Final Report Design 1

8/8/2019 Final Report Design 1

1/107

FUEL CELL INVERTERINTERFACE GROUP

SENIOR DESIGN

FALL 02-SPRING 03DR. RICHIE

TEAM 16

BRIAN QUARTERMAN

DAVID NEWSOM

HECTOR VARGAS

SPONSORED BY DR. ISSA BATARSEH


2/107

Page 2

TABLE OF CONTENTS

TABLE OF CONTENTS .......................................................................................2Executive Summary............................................................................................4Introduction.........................................................................................................6

Competition Vision and Goals ...........................................................................6Significance.......................................................................................................7Our Vision and Goals ........................................................................................7Our Project ........................................................................................................7Groups Responsibilities Diagram .....................................................................8Inverter Specifications .......................................................................................9

Projects DSP ....................................................................................................12Introduction .....................................................................................................12

DSP Background.............................................................................................12Digital Signal Processing.................................................................................13ADC & DAC.....................................................................................................13DSP vs. Microprocessor..................................................................................14DSP Language................................................................................................15DSP Fault Protections .....................................................................................17DSP Filter ........................................................................................................17DSP Communications with Project ..................................................................20Communication with Interface .........................................................................21DSP Fault Protections .....................................................................................24

Interface Design................................................................................................25

Introduction .....................................................................................................25Interface Auxiliary Power Source ....................................................................29Battery Safety..................................................................................................32DC-DC Topology .............................................................................................33DSP.................................................................................................................33Voltage and Current Flow................................................................................36Calculations.....................................................................................................36Schematics of Design......................................................................................41Components....................................................................................................42Layout Design .................................................................................................45Serial Connection ............................................................................................47

Design Results ................................................................................................47Inverter...............................................................................................................49Introduction .....................................................................................................49Half-Bridge Inverter Circuit ..............................................................................51Previous Designs Using the Full-Bridge Inverter.............................................65

Projects Inverter Design..................................................................................67Introduction .....................................................................................................67Simulation .......................................................................................................72


3/107

Page 3

Simulation Results...........................................................................................72Fuel Cell.............................................................................................................80

Introduction .....................................................................................................80What is a Fuel Cell? ........................................................................................81Basic Behavior and Components ....................................................................81

What Makes Fuel Cells Favorable...................................................................82Their Performance...........................................................................................84Key Variables affecting their performance.......................................................86Materials and Design Approaches...................................................................87Reactant Gas Composition and Utilization ......................................................88Overview: ........................................................................................................88Fuel Cell Summary..........................................................................................90

Design Issues / Troubleshooting Guide .........................................................92Introduction .....................................................................................................92

Power Stage ................................................................................................92Driver Ground: IXDD414..............................................................................93

Controller Power Supply: IR2110.................................................................93Switching Conditions ...................................................................................94Labeled Inductor Value................................................................................95

Summary & Conclusion ...................................................................................98Administrative.................................................................................................100

Acknowledgements .......................................................................................100Fall Semester ................................................................................................100Spring Semester............................................................................................101Budgeting and Parts Acquisition....................................................................103

References ......................................................................................................105Appendix Table ...............................................................................................107


4/107

Page 4

Executive Summary

The United States Department of Energy and United States Department of

Defense presented the International 2003 Future Energy Challenge competition.The purpose of the competition is to create a DC-AC inverter system. Thefollowing documentation describes the specifications and the hardwaredeveloped.

The objective of this competition is to produce a cost-effective, manufacturable,innovative DC-AC fuel cell inverter system to be used in homes and businessesaround the world. This, in turn, would reduce the electrical consumption fromtraditional utilities and provide power to newly connected homes in developingcountries. An additional benefit to the use of fuel cells as a power source is thatthey provide an ultra-clean power source, benefiting not only the user, but also

the environment of the world.

The goal of the competition is to produce an inverter system which producespower at a targeted cost have less than $40 US/kW for a 10 kW system. Thepower source is a fuel cell that is provided by the Department of Energy (DOE)for the purposes of this competition, which is built by Fuel Cell Technologies.This 5 kW solid-oxide fuel cell, supplemented with a 5 kW battery in order tomeet extended-duration demands for power in excess of 5 kW, as well as short-term transient power loads. This yields the stated 10 kW rating. The emphasisis placed on efficiency because this will affect the final cost and size of thecomplete system.

The competitions creators vision is to encourage the development of domesticenergy sources by competing designs, which employ fuel cell technology. Thedesigns should aid in progress toward the advancement of this technology aswell as improving the engineering education for the students involved.

The University of Central Florida Engineering School, along with Issa Batarseh,Associate Professor, gathered three teams of Senior Design students to handlethe three major portions of the system. These teams were assigned the areas ofDSP, inverter design, and inverter interface. The attached document outlines theresearch and the design for all groups, majority being the inverter interface.

Research in the area of digital signal processing (DSP) is amazing. DSP hasbeen around since 1960s but not many controllers have ever made moreimprovement then any other controller. Now DSP controllers are being built tohave the same abilities as a microcontroller along with its current abilities. ADSP controller have parallel communication capabilities for both transmitting andreceiving, along with its high level computations of algorithms makes itsusefulness in any high performance project possible. Controller unit inside the


5/107

Page 5

DSP give it the ability to transfer signals from analog to digital (ADC) or digital toanalog (DAC). This ability to do ADC and DAC make it possible to have twodifferent signals on a project and computation of each mathematical problem.DSP has the ability to do certain filtering techniques, which has been donethrough circuits before. This cuts back on building cost and the cutting back in

production size, not having to layout a circuit.

Research in the area of the inverter entails two types of inverter design, the half-bridge and full-bridge inverter. Both inverters are employed to convert a DCsource voltage to an AC output voltage. The benefits of the half-bridge inverterinclude simple topology, compact design, ease of control, and low cost.However, the output voltage is limited due to these factors. The full-bride inverterrequires a slightly more involved topology and more elaborate control system.The benefits of the full-bridge inverter lay in output control. A waveform isproduced which is closer to the desired AC waveform.

Research in the area of the interface yielded many options to steady the inputvoltage to the inverter system. Options include the placement of anultracapacitor in line with the input voltage, a series of batteries placed in thesame manor, and an active filter and battery system as a precursor to theinverter. The use of the ultracapacitor produces a severe oscillation aftertransient loads, rendering it insufficient for this project. Placing batteries in linewith the inverter provides acceptable steadying of the input voltage in the shortterm, but damage to the batteries occurs over time. The use of an active filterand battery is much more complex than the other options explored, but itprovides the most desirable input voltage to the inverter, while protecting thebattery from damage in extensive use.

The final design proposal includes the use of digital signal processor controller, aSPMW full-bridge link inverter in combination with a high frequency transformerand cycloconverter, and an active filter and battery. In the current projectdescription, DSP is to be connected to the interface, being the other portion ofproject was dropped along with the competition. The interface unit and DSPcontroller has the ability to have fault protections and other routines. The activefilter and battery combination was chosen due to its ability to provide the requiredsteady input voltage to the inverter while protecting the battery over long periodsof time.

Keeping the efficiency ratio the same and cost down on our design is a prioritystill after the competition was dropped because the lack of funding for testingequipment and other related materials. The final design of the interface systemwas tested in the Power Electronics Lab at the University of Central Florida andis currently producing a power output of 100 watts.


6/107

Page 6

Introduction

Competition Vision and Goals

The Department of Energy and all the other sponsors have a visionassociated with the success of the 2003 Future Energy Challenge. The vision isto encourage the development of new technologies that will minimize the cost ofinverters to be used in domestic energy systems; also, to incorporate practicalityand affordability into the competition process; show technical progress towardsthe goals pre-established by the competition; and to improve engineeringeducation and promote learning through innovative team-based engineeringsolutions to complicated technical problems.

The same way the organizers and sponsors of the Future EnergyChallenge have a vision, they also have some goals they want met in order forthe competition to be considered a success. Their goals are for the teamsparticipating to build an inverter with the following characteristics: that have amanufacturing cost of less than forty dollars per kilowatt per unit; that achievemaximum efficiency; achieve minimal size and weight requirements; and willminimize cooling requirements.


