For Solar Power and Batteries
Solar power and battery technologies play an important role in renewable energy resources, whereas the PhotoPod can be a useful device for renewable backup.
Although electric vehicles (EVs) are an environmental upgrade over the widespread internal combustion engine, some outlying concerns and drawbacks still limit the potential of EVs. The production and eventual disposal of EV batteries is an increasing problem for both the lithium mining industry and recycling plants. With EVs’ huge increase in popularity over the past decade and an approximate lifespan of 10-20 years for EV batteries , many of the batteries are now being decommissioned and frequently end up in landfills. This is their fate, even though these batteries still retain the majority of their storage capacity, which is completely suitable for certain uses, but perhaps not ideal for an EV (due to decreased driving range). We set out to design a product that would enable a longer lifespan for these batteries, through their repurposing and reuse.
Our device, the PhotoPod, is a solar micro-converter, capable of charging an end-of-life EV battery directly from a solar panel, for recreational use or residential backup power. This device converts 12VDC to 24VDC from a solar panel to ~400VDC, aimed at charging a Generation 1 Nissan Leaf battery pack. We built this project anticipating a future increase in electric vehicle use, and the number of used EV batteries that will result from this increase. This article covers the back story of this idea, how we built the product, the challenges faced, and the next steps for our concept.
To design and build this device effectively, we set out with a focus on safety and efficiency. This meant our device had to include maximum power point tracking (MPPT) to extract the most power possible from the solar panel, and a DC to DC converter to reach the required output in a single stage. We accomplished this with the incremental conductance algorithm tracking maximum power, and a flyback converter for the voltage conversion. We programmed the algorithm on the Arduino Uno using C code, which also generated the high-frequency switching signal (PWM) for flyback operation. We designed the PhotoPod to be rated for marine use; accordingly, the connectors and enclosure are all rated for IP67 (waterproof rating), corrosion-resistant, and durable. The internal display, which is visible through the enclosure’s clear lid, shows the input voltage, current, internal temperature, and duty cycle of the PWM signal.
The most critical portion of our product development process was designing the flyback converter. Having the correct output was key for safety, battery protection, and getting a decent transfer efficiency. Although the solar output is DC, the converter uses transformers in the converter to obtain high output voltages, and finding a transformer with the correct specifications for our design proved challenging. Ultimately, we had to design, build, and test our transformers from scratch, a daunting but exciting endeavor.
During our initial weeks of planning, we came up with an approximate idea of how the PhotoPod would work, choosing a converter topology, controller, and the project requirements and features. We calculated the component values for the converter and set up our first prototype (Figure 1). For this iteration, we used an off-the-shelf transformer (Wurth Electronic 750311771) that had similar characteristics to what we required, but it eventually led us to a dead end. Despite setbacks, we used these situations to continually refine our design and improve the PhotoPod prototype.
As we developed our prototype and worked through a wide variety of problems, we optimized our schematic design (Figure 2). For input current readings, we attempted a variety of measurement methods, including a Hall effect sensor, but eventually achieved the highest accuracy with a 40mΩ shunt resistor, boosted with an instrumentation amplifier (INA126), and ultimately sent to the controller for reading. We also learned that a driver chip (UCC27322P) was necessary for the MOSFET (IRF540N) to switch on and off with minimal noise. The schematic allowed us to move ahead with upcoming challenges, such as coding the MPPT algorithm and refining the flyback converter.
When choosing the correct microcontroller for this project, we needed something robust and readily available. The Arduino Uno development board (Figure 3) had all our controller requirements and specs, with the pulse-width-modulated (PWM) output signal being key. By default, the PWM frequency is set to 490Hz, but by manipulating the clock speed and timings, we adjusted this to 111kHz, which is fairly ideal for our flyback circuit. The analog inputs were required for (1st + 2nd) reading input voltage and current measurements, which are used to calibrate the PWM to the maximum power point of the solar panel, using the incremental conductance algorithm. Additionally, the PhotoPod powers down if the internal temperature (3rd) of the enclosure reaches 80°C, or if the output voltage (4th) is outside the allowable range. The digital ports control the display (6) and communicate with the battery (2) when power is available from the panel and the battery is ready to be charged.
From the conception of the project, we all knew that the transformer was the key component to the success or failure of the PhotoPod. With our input and output voltage limits and current requirements, the number of off-the-shelf flyback transformers quickly dwindled to zero. We contacted custom transformer manufacturers, but the ETAs for our design were longer than the entire Capstone semester, so we re-thought our approach and used the DIY method! We first gathered some research documentation pertaining to flyback transformer design, and decided to follow the product area (AP) method (Figure 4) , which is the industry standard for transformer design. Since we required a certain size of transformer, we had to choose a compatible core, but the material of the core would also affect the calculations, resulting in a vicious cycle, eventually leading to an EE core design with three primary turns (22AWG) and 60x secondary turns (30AWG) on the winding bobbin (Figure 5).
While seemingly straightforward, our enclosure had its own list of requirements. It had to be: non-corrosive, waterproof, partially transparent, and as compact as possible. These characteristics allow for its use in harsh weather environments and help prolong the lifespan of the PhotoPod. Although stainless steel or aluminum enclosures are optimal for our application, the price point for these cases would have doubled the project budget, so we chose a polycarbonate enclosure as a reasonable substitute. This reduced the thermal conductivity of the device, but negated the corrosion issue, while still allowing the display to be viewed when the case is sealed shut. We attached waterproof connectors for the input, output, and communication cables, giving the PhotoPod an IP67 rating, which we tested by submerging it underwater for 45 minutes (Figure 6).
