Project PlanRev. 0.1Page 1
Project Bluebird
University of Portland / School of EngineeringPhone 503 943 73145000 N. Willamette Blvd.Fax 503 943 7316
Portland, OR97203-5798
Final Report
Project Molalla: Microprocessor-Based Charge Controller
Contributors:
Andrew Melton
Antoinette Realica
John Turner
Approvals
Name / Date / Name / DateDr. Lu / Dr. Lillevik
Insert checkmark (√) next to name when approved.
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Revision History
Rev. / Date / Author / Reason for Changes0.9 / 04/12/2006 / A. Melton / Initial draft
0.95 / 04/18/2006 / A. Melton / Suggestions from advisor
1.0 / 04/19/2006 / A. Melton / Suggestions from approval mtg
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Acknowledgements
We would like to express our gratitude to:
-Dr. Lu, our Faculty Advisor, and Mr. Hui, our Industry Representative, for the guidance and support they have given us through this project.
-Ms. Sandy Ressel, our Electronics Technician, for providing us with parts and soldering services
-Mr. Gary Carlson for helping us formulate the idea for the project
-The University of Portland Multnomah School of Engineering for funding the project
-The Electrical Engineering faculty for lending to us their wisdom through the advice and insight they gave us
-Family and friends for the encouragement and patience they showed us during the Senior Design process
Table of Contents
Summary
Introduction
Background
Methodology
Results
Technical
Intermediate Battery
Vehicle Battery Pack
PIC Microprocessor
Voltage Sensors
Boost Converter
Relay
Process
Assumptions
Milestones
Risks
Resources
Contingencies
Change Control Board (CCB)
Final Outcome
Conclusions
Appendices
Appendix A
Appendix B
Appendix C
Appendix D
List of Figures
Figure 1 Molalla Product
Figure 2 Molalla System Block Diagram
Figure 3 Boost converter and LM3488
Figure 4 Molalla Schedule
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Summary1
Project Molalla is a microprocessor based charge controller that will supplement the pre-existing charging power control unit of a hybrid vehicle with an additional solar charging source. In designing our project we followed standard engineering procedures in a top down design method with bottom up implementation. We specified major inputs and outputs, divided one main block into several small functional units, then researched and designed each block.
There are two key components that comprise Project Molalla: the boost converter and the microprocessor. The boost converter steps the voltage from the intermediate battery (24V) to the desired level (150V) to charge the vehicle battery. The microprocessor monitors whether or not the vehicle battery needs charging and selects an available charging source, where the pre-existing power control unit is the default source.
After assessing our resources, we went forth with the design that was the best solution given our budget and time constraints. Other boosting methods would have required transformers or specific inductors, the production of which wouldneedspecial design and manufacturing. This would have called for more time and money than we were available to us.
At the time of our prototype release, we were able to demonstrate a successful product. Our boost converter block was capable of boosting from 24V to 150V and has successfully charged batteries with 0.5A constant current, and our microprocessor block was able to monitor input voltages and control relays accordingly. In encasing our product, we encountered problems that we did not predict or expect and remedied them as best we could. Thus incorporating the product into a single case would lend itself well to future development.
The remainder of this document will further introduce our project, give greater detail about our design and implementation methodology, and finally explain the results of our work.
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Introduction2
This document describes how product Molalla, the microprocessor-based charge controller, was constructed and the end results of the product. The methodology section provides a description of design and build process. The results detail the performance of the product, and the conclusions include future developments and improvement suggestions to the product.
The rest of the document is laid out as followed:
- Methodology
- Results
- Technical
- Process
- Conclusions
- Appendices
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Background3
As prices for crude oil continue to rise to historic levels, it is evident that reliance on fossil fuels will be an unaffordable peril to future generations. Scarcity and cost of these fuels combined with their emission of greenhouse gases make a powerful case for the development of non-fossil fuel based methods of transportation. Project Molalla seeks to increase the technology base for alternative energy based cars by providing an upgradeable power mediator that will increase fuel efficiency.