7/107

Page 7

Significance

The importance of the project lies on the need for clean, inexpensive, andreliable energy source. If not for the need of this type of energy source in our

nation, then also to provide efficient, cost-effective electricity for homes indeveloping nations. Providing electricity of this sort for under developed nationsimplies involving low-cost local energy sources, and stresses how important it isto have innovations that would allow small amounts of power to make greatimpact on standards of living.

Fuel cell technology provides just what is needed in terms of clean andreliable energy source. Although fuel cells provide clean and reliable energy, theyare not cost-efficient. It is an expensive technology, and for years researchersand developers have tried to make the technology more cost-efficient. The ideais that in the future, household electricity could be provided by means of fuel cell

technology. The teams would not have to provide the fuel cell itself, just thesystem that would be responsible to change the energy that the fuel cell puts outfrom DC voltage to AC voltage, the one capable of being handled in our homes.

Our Vision and Goals

Our goals and vision as a senior design project were kept in line withcertain aspects of the inverter design. Our groups main goal was to carry overcertain specifications that the competition required, even though University ofCentral Florida dropped out of the competition due to the unavailability ofequipment to test the circuits. This research been completed, will lead to futuredevelopment and encouragement in other challenges.

Our Project

The project set before us was to create three different groups: DSP,Inverter, and Interface. The Interface group, our research area, was to find away of producing a steady state DC source from an unsteady DC steady system.


8/107

Page 8

Groups Responsibilities Diagram

Simulation(Corporation

with thesimulation

group)

InverterGroup I

InverterGroup II

DocumentSearch

Twosolutions

comaprision

Simulation(Corporate

withsimulation

group)

Theoreticalanalysis and

proptypebuilding

Theoreticalanalysis and

proptypebuilding

Paper readingunderstanding

of theoperationprinciple

DocumentSearch and

pricipleunderstanding

MathematicalModeling

Pspicerealization

Interfacesimulation

Inverter

Systemintegration

Packaging

SimulationGroup

communicationwith computer

by RS232

invertercontrol

interface

DocumentSearch and

pricipleunderstanding

DSPGroup

(Reprinted with permission from Songquan Deng)


9/107

Page 9

Inverter Specifications

The design concept should aim to have a 100 W residential powergeneration system. Since there is a slow dynamic response from a fuel cell, there

is a 48-volt battery pack that will serve as supplemental energy source to supplytransient loads. There are other specifications that our Interface group kept inaccordance with the 2003 Future Energy Challenge, along with some changesthey made to meet the budget requirement.

Design items and their specifications as stated on the Future Energy Challengewebsite:

1. Manufacturing cost: less than $40/ kW, which would be $4.00 for the100-Watt design in mass production.

2. Complete package size, this includes (DSP, Inverter, and Interfacecircuits): volume less than 88.5 L

3. Complete package weight: mass less than 30 kg, not taking the fuel cellinto account. (66 lbs.)

4. Output( nominal) power capability: 50-75 W continuousOutput (overload) power capability: 100 W overload for one minute

half from the cell, half from battery.

5. Output voltage from inverter: 120 volts-nominal or 240 volts for split-

phase.

6. Output frequency: Around 60 Hz.

7. Output voltage regulation quality: Output voltage tolerance no wider than

6% over the allowed line voltage and temperature range.

8. Input source, the solid oxide fuel cell: 22-41 DC volts, 24 DC volts-nominal, no more than 275 amps from the cell.

9. Maximum input current ripple: 3% rms of rated current.

10. Battery auxiliary power: 48 volts DC nominal +10%-20%, with nominalrating of 1 work hour. Battery can be used as a provisional energysource for control power. Charging and charge management must besuch that the overall charge is unchanged at the end of a 24 hour testsequence.


10/107

Page 10

11. Overall energy efficiency: Should be higher than 90%, it should go up topeak power and lowest power with hardly any efficiency degradation.

12. Protection: The inverter should be protected for: going over the current,over the voltage, against short circuit, over the temperature, and under

voltage. There should not be any damage caused in case of output shortcircuit either. If at any moment the input voltage goes below theminimum, then the inverter should shut down.

13. Safety: There should not be live electrical elements exposed when theunit is fully configured. The system is projected for safe, everyday use ina home or small business by non-technical customers.

14. Grid and source interaction: There should be no grid and sourceinteraction. The inverter is proposed as a stand-alone unit for remotepower or backup power.

15. Communication interface: Control communication between fuel cell andinverter through RS232Digital signal processing group will be incharge. Any software to indicate the inverters internal data should beprovided by the team.

16. Storage temperature range: 20 to 85 C.

17. Operating ambient temperature range: 0 to 40 C.

18. Cooling: Air cooled

19. Shipping environment: Can be shipped by air or freight truck.

20. Acoustic noise: Should not be any louder than a home refrigerator; lessthan 50 decibels, measured 1.5 meters from the unit.

21. Lifetime of the unit: The unit should function for at least 16 years whilereceiving routine maintenance and while operated at a preferred ambienttemperature range of 20-40 degrees Celsius.

22. Technical report: The report should include design, simulation,experimental results, lifetime analysis, and cost study.


11/107

Page 11

There are also some design specifications associated with the communication tothe fuel cell controller through the RS 232 link. There are required signals.

The signals are as follows:

1. DC Fuel Cell Voltage2. DC Fuel Cell Current3. DC Battery Voltage4. DC Battery Current5. DC link voltage6. AC voltage7. AC current8. KVA output total from inverter9. kW real output power from inverter10. Run11. Inverter fault, or can also be PCU fault12. Slew rate- in order to control the fast step load changes, energy is first

drawn from the batteries. An adjustable slew rate will be transmittedfrom the Fuel Cell controls to the inverter. This rate will calculate therate at which the inverter transmits the load from the batteries back tothe Fuel Cell (Amps / Sec). The slew rate is needed because the FuelCell needs a period of time to adjust both the Fuel and Air Flow ratesas the Fuel Cell output current increases. The rate of change of FuelCell Output Current < slew rate.

System Diagnostic on LEDs should read: Run or Fault. The operators should beAC line breaker and DC fuse links. There should be some discrete time to livecontrol signals going to the controller from the cell. Those control signals are:

1. Enable Inverter: this tells the Aux bus to send the power to output.2. Enable Grid Connect Mode3. Enable Stand Alone Mode4. Enable Battery Charging5. Enable Battery Equalize Charging

Software protocol should be vendor specific for the prototype.


12/107

Page 12

Projects DSP

Introduction

To understand Digital Signal Processing, which is just a technique thatwas produced to transmit and receive a digital signal. DSP stands for DigitalSignal Processing, in that it improves the precision and consistency of the digitalinteraction processing. Any DSP circuit is technologically constructed to performa processing to make a distinction between different noises. Not in a noise thatwe would hear, as in someone yelling but a signal (figure 3.20).

All noises, analog ordigital, can come from manydifferent external things, such

as appliances. Everyappliance, especially things thatrun thru wireless features havesome sort of noise. Noises canalso be caused by a lightingstorm, which can affectcommunications. When usingwireless communications toprocess information, a noisecan cause information lost.Engineers and other digital communication have been working on the issue toresolve this problem. One solution that has been released is to minimize thesignal bandwidth. The less bandwidth a signal can utilize for communications,the more reliable processing of information can be produced. The other problemis dealing with the downfall of using a smaller bandwidth. This downfall is thespeed of transmitting the information across the line. Other lines, such as fiberoptics technology, are far less subject to signal noises.

DSP Background

When computer first came about back in the 1960s, DSP was used inmostly sonar and other governmental agencies. The knowledge andunderstanding didnt come about later when the boom of personal computershappened in the late 1970s. New applications of DSP boosted to high levelsand the new technological revolution begun. In the 1980s, DSP was mainlytaught as a graduate level course, but over time the course has now become acurriculum in the undergraduate level. If you have knowledge, research or other,

Figure 3.20 Diagram of Noise in lines


13/107

Page 13

of DSP as an engineer or scientist, you career can be promoted. Now in a lot ofhigh technical companies having this knowledge is almost required.

The author, Steven W. Smith states that typical issues you will find inlearning DSP are the challenging equations, obscure mathematical symbols and

unfamiliar terminology. This is all of one big nightmare. In our groupsindependent research, we wanted to learn the basic of DSP, which Smiths bookon the Guide to Digital Signal Processing gave us an outline for.