We began development on our project’s circuit board at the end of October, however, PCB design was a new skill set that each of us had to learn, so the learning process began early in the project term. As we continued to build and improve our prototype, we concurrently advanced our PCB design skills by reworking and optimizing the board layout. Once we had determined how to set up a PCB library and schematic library, the first PCB revision was underway. The design included header pins to mount the controller and display, allowing easy removal and replacement of these components for reprogramming or repair. Footprints for three transformers were added, with wiring traces aligning them in parallel. With the transformers being the limiting factor for power rating, this tripled the PhotoPod’s capability to 150W, meeting another project requirement. As testing progressed, we quickly realized that a second revision (Figure 7) would be ideal for including additional features, such as electrically isolated output voltage measurement, improved PCB dimensions for the housing, and communication signal protection with metal-oxide varistors (MOVs).
We needed to use multiple libraries for the PhotoPod’s software, aiding the display initialization and high-frequency PWM output. All measurements were made using a circular buffer, which averages a series of different readings over time to increase the accuracy of the measurement. We supplied an external reference of 4.096V to the controller to further optimize the measurement accuracy. We wrote the incremental conductance algorithm (Figure 8) by comparing these measurements to their previous values and adjusting the duty cycle. This further affects the measured values, and as the code runs this cycle multiple times, the current and voltage readings hone in on the maximum power point. These measurements are shown on the display, which updates with each code loop, roughly once per second.
Along with the MPPT software, the controller is set up to run completely automatically. The device powers up whenever the solar panel supplies approximately 1W of power, and begins charging whenever the battery can store more energy and the input voltage from the panel reaches 12V. A flag is used for emergency shutdowns, which occur when the output voltage is in an unacceptable range, the battery becomes fully charged, or the internal temperature reaches 80°C. To maintain this ease of use, the PhotoPod automatically resumes after the error is cleared.
Our device needed to fulfill many key requirements, including maximum power point tracking, high efficiency, powered entirely by solar energy (no batteries or external power source), and above all, safe operation.
The maximum power point tracking functioned well, constantly tracking every change in the voltage, current, power, and the adjusted duty cycle percentage. This greatly increased the output potential of the PhotoPod, with up to a 50% increase in output power after tracking was completed. Efficiency was further improved by decreasing the voltage supply to the display (reducing power consumption by auxiliary components), and optimizing the construction of the hand-wound transformers over several different revisions. This ultimately produced an efficiency of nearly 80% for the flyback converter, and approximately 70% for the device overall (Figure 9), which exceeded our expectations for a device using hand-wound parts.
The PhotoPod needs no external power source, but operates solely off the solar panel and the use of multiple linear voltage regulators. The device remains safe through the use of output voltage limitation and overheat protections within the code, MOVs used on communication signal inputs, electrical isolation from the input side to the output side (accomplished with transformers and optoisolators), and fuses on the input and output signals.
The PhotoPod (Figure 10) is a micro-power converter, capable of converting low-voltage DC from a solar panel to a high-voltage DC to charge an end-of-life EV battery. A QR code leading to more information about the PhotoPod is given in Figure 11.
Although the Capstone term is over and our project is technically complete, there is still a lot of untapped potential with the PhotoPod. Our design is built around the Nissan Leaf Generation 1 battery pack, but through the software upgrade, the device could be compatible with a wide range of EV batteries. This would enable an alternate purpose for an even larger number of batteries. Although approximately 80% efficiency is decent, it could be further improved by incorporating a switch-mode power supply to replace the linear voltage regulators, an active clamping circuit instead of the passive setup currently in place, and having a transformer manufacturer precisely build identical spec flyback transformers to transfer power as effectively as possible. This would not only improve efficiency, but also reduce the heat build-up in the enclosure and extend the lifespan of the PhotoPod.
In the end, the total cost for one unit is about $200 (scaling to ~$120 when building 1,000 or more). This could be reduced by incorporating the controller directly onto the PhotoPod PCB, rather than attaching the development board as a separate module.
As a team (Figure 12), we are pleased with what we were able to accomplish in a few short months, and we anticipate seeing where these ideas will lead. As EVs become more and more common, it is important to plan for the impacts they may leave behind. The PhotoPod is just one small example of the wide variety of environmental solutions that will help propel us into a greener future.
Additional materials from the authors are available at:
References  to  as marked in the article can be found there.
Arduino | www.arduino.cc
 Camille Charluet & Wesley van Barlingen, “How long do electric car
batteries last?,” EVBox.
 “SMPS transformer area of product,” Electrical Engineering Stack Exchange,
Visit Photoplast’s Webpage for more:
PUBLISHED IN CIRCUIT CELLAR MAGAZINE • JUNE 2022 #383 – Get a PDF of the issueSponsor this Article
Navin Francis (firstname.lastname@example.org) worked in the shipbuilding industry in India before moving to Canada to pursue further education. He completed the Electrical - Marine & Industrial program at Camosun College to accelerate a career in the marine industry. His areas of interest are power electronics, electrical machines, and control systems.
Daniel Pedro (email@example.com) became an electrical engineering technologist after gaining experience in drafting and team leadership through Camosun’s AutoCAD Graphics course and his role as a supervisor during previous employment. He is looking to acquire his Bachelor’s Degree at the University of Victoria, after which he plans to work with industrial power and control systems.
Nic Lockwood (firstname.lastname@example.org) has experience in troubleshooting electronics and 3D rendering. He enrolled in Camosun’s electrical engineering technologist program to further his understanding of solar technology. After completing the Engineering Bridge Program, he hopes to earn a degree in electrical engineering at the University of Victoria.
Philip Shuck (email@example.com) has experience in trades and technology, and became an electrical engineering technologist straight out of high school. He hopes to receive his Bachelor’s degree at the University of Victoria and pursue a career in renewable energies.