the
width of the track with a factor of 1.414, thus
 Location of the parts having connections in common affecting the characteristic impedance at the corners
should be as close as possible and the copper traces of the paths. Thereby, changes in the trajectory
short and straight. In digital circuits, if the above greater than 45 degrees are not recommended.
condition cannot be met, parts that are functionally
related must kept close or the pieces with clock  The use of decoupling capacitors helps filtrating the
signals at high frequencies and very short rising edges radiofrequency noise present in the signal power of
must be grouped. Thus, the size of the paths carrying integrated circuits. To use two or three different types
these signals can be minimized. of capacitors (tantalum, ceramic, electrolytic, etc.)
40 with values decline for decades from 0.1 nF is
 In accordance with the IPC 2251, paths greater than recommended.
λ/15 length, must be designed according to the
transmission lines theory.

Figure 7 Separation of linear polarization and analog and digital ground planes.
3. Results
In accordance with the procedure described in section 2.4,
3.1. Topology definition for the antialiasing filter sensitivities shown in
Table 1 (were recommendation [9] was also followed to filter is given by (2) and whose denominator is a second
decrease sensitivity in the Sallen-Key topology regarding degree Bessel polynomial [20]. By matching this with the
passive elements), were obtained. The statements in this transfer functions of the topologies, statements contained
table are given by the combination of the nominal values in
of the passive elements, which in turn, are in function of Table 2 are generated.
the frequency response required to implement. Such
response is established towards minimizing the distortion 3𝜔𝑐2 (2)
in phase and magnitude to the signals to be sampled [16] 𝑠 2 + 3𝑠𝜔𝑐 + 3𝜔𝑐2
[17] [18] [19]
For such reason, frequency response of the Bessel type By replacing equations of
was selected, whose transfer function for a second order Table 2 in the statements of
Table 1, in Figure 8 are presented the behaviors for the topology, taking into account different orders of
Sallen-Key topology, whose sensitivities are independent magnitude of the product 𝐶1 𝑓𝑐 . (𝑓𝑐 is the cut-off frequency
of 𝑓𝑐 , and in Figure 9 the behaviors shown for the MFB of the filter).

Table 1 Sensitivities Calculation.

𝑦
𝑆𝑥 Sallen-Key Topology Multi feed back Topology
𝑄 𝑑𝑄 𝐶1
𝑆𝐶1 = ∗ 1/2 1/2
𝑑𝐶1 𝑄
𝑑𝑄 𝐶2
𝑆𝐶𝑄2 = ∗ −1/2 −1/2
𝑑𝐶2 𝑄
𝑅2 𝑅3
√
𝑑𝑄 𝑅1 𝑅2 − 𝑅1 𝑅1
𝑆𝑅𝑄1 = ∗
𝑑𝑅1 𝑄 2(𝑅1 + 𝑅2 ) 𝑅 𝑅 𝑅 𝑅
4 [√ 2 3 + √ 2 + √ 3 ]
𝑅1 𝑅3 𝑅2
𝑅2
𝑅1 − 𝑅2 2𝑅1 𝑅2 𝑅3 + 𝑅1 (𝑅2 − 𝑅3 ) + 𝑅2 2 𝑅3 2
2 2

𝑑𝑄 𝑅2 √𝑅1
𝑆𝑅𝑄2 = ∗ 2(𝑅1 + 𝑅2 ) 2
𝑑𝑅2 𝑄 𝑅 𝑅 𝑅 𝑅
4𝑅1 𝑅2 𝑅3 [√ 2 3 + √ 2 + √ 3 ]
𝑅1 𝑅3 𝑅2
𝑅3
2𝑅1 𝑅2 𝑅3 + 𝑅1 (𝑅3 − 𝑅2 ) + 𝑅2 2 𝑅3 2
2 2

𝑑𝑄 𝑅3 √𝑅1
𝑆𝑅𝑄3 = ∗ 2
𝑑𝑅3 𝑄 𝑅 𝑅 𝑅 𝑅
4𝑅1 𝑅2 𝑅3 [√ 2 3 + √ 2 + √ 3 ]
𝑅1 𝑅3 𝑅2
𝑑𝑄 𝑘 𝐶1 𝑅1
𝑆𝑘𝑄 = ∗
𝑑𝑘 𝑄 𝐶2 (𝑅1 + 𝑅2 )
41
𝜔 𝑑𝜔0 𝐶1 1 1
𝑆𝐶10 = ∗ − −
𝑑𝐶1 𝜔0 2 2
𝜔 𝑑𝜔0 𝐶2 1 1
𝑆𝐶20 = ∗ − −
𝑑𝐶2 𝜔0 2 2
𝜔 𝑑𝜔0 𝑅1 1
𝑆𝑅10 = ∗ − 0
𝑑𝑅1 𝜔0 2
𝜔 𝑑𝜔0 𝑅2 1 1
𝑆𝑅20 = ∗ − −
𝑑𝑅2 𝜔0 2 2
𝜔 𝑑𝜔0 𝑅3 1
𝑆𝑅30 = ∗ −
𝑑𝑅3 𝜔0 2
𝜔 𝑑𝜔0 𝑘
𝑆𝑘 0 = ∗ 0
𝑑𝑘 𝜔0

Table 2 Resistance values for frequency response of the Bessel type.