Digital Signal Processing

Digital signal processing has a wide verity of applications its good for.Some include its ability for numerical analysis, decision theory, probability andstatistics. Not only does DSP have the capability to overcome all the above

features, but also analog and digital processing has a major effect of its greaterabilities.

Normal methods try to produce a low level of noise for informationtransmitted and received can be processed in the more accurate manner.Solution to the noise level in a line is digital signal processing. Digital is moreeffective then analog because the well defined signals, instead of using awaveform. A method to lower the signal-to-noise ratio (S/N ratio) was done byincreasing the signal power and receiving signal. The only issue, with every up,there is a downfall. The downfall on DSP is the signals that are not strongenough to override that of the noise, the noise takes over. A noise does notattack, but if the DSP cant find some sort of stability in the information beingreceived, then no signal is received. This is not good, because the informationbeing transmitted may be important, but at least the signals coming acrossshould be acceptable.

ADC & DAC

Received signals can be in different formats. To have the ability toconvert these formats upon receiving them is greatly considered necessary forcommunications. Many different components have different way ofcommunicating with other components, so DSP helps control this process. Twoformats the DSP has standards for is analog to digital (ADC) and digital to analogconverters (DAC).

Before discussing the conversion process of ADC and DAC, lets talkabout what a signal is for DSP. An analog signal is gathered by a measure of avoltage across time, which is known as a continuous signal. This is then putthrough an ADC which changes the unit of data, giving it a new form, which is


14/107

Page 14

discussed later, but described at discrete signal or digitized signal. Tounderstand more clearly, digitized signal are related to computer communicationswhile continuous signal relates to nature or man.

Analog to digital is a conversion process that is taking an analog signal

and changing it to a digital format without changing any content. Analog signalsare an unbounded number of values that can be represented in many differentwaveforms. An example waveform is the sine wave. The conversion process toan analog-to-digital is done by internal computations. This is then converted toan output power, which is in powers of two. Example of powers of two is two,four, eight, and so on. The easiest digital processing is binary, which consist ofones and zeros. These two states are states ON or OFF.

Digital to analog state is conversion process in the reverse direction thenADC. DAC is usually in the format of two states, or some set process. Anexample of a DAC is a modem in a computer. A computer creates a digital

signal, which is converted into an analog signal so a transmission over a twistedpair is possible. A twisted pair referred to here is a lines used in telephonecommunications. That is why the world can dial into a computer around theworld, because twisted pair lines are almost everywhere.

To a human, digital communication is not something is easily read. Beingit consists of ones and zeros, and a long transmission may have loads of datacoming across, DAC coverts this information with no readability issues. Theseones and zeros can make up anytime from voice, pictures, to a music file. This isquite amazing, in that no communications such as my words can be places overa one or zero, just like an on or off switch.

The DSP has the ability to not convert data if needed. If the data or signalis already in digital format then the DSP acts independently, not needing theanalog to digital or digital to analog converter. By not using the ADC or DAC,lower noise level is produced, therefore minimizing number of signal errors.

DSP vs. Microprocessor

To give a brief overview of the difference in a microprocessor and a DSPcontroller is the design in performance. DSP can use numerous amounts ofnumerical values to process rigorous & repetitive tasks. On the other hand,microprocessors and most other processors are not developed to transmit thesetypes of task but more designed for control-oriented applications.


15/107

Page 15

Applications that show the performance of a DSP are presented by its:

ability for high-performance, usually having two or more multipliers. focus on addressing modes, which help assist the modifications of points

and other bit modifications. ability to address memory more then once in one instruction cycle. control over execution statements, this includes ability for loop control, not

needing to recall each time. Includes updating the current counter. to have irregular instruction sets. This includes the ability to do parallel

computations along with data locations modifications.

The hardest part is not finding a processor but find a processor that will bebest suited for the needs of completing a task for the project. If your need theability of both a microprocessor and DSP, some newer microprocessors, such as

GPP/ MCU now carry DSP features.

DSP Language

Programming in digital signal processing language is been increasingeasier. When programming for DSP first came about, it was done in some sort ofassembly language. This was more difficult then today, were it may take one lineof todays programming language to complete a task of five to ten command linesin assembly. Other language used, besides assembly is C and BASIC. BASIC

programming being the most widely used language, for those programmers thathardly write in any language.

The language is not the key topic on choice but precision. If precision isneeded, certain languages cut or do weird round off computations. Other keyissues are the bit formations and processing speeds of any type of computation.An example of a computation may be adding two digits together. The addition of1 and 0.00000001 would come out to be 1, when really we would like for theanswer to show 1.00000001. Therefore, before doing any computations on acomputer, make sure an understanding of how digits are manipulated beforedoing meaningless amount of programming. This will save both time and money.

When programming in DSP, it can be divided into levels of sophistications.To give an outline of the three, which are Assembly, Compiled and ApplicationSpecific, see table (table 3.20) below.


16/107

Page 16

GeneralRegisters

Size inbits

OtherRegisters

Assembly AX, BX, CX, DXEx. AX 1234

16 SI, DI, SP,BP, CS, DS,

SS, ES, IP

Directcommunication

with hardwareCompiled Variables, suchas A=1234

32 Any variable,which can be

in alphanumerical

point

Easiertransportation,

easier tounderstandwith internalknowledge

ApplicationSpecific

Just likeCompiled, but

suppliedapplication with

routines/algorithms

32 Any amount,usually donein C, for high

level

languages

Built infeatures.

Package ofroutines/

algorithms forprocessing.

Using an application specific programming language doesnt alwaysmeans its own independent language, but pre-defined routines and algorithmsmay be supplied. Other tools, I/O support, or debugging applications to help youovercome your objectives. Using an application for the manufacturer can alsoinclude more intensive packages for testing the spectral analysis,instrumentations, and the digital filter simulations.

Issue with using a compiler or application specific platform is theprogrammer who built each of these. If you think about it, each routine that wasbuilt was discovered in a process of having a problem in the first place. So thismeans, do you think every problem has been discovered and the problem thatyou face now is something discovered. If it is discovered, do you think theprogrammer did it correctly? Wow, a lot to think about, therefore; using assemblylanguage may be the correct path for languages. Just remember the basic, thateach high level programming language has drawbacks on the memory usage,speed, and numerical precision.

Table 3.20 Levels of So histications


17/107

Page 17

DSP Fault Protections

The digital signal-processing group can supply a setup of fault protections sonothing harmful is done to the interface and load in case of a malfunction. Such

fault protections as listed in block diagram (figure 3.21), show a fuel cell tripping.

This beginning routine shows that a fuel cell, for some reason, has a faultand trip the circuit. This sends a signal to the DSP controller for shut down of thefuel cell. To run this, the power is switch over to the battery backup power so theinverter can continue to try and supply power. If the battery falls below a presetlevel, then the controllers goes into shut down mode, first stopping the powerconsumption. After the inverter has been stopped, the controller leaves on fuelcell fault warning light on and shuts down the controller. Other fault protectionswould include an inverter fault and voltage fault routines. Many other types ofprotection may be set forth to help protect the fuel cell and any external faultissues.

DSP Filter

Most digital signal processing has some sort of filtering processing. Thisis used for two main reasoning, one being the separation of signals that havebeen combined. The second is the restoration of a signal that is been distorted.Filters can come in both analog and or digital, but the digital filtration of signals isfar superior then an analog signal.

Receive FuelCell Trip

Signal

Disable FuelCell and

switch to

backup power

Switchexternal

notification to

red

Wait forbattery to fall

at low power

point

Run shutdown of

inverter and

load

Wait for time toelapse, shut

down DSP

Controller

Figure 3.21 Block Diagram of F.C. Fault Trip Routine


18/107

Page 18

When you go to restore something, weather it is a house, car or somesignal, its something we use as a process to make some idea work again. Anexample of restoration may be an audio file, running thru a restoration processshould clean up the quality to play actual recording.

The block

diagram in figure3.21a shows thefiltration process ofa DSP system.This particularsystem is foranalog, which isstated to be themore unreliablefilter, but still getsthe job done. If you

see the block diagram, it shows the starting point to have an analog filter, whichtries to get rid of most noise issues. After the filtering process, an analog todigital conversion is completed, which has been described above. The rest ofthe process you can read above, but even in conversions, noise can be addedinto signal, therefore; another analog filter is on the output side. The startprocess is known as the anti-alias filter while the output or finish is known as thereconstruction filter. Review the half simulated graphs in figure 3.21b, whichshow the input and output of the DSP block system. This graph show thereadability of the waveform has smoothened out, looking like a functioningwaveform with no noise.

