Custom Maximum Power Point Tracker - Solar Gators Project
Purpose
In solar car racing, Maximum Power Point Trackers (MPPTs) are vital for optimizing energy harvesting and use. They ensure solar panels operate at peak efficiency despite changing conditions, convert variable voltage to stable output, and can increase overall efficiency by up to 30%. MPPTs continuously adapt to race conditions, managing power generation and battery charging to support race strategy. In competitive events where every watt matters, effective MPPT systems can be the difference between winning and losing, prompting teams to develop custom solutions that balance efficiency, weight, and integration with their specific vehicle design.
Tldr; Solar Cars use MPPT’s as a safe way of utilizing the power coming out of the solar panel. The companies that make them for University Solar Car teams are starting to die out (There was 6 companies, now there are 2) and getting really expensive. So we need to make them ourselves.
Project Timeline
March 2024 - June 2024
In this phase I started digging through a large amount of the relevant theory with regards to MPPT’s, especially since this project has been floating Solar Car for years with no real movement towards a physical products. My first task was to understand how Buck and Boost Converters so I began to eat through random videos from GreatScott Electronics, EEVBlog and random instructables articles that were online in to really understand the problem itself.
This process of understanding the theory had to managed with studying for the later half of my Spring Semester, where I was taking Complex Analysis and Electronic Circuits. Both of these classes ended up being quite useful as I later realized. After spending some weeks reading into the theory, I ended up making some simulation circuits on LTSpice to model Buck Converters, Boost Converters and IV Curve Tracing in order to gain a better understanding of the theory behind it.
Reviewing various parts of theory including but not limited while working full-time at CACI and taking a class online.
Power electronics fundamentals
Semiconductor devices (MOSFETs, diodes)
Magnetic component design (inductors, transformers)
Control systems and feedback loops
PCB design and layout techniques
Thermal management
EMI/EMC considerations
Microcontroller/Embedded programming
Battery charging principles
Solar panel characteristics with the IV Curve
June 2024 to Now
Figure 2: Comparison between Asynchronous and Synchronous Buck Converter topologies
Currently I am in the testing phase for this device right now with various objectives being focused on and achieved at the same time. The configuration that I am using is a Synchronous Boost Converter system with the panel output connected with a Voltage Sensor and Current Sensor system. The Sensors would straight pipe into an ADC to an ESP-32 MCU in order to run the Perturb and Observe Algorithm to then be able to switch ON and OFF for the Boost Converter.
Basic Operating Principles
Typical Boost Converters are considered Asynchronous, meaning that the timing for the duty cycle is not periodic with respect to the power inputs.
Consider that a normal converter topology for either buck or boost, contains a diode that has a very high turn-on resistance, which is an issue that would lead to inefficiencies due to heat loss and imprecise timing of turning ON and OFF the system.
Refer to the image on the right for Asynchronous/Synchronous Buck Converters
Please be aware that the image on the right shows the same relevant principles as our Boost Converter, but I spent most of the summer working on a Buck Converter design. Starting in the Fall after a design review with my Solar Gators alumni, a design change has been made to switch to Boost Converters instead. Most of the circuits that will be shown here on this page will be BUCK -Converters, but the final goal is to have a BOOST converter.
Prototype #1 Review
On the right we have a prototype of a synchronous buck converter from earlier this summer. The 3.3V, 5V, 12V power lines for this section are not shown as the device pipe directly into the relevant power lines for the Solar Car. This has a benefit of better DC Isolation and reduces design complexity as I do not have to recreate the wheel when I can just go to auto shop and get a pair of tires for less work.
The current issues with this design are as follows
ESP-32 VS STM-32: The original design was based on using ESP-32s, which is best for the optimized switching ratio with regards to the switching that would occur within Buck Converter systems. Needs to be switched to an STM-32 in order to have a coherent refactoring in order to integrate into Solar Car’s basic command functions.
No CAN: The code for the perturb and observe algorithm is not configured to work with the massive code repository, more specifically the CAN communication protocol. This is a massive issue as the main control protocol for the car is based in CAN, a communication protocol that is standard practice for both Solar Cars and modern electric/gas cars as well.
Prone to shorting/broken components: In portions of LTSpice testing, the voltage ratings for the MOSFETs are way too weak to account for the occasional solar irradiance spikes that can occur when a cloud clears thus can be way too high and break the fuses that would enter the circuit. Luckily the fuses were there to prevent any issues. Electronic fuses need to be considered down the line.
After reviewing my design with Lucas Magno from Gator Motorsports at University of Florida, he suggested I look through the INST rating scheme from NASA in order to “overengineer“ the voltage conditions that would subsequently form.
Prototype #2.1A/2.1B
Here are the current areas of focus right now for testing the MPPTs, each being designated as a path of development based on the issues with Version 1.
CAN Communication (Version 2.1A): Be able to configure the CAN Communication protocol of the Solar Car to be able to tell the MOSFET driver present in the circuit to be able to turn ON and OFF and do a NUKE-Shutdown
Higher Voltage Rating (Version 2.1B): Redesign the board in order to prevent the voltage spikes from breaking the fuses. This would most likely include thicker routing traces, the NASA re-rating for components and the incorporations for eletronic fuses.
Overall, switch the system to be a boost converter design as well as a general note.
Figure 1: Current Configuration of the ELMAR MPPT’s used in the Car 3, Sunrider (Dated September 19th, 2024)
V1 Buck Converter MPPT
Current Testing Procedures For MPPT Testing
In the version 1 of the MPPT, the original testing procedure is just randomly modulating the power supply in order to simulate the randomness of a panel. After some reading of the literature online and some broken LT-Spice Simulations, there has been an improved testing infrastructure for testing the MPPT.
MATLAB Simulations: Using reference videos from YouTube, one can actually simulate the IV Characteristics and then simulate the actual timing values for Switching on an MPPT in order to create
Programmable Power Supply: The RIGOL family of products allow for programmable output voltage and current forms similar to that of a solar panel.
Final Product testing will be done with connecting the actual MPPT to the Solar Car and having the data from it show up on the Solar Gators telemetry website.