RX Sallen-Key Topology + Bessel Multi FeedBack Topology
1 √3 1 1 √3 1
R1 𝑅1 = ( √3 − 4𝜌 + ) 𝜌 = 𝐶2 /𝐶1 𝑅1 = 𝑅2 = ( √3 − 8𝜌 + ) 𝜌 = 𝐶2 /𝐶1
𝜌𝐶1 𝜔 6 2 𝜌𝐶1 𝜔 6 2
1 √3 1
1 √3 1 𝑅1 = 𝑅2 = ( √3 − 8𝜌 + ) 𝜌 = 𝐶2 /𝐶1
R2 𝑅2 = (− √3 − 4𝜌 + 2) 𝜌 = 𝐶2 /𝐶1 𝜌𝐶1 𝜔 6 2
𝜌𝐶1 𝜔 6

1 √3 1
R3 𝑅3 = (− √3 − 8𝜌 + ) 𝜌 = 𝐶2 /𝐶1
𝜌𝐶1 𝜔 12 4
 𝑸 𝑸
Figure 8 (𝒂)𝑺𝑹𝒙 , (𝒃)𝑺𝒌 Sallen-Key Topology

42

𝑸 𝑸 𝑸
Figure 9 𝑺𝑹𝟏 , 𝑺𝑹𝟐 , 𝑺𝑹𝟑 Multi FeedBack Topology.

Figure 10 Effect of external interrupt (TZ1) over the PWM1x channel.

 3.2. Operating Tests 3.3 Comparison with similar boards commercially
available
Figure 10 shows the behavior of the signals when pushing
the button P1 (associated to TZ1), imposing a high level Although on the market there is a variety of development
for PWM1A and a low level for PWM1B. These levels are kits, evaluation boards and prototypes training, these are
predefined in the TZCTL configuration register associated designed for general purpose applications, thus limiting
to each PWMxx channel. On the other hand, resulting the scope that can be achieved, since they include
signal of the voltage booster was captured in channel 3 of peripherals that are not specific for power electronics
the oscilloscope, which is responsible for increasing the applications, taking up space on the PBC which can be
amplitude up to 15 V. used for conditioning or addition of related modules.