Analog

FilterA to D

DSP

Controller

D to A

Converter

Analog

Filter

Figure 3.21a Block diagram of a DSP system

Start Finish

Figure 3.21b Waveform (a) before analog filter processing and (b) after

filtration processing.

a b


19/107

Page 19

Figure 3.22a Reprinted per non-commercial permission use by Steven Smith

Filters can be a greatly help the by adding in killing out a lot of unwantednoise. Shown in figure 3.22a, we describe a filter known as Window-Sinc. This

type of filtration is use to separating multiplexed signals or noise reductions.Cost of each component and time to build PCB boards may take a little morethen doing a program. You ask, how is it possible for a circuit, such as figure3.22a to be a program? By converting it to a digitized signal, the simple circuitabove has just been obsolete. Below is a simple program that does the samething as circuit shown in figure 3.22a. This program has been configured toappraise a simple signal of 10 kHz.

100 'LOW-PASS WINDOWED-SINC FILTER110 'This program filters 5000 samples with a 101 point windowed-sinc120 'filter, resulting in 4900 samples of filtered data.

130 '140 ' 'INITIALIZE AND DEFINE THE ARRAYS USED150 DIM X[4999] 'X[ ] holds the input signal160 DIM Y[4999] 'Y[ ] holds the output signal170 DIM H[100] 'H[ ] holds the filter kernel180 '190 PI = 3.14159265200 FC = 0.1 'The cutoff frequency (0.1 of the sampling rate)210 M% = 100 'The filter kernel length220 '230 GOSUB XXXX 'Subroutine to load X[ ] with the input signal240 '250 ' 'CALCULATE THE FILTER KERNEL

260 FOR I% = 0 TO 100270 IF (I%-M%/2) = 0 THEN H[I%] = 2*PI*FC280 IF (I%-M%/2) 0 THEN H[I%] = SIN(2*PI*FC * (I%-M%/2)) / (I%-M%/2)290 H[I%] = H[I%] * (0.54 - 0.46*COS(2*PI*I%/M%) )300 NEXT I%310 '320 'FILTER THE SIGNAL BY CONVOLUTION330 FOR J% = 100 TO 4999340 Y[J%] = 0350 FOR I% = 0 TO 100


20/107

Page 20

360 Y[J%] = Y[J%] + X[J%-I%] * H[I%]370 NEXT I%380 NEXT J%390 '400 END

Reprinted per non-commercial permission use by Steven Smith

So if a simple program like the above can do the same thing as the circuitin figure 3.22a, why would we use a circuit? Cutting back on cost and havinghigher efficiency is a major portion to the competition rules. This same programcan be view in appendix C table 16-1 which shows a different efficiency setting.Other routines are supplied for use in appendix C as well.

DSP Communications with Project

In the Future Energy Challenge for 2003, requirement stated earlier in thisreport show the basic outline on communications for DSP. Just by reviewing therequirements, you can guess some of the minor functions of the DSP.

In the process of starting the review of different components, the DSP andInterface group, along with graduate students produced a basic setup. The blockfigure shown (figure 3.23) below gives a general overview on a small amount ofcommunications needed.

Sol idOx ideFue l Ce ll

Inver te r

In ter face&

Bat te ry

D S PCont ro l le r

Ou tpu t Sam p l ing( Vo l tage&C ur ren t)

Dr i ve S igna l

Dr iv ingCi rcu i t

Samp l ing

Samp l ingDr iv ingCi rcu i t

Fue l ce l l power ava i labe l s igna l

H 2 f low ra te contro l

Figure 3.23 Block diagram of DSP general controller setup

Reprinted with permission from Songquan Deng


21/107

Page 21

The interface design team dealt with the steady state portion, but to do sothey need something to control the analog signals. This is where the DSPcontroller came into action. The switching control in the DC-DC must be timed

correctly and in accordance with the duty cycle. For further explanation on theDC-DC process, see the interfaces layout and design.

Besides creating a communication on the DC-DC switching, a process onthe backup power system must be implemented. If you view figure 3.40, you willsee the interface and battery portion of the project. The signal driving from theDSP controller to the interface must communicate the accuracy of when the fuelcell being ready or not. DSP must do the following:

Create a voltage average from the fuel cells power being supplied. Comparing sample of what the VI characteristics of what the fuel cell

should be. This is stated in the fuel cell portion of the report. Sample the output voltage coming from the interface compared to thevoltage requested.

Be prepared to switch over the battery power source when fuel cell is in atransient state or not able to meet the requirements of the load.

These are just some of the communications that the controller must process,along with mathematical computations.

Communication with Interface

The DSP controller group created a protocol with the DSP unit, on thefunctions they had ready to work. After discussion and group meetings, thedesign teams decided to communicate with the interface circuit through a serialport. Being there are many different controlling ideas and future ideas, this wouldstill give room for additions. Why do we need a communication port like a serialport? The senior design academic advisor states the projects and teams need tobe separated but can incorporate for communication purposes to control andsend signals. This was good for the project described but for high level ofproduction of such a unit, putting the units together would lower the cost.

The interface current design conditions only used five of the lines. Theselines are described in the diagram. Each of the lines shown below show thatlines were producing or receiving what kind of signal.


22/107

Page 22

Pin1-Signal for switch one Pin6-Low level test pointPin3-Signal for switch twoPin5-Ground Pin9-High level test point

Using an RS-232 cable, the Fuel Cell Energy Computation suggested allof the following ideal communications. All of the communications were reviewedbut only some were added to the current design, as stated above. We includedthis for the idea of future enhancements and design productions.

DC Fuel Cell Voltage DC Fuel Cell Current DC Battery Voltage

DC Battery Current DC Link Voltage Run Inverter Faults

AC Voltage AC Current KVA Output Total from Inverter

KW Output Real Power fromInverter

Other diagnostics might include things with the LEDs on the DSP controller unit.As shown, two LEDs were placed in the bottom right of the figure 3.24.

Figure 3.24 DSP Controller Unit


23/107

Page 23

To get a bigger picture of the communications between the DSP controller unitand interface unit, see the diagram (figure3.24). shows how the TMS320C240

evaluation board is hooked up to the interface unit. This only is a brief overview,because as reviewed in the interface design, the DSP doesnt send signalsdirectly to the mosfet switches, but first going through a drive (IXDD414).

Figure 3.24a DSP & Interface Communication Setup


24/107

Page 24

DSP Fault Protections

The digital signal-processing group can supply a setup of fault protections so

nothing harmful is done to the interface and load in case of a malfunction. Suchfault protections as listed in block diagram (figure 3.25), show a fuel cell tripping.

This beginning routine shows that a fuel cell, for some reason, has a faultand trip the circuit. This sends a signal to the DSP controller for shut down of thefuel cell. To run this, the power is switch over to the battery backup power so theinverter can continue to try and supply power. If the battery falls below a presetlevel, then the controllers goes into shut down mode, first stopping the powerconsumption. After the inverter has been stopped, the controller leaves on fuelcell fault warning light on and shuts down the controller. Other fault protectionswould include an inverter fault and voltage fault routines. Many other types ofprotection may be set forth to help protect the fuel cell and any external faultissues.

Receive Fuel

Cell Trip

Signal

Disable Fuel

Cell and

switch to

backup power

Switch

external

notification to

red

Wait forbattery to fall

at low power

point

Run shutdown of

inverter and

load

Wait for time toelapse, shut

down DSP

Controller

Figure 3.25 Block Diagram of F.C. Fault Trip Routine


25/107

Page 25

Interface Design

Introduction

Interface of the fuel cell inverter project was separated from the invertergroup because no University of Central Florida topology has been created orpublished, to our knowledge, in the past. We have learned great backgrounddetails in all the sections of our overall project and did majority of the researchthrough old Future Energy Challenge reports from 2001.

The Interface group is to stabilize the power being consumed from theinverter or load (Fig.3.40). Our group has decided to use a topology that uses abattery for its auxiliary power source. The main power source being consumed isfrom the fuel cell, when a steady state has been established. Fuel cells are stillunreliable to hold a steady current flow, which is why establishing an auxiliary

power source is needed.