There exist several categories and schemes of battery chargers. In general, our scheme falls under “opportunity charging” as it incorporates power whenever it is available. This is an apt description for an onboard charger in a hybrid vehicle as power sources (such as regenerative braking, alternators, and solar panels) are not constant. Project Molalla must choose from available power sources. The first available source is the pre-existing power control system in the hybrid that mediates charging between regenerative braking and a gasoline powered generator. The second source will be an intermediate battery charged by a solar panel (see Figure 1). Project Molalla will act as a smart switch between these sources, allowing the intermediate battery to charge the vehicle batteries when the existing power unit is not available.
Figure 1 MolallaProduct
It is important to note that actual sources and loads such as working car alternators, regenerative braking systems, multiple car batteries, and solar panels are costly and difficult to obtain. This introduces a possible limitation on the ability to test the device, requiring accurate source and load modeling.
A final limitation on Project Molalla’s design is in its ruggedness. Adequate research and implementation of a rugged device capable of extreme operation over hundreds of thousands of miles and several years is a project in itself. For this reason, the product will be theoretical rather than actually mounted inside of a vehicle.
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Methodology4
In designing our overall product, we began with a hierarchical approach. Our goal was top down design and bottom up implementation. We defined how we wanted our project to function, and then divided it into several functional blocks. Each block was then independently designed and finally integrated. Before doing any design however, we had to perform significant research to resolve various problems unique to our project.
Before performing any design work, we also made assumptions, assessed various risks, created contingency plans, and created a detailed schedule with important milestones. In addition, we listed all resources that we would need, assigned necessary tasks to specific group members, and created a budget that encompassed the entire project.
After design began, our work was organized as follows:
-Drew schematics
-Selectedand ordered parts
-Building
- Boost converter – first on bread board, then designed PCB
- Programmed PIC using MPLAB ICD2
-Unit testing
- PIC monitoring – completely debugged PIC control and relay switching separate from boost circuit
- Operated boost converter and analyzed various waveforms. Replaced key parts to achieve stability.
-Integration
- Made all appropriate connections to use PIC to monitor and control the boost converter and relays.
- Encased finished product.
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Results5
Technical
As shown in Figure 2, the Molalla Product consists of a PIC microprocessor, a boost converter, voltage sensors, and relays. The PIC microprocessor monitors the output of the pre-existing power unit of the hybrid vehicle and the intermediate battery through the voltage sensors. When there is no output from the pre-existing power unit, the PIC checks if there is enough charge in the intermediate battery to charge the vehicle battery, and if there is a sufficient amount of charge, it switches the relays so that the intermediate battery is charging the vehicle battery. The intermediate battery is able to provide the necessary voltage/current needed for charging through the boost converter. Whenever output appears from the pre-existing power unit, the PIC switches the relays so that the pre-existing power unit is charging the vehicle battery (default mode). The overall Molalla schematic can be found in Appendix A.
Figure 2 Molalla System Block Diagram
Intermediate Battery
- Modeled as two 12V Yuasa lead acid batteries
Vehicle Battery Pack
- Modeled as eleven 12V Yuasa lead acid batteries
PICMicroprocessor
-PIC18F452
-Enables/disables boost converter based on the status of the voltage sensors.
-Contains an A/D Converter used to verify charge from the pre-existing power unit and from the intermediate battery
-Output pins send out a logic 1 (approximately 5V) or 0 (0V)
- Controls relays.
- MOSFETS for voltage sampling
-PIC Program Flow Diagram – see Appendix C
Voltage Sensors
-Sense the battery voltages and provide inputs to the PIC as feedback.
-Consists of two resistors that provide a voltage divider scheme
-Chosen to output a small current (μA range) for small i2R loss
- For intermediate battery: 43.5 kΩ and 5.1kΩ resistor combination
- For vehicle battery: 1.5 MΩ and 27 kΩ resistor combination
Boost Converter
-Steps the intermediate battery voltage from 24V to 150V and charges the vehicle battery pack with a constant 0.5A of current.