Figure 11 shows the transmission of the word Table 3 shows those commercially available boards,
“hexadecimal 7555”, along with a configuration mode of similar to the ones developed in this project. Even though
the clock signal (0,0), that is to say, the bit transmission is these are specific boards, do not include conditioning of
given by each rising edge without delay. MISO signal is their peripherals such as: different voltage levels for the
captured in channel 1 of the oscilloscope for which the PWM signals and specific purpose inputs/outputs, anti-
logic level at an idle state is high and a transmission rate aliasing filters and signal conditioning for the ADC
of 300 kbps. Channels 2 and 3 respectively record SS and module. Another factor to consider is the cost per unit, due
SCLK signals. to the 50% decrease compared with the cheapest board
registered on Table 3.
Table 3 Comparison of the most representative similar
boards on the market.
COST
BOARD MAIN FEATURES PER
UNIT
Based on the TMS320F28335
DSP, is specifically designed for
industrial automation. Includes
TI28335DSK-II LCD, two communication $ 950 000
terminals RS-232, SD card
interface, switches for interrupting
and Ethernet interface.
Figure 11 SPI Standard, configuration of the clock signal Its main processor is the
CPHASE =0, CPOL = 0. TMS320F28335 DSP. Includes 16
ADC channels for analog signals
ICETEK- $ 1 586
On the third test the ADC module was used, in which a F28335-A
with width from -5 V to 5 V. 88
000
sinusoidal input signal to the filter entrance was input/output of general purpose.
Four (4) audio interfaces and two
introduced with a frequency corresponding to the cutoff D/A0 converters. 43
frequency (600 Hz), with amplitude of 15 V (Figure 12, Motor Control
It provides hardware and software
Ch1). Channel 2 has the output signal of the filter with Development
for motor control, software for
amplitude 10 V. Subsequently, the test signal is attenuated for $ 1 746
using fuzzy logic and PID
TMS320F28335 100
10% and elevated to a DC level of 1.5 V, effectively eZdsp™ with
algorithms. It is built upon the
obtaining a signal amplitude of 1V and 1.5V above the TMS320F28335 processor.
Socket
reference of channel 3 (Ch3). Finally, the test signal is
digitized at a sampling rate of 735.3 ksps and transmitted
to the DAC module for its conversion to analog (channel 4. Conclusions
4, Ch4).
A development board for power electronics applications
based on the Texas Instrument TMS320F28335 DSP,
which meets the proper characteristics to developments of
prototype and research was designed and implemented. It
has the necessary circuitry for setting and conditioning of
different modules associated to the DSP (PWM, DAC,
ADC, SPI, I2C, CAN, SCI, eQEP, eCAP, GPIO)
assembled on a PCB, which was designed based on
recommendations made by international organizations and
scientific articles published in the IEEE to counter the
adverse effects of electromagnetic interference.
b
Figure 12 ADCIN0 Block -Filter- Conditioner- DAC
 Acknowledgments [9] S. Muralikrishna y S. Sathyamurthy, «An overview of digital
circuit design and PCB design guidelines - An EMC perspective,»
de Electromagnetic Interference & Compatibility, 2008.
This paper is the result of the research project “Design and INCEMIC 2008. 10th International Conference on, Bangalore,
implementation of a development board with digital signal 2008.
processor and specific purpose in power electronics [10] N. L. Eastman, «Considerations for mixed analog/digital PCB
design,» de WESCON/96, Anaheim, 1996.
applications” (“Diseño e implementación de una tarjeta de
desarrollo con procesador digital de señales, de propósito [11] J. Jing y K. Lingwen, «Study of signal integrity for PCB level,»
de Electronic Packaging Technology & High Density Packaging
específico en aplicaciones de electrónica de potencia”) (ICEPT-HDP), 2010 11th International Conference on, Xi'an,
partially funded by the Research and Scientific 2010.
Development Center of the Universidad Distrital. [12] J. López Sánchez, F. A. Rojas S., C. L. Trujillo y J. Guacaneme
Moreno, «Recomendaciones para el diseño de circuitos impresos
de potencia,» Revista cientifica y tecnologica de la facultad de
References ingenieria universidad Distrital Francisco José de Caldas, vol. 9,
nº 2, pp. 44-47, 2003.

[1] Varun, R. Prakash y I. Krishnan Bhat, «Energy, economics and [13] See, «Impact of PCB Layout Design on Final Product’s EMI
environmental impacts of renewable energy systems,» Renewable Compliance,» de 17th International Zurich Symposium on
and sustainable energy reviews, vol. 13, nº 9, pp. 2716-2721, Electromagnetic Compatibility, Zurich, 2006.
2009. [14] P. Chand, B. N. Gangopadhyay y M. Kumar, «PCB layout design
[2] P. K. Steimer, «Enabled by high power electronics Energy for EMI/EMC compliance,» de ElectroMagnetic Interference and
efficency, renewables and smart grids,» de Power Electronics Compatibility (INCEMIC), 2006 Proceedings of the 9th
Conference (IPEC), 2010 International, Sapporo, 2010. International Conference on, Bangalore, 2006.

[3] Y. Huang, J. Li y R. Gong, «A novel SVPWM control Stand-alone [15] W. Saraga, «Sensitivity of 2nd-order Sallen Key type active RC
Three-phase PV power system based on TMS320F28335,» de filters,» Electronics letters, vol. 3, nº 10, pp. 442-444, 1967.
Consumer Electronics, Communications and Networks (CECNet), [16] W.-. K. Chen, Passive, Active and digital filters, CRC Press, 2005.
2012 2nd International Conference on, Yichang, 2012. [17] F. Miyara, Filttos Activos. E-book, Rosario: Universidad Nacional
[4] J. Salazar, «Procesadores digitales de señal (DSP),» Mundo de rosario, 2004.
electronico, nº 314, pp. 46-57, 200. [18] P. Horowitz y W. Hill, The art of electronics- 2nd Edition,
[5] Texas Instruments, «Data Manual: TMS320F28335, Cambrige University Press, 1989.
TMS320F28334, TMS320F28232, Digital signal controllers [19] M. Thompson, Intuitive Analog circuit design, 2nd Edition,
(DSCs),» Texas Instruments, 2007. Nownes, 2013.
[6] P. Embree, C algorithms for real-time DSP, Prentice Hall, 1995. [20] J. L. Dominguez Barragan, «Acondicionamiento de Mfilters,»
[7] A. S. Sedra y K. C. Smith, Circuitos Microelectronicos, Oxford: Universidad de las americas Puebla , Puebla, 2003.
Oxford University Press, 1999.
[8] H. G. dimopoulos, Analog Electronic Filters: theory, design amd
synthesis, Springer, 2012.

44