This auxiliary power source is to steady the voltage and current, when the slowdynamics of the fuel cell are unable to. A solution that was presented, even afterresearching different DC-DC converters, was the buck-boost converter.

For the interface module, an understanding of the characteristics of thefuel cell had to be learned, which was the portion of the power being supplied to

the interface. The main characteristics of the fuel cell are described in the fuelcell research portion of this report. To give a general overview of the fuel cell, itproduces power of a non-steady state voltage, approximately 25 volts nominal.This power supplied must be controlled to make our steady state voltage beaccomplished. The interface group was set to research different topologies tocontrol this voltage, protecting the fuel cell, inverter and the load. Reversecurrent into the fuel cell is not permitted and can cause damage if a current isleaked back in a reverse direction. Other prevented protection is the interface

FuelCell

24v

Filter

+

+

-

- Inverter

Figure 3.40 General Outlined View of the Interface Module


26/107

Page 26

module. Having stability from the fuel cell is unlikely, therefore; an interfaceuninterruptible power supply module must help maintain power stability. Thisprotects the overall inverter and its load.

Protection that the interface has control over is its ability to have an

uninterruptible supply of power. This isnt for the fear of power loss by the fuelcell, but more for the unsteady state of power supplied. Fault protection, such aspower loss by fuel cell is discussed in the DSP portion. The interface group hasstarted to set up time frames, which has seem to be the average in most pastfuel cell inverter projects. These frames give, as an idea, of how long protectionis needed to keep from losing full power. A brief outline of time frames for theback up power source to work is giving in frames of seconds. Described below isan outline of these frames, but digital signal processing group will have the finalsay and control over how protection and times are done. Our suggest for thesetime frames are listed as:

Start up of fuel cell and Inverter, supplying for 90 seconds. Short term transient state, supply for 10 seconds. Long term steady state, supply for 60 seconds.

During the start up of a fuel cell, there is a time frame of about one minute wherethe fuel cell must meet up to the power required. A current reference (Iref) fromthe DSP controller unit (see figure 3.40a) would show this to compare against theactual current (Iav) being supplied.

FUEL

CELL

Filter

Controller

Unit

+ -

-

+

+

-

-

+

Vref

Ibatt

Charge/Discharge

24 Nominal

Iav

Iref

Inverter

Figure 3.40a Controller Outline View of the Interface Module

NOM


27/107

Page 27

Along with supplying uninterruptible power from some source, an interfacemodule must re-supply to that same source, recharging for future use. Themodule must have fast response times, to make uninterruptible power possible.

Along with lines of uninterruptible power supplied, the interface controls anon-steady state voltage by using a filter. This type of filtering started the reviewof different DC-to-DC converters along with a filter. There are many differenttypes of converters that have been produced to meet specific needs.

The first converter discussed has the ability to obtain a lower outputvoltage from an average input voltage. This type converter is known as a buckconverter (fig. 3.41a) or a step down converter. This would be necessary whenthe source is supplying more power then needed. Next reviewed is a boostconverter (fig. 3.41b), which is known as a step up converter. This gives a higher

voltage supplied then the actual input.

Our research showed how both topologies would be helpful to our design.By explaining how a simple switching circuit works, we can show you how aconverter helps out the stability of our design.

If you take a switching circuit (fig. 3.41c) were it runs over a time period Tand the switch S is turning on and off, the voltage output would have somestability as shown in figure 3.41d.

+

-

Vo

+

-Vin

L

DC

Figure 3.41a Buck Converter w/ diode implementation

+

-

+

-

VoVin

L

D

C

Figure 3.41b Boost Converter w/ diode implementation

+

-

+

-

VoVin

Figure 3.41c Simple Switching Circuit

R

S

Vo=DVin

Vin

Figure 3.41d Simple Switching Circuit Waveforms

ON ONOFF

S


28/107

Page 28

If you combine both these topologies together, the basic outline gives a

buck-boost converter. An actual buck boost converter has a different topologythen that we reviewed. Our buck boost converter topology was for a differenttype of project, using two different power sources, one being a load, while theother supplies some source of power. A regular buck-boost converter topologymaintains the over and under voltage supplied by the ideal fuel cell or powersource. A basic buck-boost converter setup (fig. 3.41e) gives voltage in stages,which is less in the buck stage and more in the boost stage.

This type of topology issomething the interface group

reviewed and researched. The buck-boost converter makes up when thepower lost goes below a set tolerancelevel and above. This is a perfect typeof general topology, because thepower can go positive or negative.

The reviewed requirements set forth by the competition rules and the USDepartment of Energy (DOE) show the fuel cell will supply twenty-nine volts(nominal), with a range of twenty-two to forty-one volts DC. Along with standardson the voltage, the supplied amps will be 275 maximum from the fuel cell. Thisdoes not attempt to include or be the auxiliary power being supplied to theinverter. The standards for the auxiliary power for the inverter by the DOE arestated to be forty-eight volts (nominal) with a +10% to 20% with a nominal of500 watts per hour. The auxiliary power being supplied by the DOE at thecompetition is not for the interface module but for the extra load testing abovefive kilowatts. Being this was the challenge we begun to research for, we tried tokeep every specification possible, with our current equipment supplied by theUniversity of Central Florida.

An active filter, which is present in a DC-to-DC converter, must besupplied. For our project, we were looking into building a DC to DC for ourspecifications required by our senior design project. Use of a converter to reachthis standard nominal voltage can be bought at standard DC-to-DC conversions.This would have been the easiest known solution and less work but defeated ourpurpose to have a DC-to-DC converter built. Pre-built DC-to-DC converters(table 3.40) are priced as a reference point, both for our project budget area.

+

-

+

-

VoVin L

D

C

Figure 3.41e Buck-Boost Circuit with diode implementation


29/107

Page 29

Manufacturer Input Power Output Power PricingLambda

Electronics18 to 36 volts

24 volts nominal4.2 Ampere

$ 189.79

Table 3.40 Chart of priced solutions for DC/DC Converters

Other topologies are available for review, such as the cuk, forth order, andbi-polar output voltage converters. None of which seem to present the correctnominal voltage requirement needed for supplied power. Our currentspecifications for the interface module of the inverter are specified in a latersection.

Interface Auxiliary Power Source

The fuel cell would be supplying the main abundant power after thecompletion of the start up process has been completed. This process takesapproximately one minute or less, which we are including a time frame of aboutninety seconds.

The auxiliary power source that the interface group is using for it project isthe simplest and least expensive. The battery, yes it has the ability to hold acharge over many days and can be recharged. Being the battery would have acontinuous recharging capability and doesnt keep track of when last charged, weneed a material that has no memory. Why would a battery have a memory? Notthat we want a battery to have a memory, just certain material have this already.Therefore, by purchasing a battery that does not have a memory, when it isrecharged, it is completed to the highest level. This gives the ability to rechargethe battery all the time and never have to worry about lost of batteries ability torecharge fully. Solution to this is descried in the table 3.41. This shows that theNi-MH or Li-Ion would be better suited, how ever it is more expensive then abasic battery.

The easiest way to create a idealsolution for meeting the requirement oftwenty-four power source from the battery isby putting the battery in series of two twelvevolts as described below (figure 3.41f).

Now to create a fast response to add in extra current needed to theconverter at the time the load request such power usage is something the

Figure 3.41f Diagram of (a) seriesof two batteries verse (b) a

arallel circuit version.


30/107

Page 30

inverter interface group has completed a circuit for. On circuit switching we havenoticed a creation in not allowing the switch S1 and S2 to be off at the sametime. Other topologies use a fast turn off and slow turn on delay stage.

The power for the interface, as briefly discussed earlier, is to have the

capacity to take on a load change. This includes from the start up point to thefinal shut down of the inverter. Certain aspects must be looked at, which include:

Start up of fuel cell, supplying for 90 seconds. Short-term transient state, supplying for 10 seconds. Long term steady state, supplying for 60 seconds.

Other aspects of the interface extra power include:

Twenty-four volts (nominal)

Supply controller power before and at end of cycle

The difference in the extra power verse the main power is the source. Weknow that the main power that was going to be supplied was the solid-oxide fuelcell, which is then converted to a state of twenty-four volts. For our purpose, themain power source is going to be a DC power generator. The extra power canbe made up in different topologies or hardware.