-Consists of n-channel MOSFET, N-Channel IC Controller (LM3488) which maintains constant current for battery charging, an inductor, capacitors, and a diode rectifier as shown in Figure 3. Refer to Appendix A and Appendix Bfor actual part values.
Figure 3 Boost converter and LM3488
Relay
-Acts like a DC transfer switch, switching between two power sources: the intermediate battery and the pre-existing power unit of the hybrid vehicle
- Consists of one normally open relay connecting the intermediate battery to the vehicle battery and one normally closed relayconnecting the pre-existing power unit of the hybrid vehicle to the vehicle battery
- Magnecraft 199BX-11 Power Relay
- CrydomSolidState Relay D2D07
- In case the product fails, the normal operation of the hybrid vehicle is not hindered (connection between pre-existing power unit of the hybrid vehicle and the vehicle battery remains intact)
-Controlled by the PIC.
Process
Assumptions
In this project we assumed that:
- Batteries can provide at least 12.5 amps continuous current.
- Parts needed are available and affordable.
Our assumptions were adequate. We were able to draw enough current from the batteries and found and acquired the parts needed.
Milestones
Our milestones were:
Number / Description / Original10/28/05 / Previous
03/02/06 / Present
04/13/06
1 / Product Approval / 10/07/05 / 10/07/05 / 10/05/05
2 / Plan Approval / 11/11/05 / 11/11/05 / 11/11/05
3 / Design Release / 12/09/05 / 12/09/05 / 12/09/05
4 / Boost Converter Built (added 12/08) / N/A / 12/15/05 / 12/15/05
5 / Parts Received / 01/27/06 / 03/24/06 / 03/24/06
6 / TOP’s Approval / 02/24/06 / 02/24/06 / 02/24/06
7 / PIC Program 0.95 / 03/03/06 / 03/30/05 / 03/30/06
8 / Prototype Release / 04/07/06 / 04/03/06 / 04/05/06
9 / Founder’s Day / 04/11/06 / 04/11/06 / 04/11/06
10 / Final Report / 04/13/06 / 04/13/06 / 04/13/06
Dropped Milestones
Boost Converter Stage 1 Built / 12/15/05 / N/A / N/A
Voltage and Current Sensor Built / 02/10/06 / N/A / N/A
Boost Converter Stage 2 Built / 03/03/05 / N/A / N/A
We were able to meet most of our milestones on time. Some we accomplished before the expected deadline. A few we had to postpone, but we adjusted the schedule accordingly.
Risks
We assessed and experienced the following risks:
- Boost converter failure – high risk.
- Since the boost converter is an important component to Project Molalla. It provides the core functionality of the product by creating the boosted voltage signal used to charge the vehicle batteries. Hence its failure will affect our project severely.
- We experienced:
- Stability issues – bread board vs. PCB
- Extensive debugging
- High voltages and current resulting in high parts fatality. We had to reorder parts and use overnight shipping on several occasions.
- Part availability – medium risk.
- Special parts required for the completion of the product may not be available to the typical consumer. This will restrict the choices available when designing the product.
- We experienced:
- Difficulty in finding DC relay to fit the task.
- Difficulty in finding the proper gate driver to drive the MOSFET that would provide the proper on-voltage
- Late arrival of parts – medium risk.
- Delays the progress of Project Molalla.
Resources
These resources were available to us:
- Personnel
- Andrew Melton. Team leader for Spring 2006. Design, implement, and test PIC program.
- Antoinette Realica. Recording Secretary. Design, implement, and test relays and voltage sensing portion. Webmaster.
- John Turner. Team leader for Fall 2005. Design, implement, and test boost converter.
- Dr. Wayne Lu: Project advisor.
- Mr. Andrew Hui: Industry representative.