One type of hardware or chemical reaction is the battery. A battery ismade up of many different chemicals, which make the length of power beingsupplied or the quality of the years performed. Table below show the outline of

the different chemical make up of batteries.Battery Type Abbreviation Typical Use Characteristics Advantages

Sealed LeadAcid

SLAEmergency

PowerHolds chargeup to 3 years

Inexpensive

NickelCadmium

Ni-CdAppliances,toys, most

popular

Fast, evendischarge

Inexpensive,available

Nickel MetalHydride

Ni-MH

Same as Ni-Cd plus cell

phones,portable

computers

1.2V 1200 to1500 mAh; runs2.5 to 4 hours

No memory,unusedcapacityremains

usable

Lithium Ion Li-IonSame as Ni-

MH

Stable andSafe, highest

energy capacity

Doublecharge

capacity ofNi-Cd, slowdischarge

Table 3.41 Show the different characteristics of a battery.


31/107

Page 31

Other batteries not listed in the above table (table 3.41) are Zinc-Air,Flooded Lead Acid, and Alkaline. This topic is important because the USDepartment of Energy wants to have a system that will last over a period of yearsand be low maintenance.

This one ideal solution to meet the requirements of forty-eight volts extrapower to create a steady voltage to the inverter is done by using differenttopologies. A battery of twelve volts in series, times by four, creates therequirement of forty-eight volts. This extra power source is to power the controlunit of the modules besides just creating stability in the power. The control unit isstarted up by the extra power because the fuel cell is not in a steady state toperform duties.

The characteristic for general batteries suggested for the projected are

listed in table (table 3.42) below.

Manufacturer Voltage Ampere/ Hour(Ah)

NP Series (1)NP38-12

12 V 38.0

NP Series (2)NP65-12

12 V 65.0

NPX Series (1)NPX-35

12 V 35.0

NPX Series (2)NPX-80B-HYC 12 V 80.0

Table 3.42 Suggested battery types for topology.

To meet the minimum requirements of maintaining the power needing tobe supplied, we called around looking for a battery that would carry a current forone hour of about five amps. An Intrastate battery was found that meets therequirements of twelve volts, times four, with five amp hours. The cost for thistype of battery ran around ninety-eight dollars.

Another solution to the auxiliary power is a capacitor. A capacitor is adevice that stores energy for use when a surge of power loss or unsteady state.It fast response is very useful and is used in small electronics such as memory.When used in such a small volume, it has to be updated hundredths of times in asecond to keep its current state. The larger the capacity and the chemicalvelocity used give it great power. The chemical used is called dielectric, whichhas a low and moderate make up. A low dielectric includes a dry gas of purehelium and nitrogen, along with being in a perfect vacuum. This type of chemical


32/107

Page 32

makeup gives the most effectiveness to the material but less effective of aconductor. A much higher on conductance would use chemical such asaluminum oxide. The chemical dielectric is insulated between two conductingplates, which transform the chemical into an electrostatic field.

All capacitors standard unit rating is farad (F). The unit farad is thestandard units of one second to the fourth power ampere squared per kilogramper meter squared (S^4 * A^2 * kg^-1 * m^-2).

Manufacturer Farad Volts PriceBOSS Cap 60 10 F 20 V/ 24 V Surge $ 379.99

Table 3.43 Suggested battery types for topology.

For service years and quality, capacitors are better then a regular battery.

The capacitor is safe, in that no hazardous chemical make up. The standard oflife expectancy is higher for a battery, but the cost of all this makes a major downfall. The table (table 3.43) shows an ideal capacitor that are suggested for usefor the project but cost being a major factor for the Department of Energy, we areplanning on using a battery.

Battery Safety

Just like in cars, most have a standard twelve-volt battery hooked up tothe ignition and alternator. This is so the car can be started, but why doesnt thebattery run down? An alternator is used to charge a battery up while the car is inuse, but is used as a buffer, just like in the fuel cell design. The fuel cell is therecharge, just like the alternator of the system, when it has the extra power togive out. Safety is a big issue, when charging a battery, the wires are stillhooked up the same way, but material used must be able to accept the currentdrawn back in to recharge. In our design, we just control when the battery isgetting charge or discharge but the buck-boost converter. Biggest safety is tonever cross wire a battery nor pull a battery off while a system is running. Theinverter controller and auxiliary power source come from this extra supply. If itisnt present when needed, failure of the system will take place and everything onthe load end would stop running. If you need to replace a set of batteries, placeanother set of new batteries in series, but in parallel with the older set. Attachthe new set some how and then pulling off the older. Dont change out, nothaving any auxiliary power source, could cause some major problems.


33/107

Page 33

DC-DC Topology

To explain the buck-boost converter using the topology our design groupcame up with, it has the ability to increase or decrease voltage being supplied,creating an average output voltage. This topology is known as the interface of

our project (figure 3.42), were the conversion is generated just like a DC-to-DCconverter. Certain things control a circuit like this, for a general over viewaccording to our purpose; we used these settings or setup.

Buck Mode: switch one(off) and switch two (on)

Boost Mode: switch one(on) and switch two (off)

When referring to a switch being ON, it does not mean constantly. As shown infigure 3.4, any of the switches are being turned ON and OFF, just like a pulse.

DSP

The digital signal process controller is a solution that was present throughthe process of researching different types of controllers. Another reason thiscontroller was used is because of its availability to the power electronic lab atUCF. This controller is to supply a reference to compare to average currentbeing output by the high and low side sample points.

Fuel

Cell

22-41V

S

S

LO

Batter

F

Line Out

Figure 3.42 DC-to-DC Topology Converter

ON ONOFF

S

Figure 3.42a Switch turning ON and OFF


34/107

Page 34

Another type of controller, besides DSP controller, is the hystereticcontroller we reviewed from another project. The condition of the switchingprocess was about the same except using a inductor as a reference point. Whenthe current across the inductor increases, the switching process is set at S1 onand S2 is off. This would happen if the current across the inductor falls below the

Iref- I. Then when you want the reverse operation, decrease the current acrossthe inductor, setting the S1 off and S2 to on will accomplish this. This state iswhen the current across the inductor exceeds Iref + I. Implementation of thistype of topology is shown in the figure below (figure 3.43).

S Q

R Q

Charge

DischargeIref

2

+

2

IL201

+

+

Figure 3.43 Hysteretic Circuit Controller

The points located at two different locations on figure 3.43a, low sidereference point is located at the top of the figure off the fuel cell. The otherreference point, high side, is located at the bottom middle of the figure. It gives

the average of the current available from the battery. The reference currentsupplied by the digital signal-processing group compares the samples giving bythese points and makes the appropriate changes. This type of controller setupseems to have its straightforward setup, making an ease to use.


35/107

Page 35

Figure 3.43a -The block diagram of the interface and controller units

The reference points are the controlling for the switching of the active

filter. The DSP controller has the ability to send signals to the interface drivers,to control the switching process. When switch one is on, switch two is off, theinterface is in boost mode. When in boost mode, the fuel cell is supplying thepower to the load. The load being, at this point, the external source beingpowered. If the fuel cell has enough power being supplied to the external load,then the batteries are being charged at this point, unless not needed. In buckmode, the battery is supplying power to the external load. This is because thefuel cell may not be in a steady state, may be low, or be in a start up phase.

Along with running reference signal points and other switching controls,the digital signal-processor has to be prepared to protect the system. This

includes reverse current into the fuel cell. If a current flow back into the fuel cell,the fuel cell would either damage or cause major faults. Faults?

Our research shows us that the fault protection could be very importantwhen dealing with fuel cells. Any component in the circuit could go back into afault condition causing a repeated issue or malfunction. To have the ability tofigure out where an issue is coming from, routines should be created to check thesystem status. If a fault occurs then a flag will be raised. The flag can be

Voltage controller withdelay prediction

Bi-directionalfilter

Fuel Cel l

DSP

Battery 48 V

charge

discharge

+-

Vref for battery

ibattery

Load

iav

power availablesignal

ire f

ifc


36/107

Page 36

implemented into a light system, where a green light states the status being goodor red fro when the flag is raised, showing a fault has occurred. The DSP grouphas started to implement the light system into their design, using two LEDs, bothbeing green.

Voltage and Current Flow

When the current direction is from left to right, the interface is producingextra power source from the battery to send more power to the load. The load inthis case is (figure 3.43b) the fuel cell. The actual load source we used in thetesting procedures is operated load machine. You can set the load to what everyou need, per amps and voltage. This mode of operation is buck.

When the current is flowing from right to left, the fuel cell is supplying thesource power needed. The current is going to flow thru switch two and diode atswitch one when there is extra current to be supplied. This is when the fuel cellrecharges the battery at the same time supplies the correct amount to the load.

Now, with this type of topology, the inverter will always have a nominalvoltage of twenty-four volts on average.

Calculations

Our calculations of the circuit design we started to work with come frombackground information on buck and boost converters. To give us an idea on thecomponents we needed to use, we separated the circuit into two stages, onebeing the buck and other being the boost.

Fuel

Cell-

Figure 3.43b DC-to-DC Topology Converter


37/107

Page 37

To figure out the different calculations, we first had to start off with abottom line or specifications that would meet our requirement. The followingspecifications we used to begin the process of using each of the calculations.

Input Voltage (Fuel Cell) = 24 v DC

Input Voltage (Battery) = 48 v DC Output Power = 100 watts Ripple current is 5% of minimum current Capacitor voltage ripple = 5% All other values we gathered by the calculations presented.

The calculations over the buck converter depended on the type of circuitto deal with. There are different components and layouts that can make up abuck converter. Figure 3.44a shows the circuit of a buck converter using a diode.

The calculations, as shown in the following, led us to the components that wouldhelp us get close to the correct values needed. The reason I say this, whendealing with any circuit, you have to create a level of tolerance and resistanceinside a circuit.

The duty cycle of the circuit is , which is the time in when the system isoperated. Usually the duty cycle is present in a format of a ratio or percentage.On the circuit, you want to have the lowest duty cycle as possible, this is becausethe higher a duty cycle, the shorter the life span.

Figure 3.44a Buck Converter


38/107

Page 38

The ideal characteristics of the dutycycle are show on the figure 3.44bgraph. As you can see, when the

duty cycle is risen from zero to one,the ratio of the voltage output overthe voltage input is compared.

For the average output of the current:

The current across the inductor has a minimum and maximum.

If the inductor value used falls below zero on the current, the mode would go intoa discontinuous conductance mode (dcm). Otherwise, if the current value

, then the converter operates in a continuous conductance mode (ccm).

To find the value of the dcm, we use:

The lowest or critical value found here, need to use a higher value so there can acontinuous conductance mode in the circuit.

To follow, the expressed the ratio of the output voltage by finding the change ofthe voltage across the capacitor divided by the output voltage. The formula used

to figure out the capacitor rating, can be translated from: as long asthe ratio percentage is assumed or known.

Power, just like for any circuit, is usually easily found out by using .

Translations of each of these equations from the specifications we presentedearlier, you can find out the values needed for each component.

Figure 3.44b Duty Cycle Ratio Graph


39/107

Page 39

The calculation for the following figure is for the boost converter (figure3.44c). This is just like the buck converter, excepts layout of components aredone differently. Here, the inductor is on the opposite end, just like the figuresshown earlier with the fuel cell.

So the voltage going in to the circuit is just like the fuel cell supplying the power.The load would be the battery; in this case the Vo is the battery. At this point theboost converter is charging the battery and is supplying the correct powerneeded for the load. As like buck converter, the calculations led us to thecomponents values or minimum value.

The duty cycle for the boost converter is:

If you work out the circuit, you will see that the average current at the diode is thesame as the average output current. Use the formula to figure out either way.

Since we have the current across the diode, we need to be able to figure out thecurrent across the Inductor, this is the minimum and maximum values.

To find the minimum and critical point, we can set the , therefore thevalue above, in a positive value, would keep continuous conduction mode.

Figure 3.44c Boost Converter


40/107

Page 40

The voltage ripple for the capacitor:

For each of the calculations above, an excel sheet to input the assumed values

in was done. This made it easier if we needed to re-assume our original values.The values that were most important were the resisters, inductor, and capacitorsthat we got from the calculations. These values the calculations came up withdidnt mean the value was available, therefore we had to assume some values inthe final analyst.

The values that were produced from the calculations are:

Buck Converter Specifications

Duty Cycle 50% Resister 5.76*Fuel Cell (In) 24v Inductor 47.6uH

Battery (In) 48v Capacitor 0.29uF

Output Power 100w

Capacitor Ripple 5%

Boost Converter Specifications

Duty Cycle 50% Resister 23.1Fuel Cell (In) 24v Inductor 48uH

*Battery (In) 48v Capacitor 14.5uF

Output Power 100w

Capacitor Ripple 5%

In the tables above there is a * next to the description of the load. The load beingthe output at that particular mode, either buck or boost. To get the calculationscorrect, you must make sure that the correct value is in the right locations, otherwise the duty cycle will be incorrect.

Another calculation is the voltagedivider. This calculation is to find theresistance ratio needed for resistorvalues. Suck as, when you have thehigh side (battery) being tested, youhave a Vin of 48 volts. The outputdepends on the expected, which forDSP is between zero and five volts. Inthis case, 48 volts is nominal, so 2.5volts output would be reasonable.Figure 3.44d shows a voltage dividerdiagram, Vout not being a load, but thetest point.

R1

R2Vout

NOLOAD

Vout = V1* IR2I(R1+R2)

Figure 3.44d Voltage Divider


41/107

Page 41

Schematics of Design

As shown in figure 3.45a, the buck converter is at CCM critical point. Perspecifications, we changed the values to meet our wanted output. By using

Pspice capture simulator,the buck converter modecould be simulated,showing the output. Forthe output, using all ourspecifications, we werelooking to get results of anominal voltage amount.The final result for thecircuit design was just asthe assumed results

should have been.

After running the simulations of the buck converter, the design of the boost(figure 3.45b) was almost completed. Changes made were the location of theload and powersource, along with thefunction generator.During the simulation,the function generatorfor a boost mode isconnected to switchtwo. This whileswitch one isgrounded, off, oropen. This designcreated the correctoutput to the resistoron the left, which is acting as the load or battery. The output voltage, per all thespecifications is approximately forty-eight volts.

After confirming all the simulations and correct values for each of thecomponents, the design was calling for a control to have the correct switchingprocess. Along with the other components, we must pick a switch to meet therequirements. Using data sheet and references on mosfets and IGBTs, wedecided to go with a mosfet switch.

Figure 3.45a Buck Mode (Pspice)

Fi ure 3.45b Boost Mode Ps ice


42/107

Page 42

Components

The first step before choosing the correct component to use in our designwas the calculating the right or approximate values to be used. The calculationspresented earlier, which they go over the minor components descriptions (see

calculations).

The first component that has been briefly described earlier is switch, whichwe have done some research on. The biggest thing reviewed for this typeproject is the switch time. Because it is so important to have a fast switchingprocess, so the converter canbe uninterruptible powersupply. For this design, theIRFP 140 mosfet switch waschosen to meet our needs. Ifyou look at high switching

process and low resistance onthe gate, this helps inneedless signal lost.

The component that we chose tocontrol the mosfet was an IC controller.This controller is used just for thepurpose of sending a high and lowsignal. The high signal switching oneand low signal switching the other,switch two mosfet. The IR2110controller was picked for our design forits:

High speed, high voltage drivers for mosfets Float Channel, operational to +500v Signal times delay (includes rise/fall) 120 on, 94 ns off (typ.)

G

D

S

Figure 3.45 c) (top) shows the size of the mosfets

compared to a quarter for scaling. d) (left side) diagram

showing the view of the gate, source, and drain. f) (bottom)

shows the size of the IR2110 controller, along with the seat

for the chip.


43/107

Page 43

The data sheet, given the connections needed to complete the hook up of the

IR2110, described the component. We have included this data description(figure 3.45f) because of the importance of the correct layout to our project.