- Sandy Ressel: Electronics technician
- Mr. Gary Carlson. Independent contributor
- Equipment
- Oscilloscope
- Bread board
- DC Power Supply
- Ammeter
- Voltmeter
- Facilities
- EH 2007 (Electrical Circuits Laboratory)
- EGR 312 (Microprocessor Lab)
- EH 2001 (Senior Design Laboratory)
- Time – see Appendix D for schedule
We created a schedule, met weekly, and prepared a budget to ensure successful completion of our project. However, we spent more time and money on the project than originally predicted due to unexpected problems that we encountered. Debugging the boost converter and the PIC code took more time than expected. As for our budget, while we were able to incur free samples from companies, our high parts fatality caused us to reorder parts with overnight shipping to keep on schedule.
Contingencies
Our contingency plans were:
- Boost Converter Failure:
- Increase intermediate battery voltage
- Purchase off the shelf 12V DC to 220V AC inverter, then rectify and buck to 150V
- Use PIC for PWM in a cascaded boost converter topology
- Part availability
- Find an alternate part or redesign to match parts available
- Late arrival of parts
- Be patient
- Order from another supplier
Our final product used our main plan, thus no contingencies were fully developed. However, initial testing of the IC based boost converter gave mixed results so we began developingour contingency plans. We were able to successfully use the PIC to provide PWM and control in a single stage boost converter and built a two stage boost converter on a PCB. Unfortunately, we were not able to successfully debug the circuit due to time constraints. We also purchased an off the shelf 12V DC to 220V AC inverter and other required parts for our third contingency. We did not, however, afford time to developing this contingency because we felt that our time would be spread too thin. Fortunately, continued debugging efforts on the IC based boost proved successful.
Change Control Board (CCB)
We performed a CCB request on December 6, 2005 to use an IC instead of the PIC to provide PWM and control for the boost converter. We had originally planned to use the PIC for the PWM signal in a dual stage, cascaded boost but later found an IC that simplified the converter significantly by requiring only one stage and eliminating the need for the PIC control. In addition, after talking to professionals in industry, we were informed that it was somewhat unconventional to take a cascaded boost approach. The impact of this CCB was that the role of the PIC in providing PWM and providing current sensing for stability were eliminated since the IC would be able to provide these, but the PIC was still necessary for voltage sensing/source selection.
Final Outcome
We succeeded in producing all of the individual functions required for a microprocessor based charge controller. At the time of our prototype release, we had a complete, functional system.This included a successful boost circuit and monitoring circuit that was controlled by the microprocessor. Upon integrating the entire product into a single enclosure, we encountered complex problems as a result of mixing high and low power systems. We narrowed them down to ground bouncing, EMI, and a sudden voltage change on the intermediate battery due to large current draw. These problems are worth exploring and would be great for future development of the product.
University of PortlandSchool of EngineeringContact: A. melton
Final ReportRev. 1.0Page 1
Project molallaup-ee-tr-06-06
Chapter / Conclusions6
Project Molalla is a step towards enhancing the current technology for alternative energy based cars by including solar energy to the pre-existing charging scheme. With the rising costs of fuel, it is our hope that the implementation of harnessing solar energy to power hybrid vehicles will increase fuel efficiency in a vehicle.
Our boost converter is functional and is based on the LM3488 controller from National Semiconductor. It outputs a constant 0.5A of current to charge the vehicle battery pack in a single stage boost from 24V to 150V. The microprocessor can enable and disable the two relays based on the input and output voltages read from the voltage sensing blocks.
We were successful in implementing the individual functional blocks of our design and in integrating them into a final product outside of an enclosure. Full integration of the productinto an enclosure was not achieved at the time of this document. Should further development of the Molalla product be undertaken, we would like to recommend analyzing EMI issues, observing ground bounce issues, and finally realizing full integration into an enclosure. Should more time and money become available, we would recommend a fly back or forward converter approach instead of using a boost converter.