IR2110 Controller Specifications

Lo Low side gate driveCom Low side returnVcc Low side supplyVs High side floating supply returnVB High side floating supplyHo High side gate drive output

VDD Logic SupplyHIN Logic input for HoSD Logic input for shut downLIN Logic low input for LoVss Logic Ground

So DSP controller cancommunicate to the interface boardcorrectly an additional component had tobe added. The IR2110 requires a highervoltage then DSP can handle, same withthe mosfets, so we used a driver. Thisdrive, IXDD414, is to produce a higher

output voltage from the low input voltagereceived from DSP controller signals.

Figure 3.45f IR2110 Controller Setup

Figure 3.45g IXDD414 Driver


44/107

Page 44

The IXDD414 is a driver component, used for driving mosfets and IGBTs,along with DC-to-DC converters. The specifications for the driver are:

IXDD414 Driver SpecificationsVcc Logic PowerIN Signal Line InEN Enable line (High if enable)

GND - GroundOut Signal Out, boosted voltage

The importance to this component is the actual hook up, where it is easybut for the drive to begin working, the enable pin must be enabled with somesource of power/ signal. This signal, when broken to the EN pin, will stop thedriver from continue its work. This could be something to review for short circuitor fault conditions in the future. For now, we set this pin to a constant on, so the

driver is able to communicate. For our design, just like the IR2110, we used aneight-pin seat, therefore if anything should happen to the chip, it can be easilyreplaced.

So to better understand the translation between all the different controllersto the mosfet switches, figure 3.45h. This translation between the controllersignal voltages is just an example. Some of the controllers could handle more, orpossibly have to handle less.

14

13

12

11

10

9

8

Vss

LIN

SD

HIN

VDD

Lo

COM

VCC

Vs

VB

HO 7

6

5

4

3

2

1

IR2110In: +10v Out: 12v

In: 5v Out:11.9vG

D

S

In: 12v (up to 20 allowed)IRFP 140

IXDD414

Figure 3.45h Conversion Across Controllers


45/107

Page 45

Layout Design

To begin the process of getting ready to mill a PCB board, we first had tocomplete a layout design of the full schematic of all the components. The valuesof the components do not have to be correct, just the way they are connected.

Polarity issues could cause a major issue after milling a board if the componentsarent connected correctly. For most of the components we used, Pspice carriedthe correct corresponding one. The controller and drivers, they had to be relayedout and placed into a library because Pspice didnt have the pint correct. Thisprocess did not take long after the first one was completed. Our schematic withall its components first looked like figure 3.45i.

After completing the schematic, when then transferred the design over toPspice Layout Plus program, here is were the layout (figure 3.45j) of the PCB isdone. Per specifications we still wanted to make our design as small andcompact as possible, this is so the future production of such board can be doneat a low production cost. So a four by four inch piece of copper PCB board was

Figure 3.45i Schematic of Final Design


46/107

Page 46

cut and milled according to out lay out design, which gave us the final product(figure 3.45k).

Top Design Bottom Design

Figure 3.45k Final Design of PCB Board

Figure 3.45j Final Layout Design from Pspice Layout Plus


47/107

Page 47

Serial Connection

To connect and send signal between the different components, such as

the DSP controller and the interface controller, we had to think of some sort ofcommunication line. The interface team and DSP team decided that thecommunication between the two could be done with a single serial cable. A RS-232 cable holds up to nine different lines.

The interface current design conditions only used five of the lines. Theselines are described in the diagram below.

Pin1-Signal for switch one Pin6-Low level test pointPin3-Signal for switch twoPin5-Ground Pin9-High level test point

Design Results

The final simulations of the interface design show the results the interfacecircuit. To give a brief description, when the circuit was in buck mode, the battery

Fi ure 3.46a Buck Mode final results In: 48 volts


48/107

Page 48

side (high side) was giving power to the load. The battery is a power supply ofDC voltage, supplying the desired test. The load being simulated load of acertain amount of current being drawn.

The first snap shot of the oscilloscope shows the interface in buck mode

(figure 3.46a). As seen, the output voltage is approximately twenty-three volts,while the input voltage is 48 volts.

The next oscilloscope screen shot shows the circuit in boost mode (figure3.46b). The input value is being 24 volts, while the output is approximately 48volts. This shows exactly what the interface board is supposed to do. At thispoint, the load is the battery while the source is from the fuel cell.

Figure 3.46a Boost Mode (final results) In: 24 volts


49/107

Page 49

Inverter

Introduction

Static inverters, or DC-to-AC inverters, employ fixed DC sources to output

symmetrical AC voltages which may be of fixed or variable magnitude orfrequency. These output voltages can be single-phase or three-phase accordingto design. Inverters provide power from the DC source to an AC load, which maybe passive or active, through the use of SCRs or gate driven semiconductors.Such semiconductors include GTOs, IGBTs, and MOSFETs. Because ofincreased power and switching speeds, along with modern control technology,inverters can now be used in a broad range of output voltage and frequency,while harmonic distortion is reduced. DC-to-AC inverters are useful when a loadrequires AC voltage and the only power source available, or practical for theapplication, is a fixed DC source. Examples of such systems includeuninterruptible power supplies (UPS), power supplies in aircraft, and motor-

drives.

L

OAD

voltage-source

Inverter circuit

Voltage

control

Dc-Ac conversion

output

isolation

transformer post filter

-

ac voltage

referance

Driving

Circuit

Comparator

Ac output

T

c

+

o

DC

Fig. 3.31 Block Diagram of Typical

Power Electronic Circuit With DC-

AC Inverter (Reprinted permission

of Issa Batarseh)


50/107

Page 50

Figure 3.31 is an example of a typical DC-AC inverter circuit as used inpower electronic applications. This diagram shows the voltage source, inverter,transformer, output filter, and feedback circuit. The output filter is added in orderto smooth the AC voltage before it reaches the load. The feedback circuit is usedto compare the AC output voltage with a reference signal. This is done to be

sure that the inverter is producing an acceptable output.

The types of switches used in the inverter and their arrangement are adetermining factor in the control method which will be used. The goal of thecontrol method is to convert the DC source to a controllable AC output. While itmay be desired that the output be a purely sinusoidal waveform, the inversionprocess will also add high frequency harmonics. Because of this, high frequencyswitching is used. This reduces, but does not completely remove, theseharmonics.

Two common inverter types are the half-bride and full-bridge. These two

inverters are shown respectively in Figure 3.32.

Vdc

S1

S2

o

+

-

+

-

Vdc

S1 S2

o

+

-

+

-

S3S4

(a )

(b )

Fig. 3.32 Single-phase inverterarrangements,

(a) Half-bridge inverter(b) Full-bridge inverter.

(Reprinted with permission of Issa Batarseh)


51/107

Page 51

Half-Bridge Inverter Circuit

Shown in Figure 3.33 is a half-bridge inverter with an inductive-resistive load, as

well as its equivalent circuit and output waveform.

(a )

S1

S2

R

Vdc

iO

o

L

a

a '

Vd c

+-

t

in o

R

L

iL

+

-

in

+ Vd c

-Vd c

T /2 T0

t

iL

IL

(T /2 )

IL

(0 )D 1 D 2

Q 1Q 2

tT /2 T0

o

t1

(b )

(c )

F ig . 3 .33 (a) Ha l f -br idge w i th induc t ive - re s is t i ve load .(b) Equiva len t c i rcu i t and (c ) S teady s ta te wave form s .

( R e p r i n te d w i th p e r m i s s io n o f I s s a B a t a r s e h )


52/107

Page 52

Switches S1 and S2 operate complementarily at a 50% duty cycle with aswitching frequency off. This provides the load with a square voltage waveform

)(tvin and an amplitude of dcV . Therefore:


53/107

Page 53

From Fig. 3.33 (c) it can be seen that:

( )

( )Tii

Tii

LL

LL

=

=

)0(

2)0(

When 2/0 Tt is givenby:

( )2( ) ( (0) )

t Tdc dc

L L

V Vi t I e

R R

=

This equation is equal to ( ) ( (0) )t

dc dc

L L

V Vi t I e

R R

= + , where

R

L= , when TtT


54/107

Page 54

This value of t1 allows the average and rms values in the switches and diodes to

be found. Figure 3.35 will help with these calculations.

iL1

iL

T

T/2t1

Io1

t

t

+Vdc

-Vdc

o1

o

1oV

Fig. 3.35 (b) Shows the average transistor

and diode current waveforms

(Reprinted with permission of Iss

Final Report Design 1

Documents