Hello, my name is Adam Wakelin, the creator of Penguin Pi.

I am a high school student in San Diego, California, and an aspiring engineer. I love to learn about electronics, science, and radio; and when I find the time, I love to work on my personal STEM projects.

Penguin Pi has been a personal project of mine since late 2021. Throughout its on-and-off development over the course of the past four years, the design, purpose, and scope of the project has changed significantly. I can proudly say that I am still learning, and that this fun, humble project of mine has pushed me to learn and grow in my designing and engineering skills.

The project has simple origins, which date back to 2021. At the time, I was very interested in portable or handheld Raspberry Pi technology, as most Raspberry Pi computers are confined to a desk, requiring a rat's nest of wires to operate. I spent a lot of time browsing the internet and reading various technology or hacker forums, where Raspberry Pi enthusiasts and hobbyists showed off their portable Raspberry Pi computers, which are commonly referred to as “Cyberdecks”.

Many of these Cyberdecks were designed to resemble devices or computers from popular video games or movies, such as Fallout, Star Wars, and more. The designs looked amazing and it was obvious that a large amount of time and effort went into their creation, but they did not seem to be the most practical due to their form factors and interfaces. (This is not to say that my device is any better, in fact, many of these devices are still far smaller and more ergonomic than mine)

Pipboy 3000 from Fallout, featured on Adafruit.com (Article written by the Ruiz brothers)

Other hobbyists popularized “doomsday” Raspberry Pi-based rigs, which were more often than not built into a Pelican case. These makers would cram as much functionality as possible into these devices, which both looked clean and high-tech, and had a clear function that they could perform well. These Cyberdecks were some of my favorites, but they were still a bit large.

The Raspberry Pi “Recovery Kit”, created by Jay Doscher

It was around this time that I started to research Raspberry Pi “handhelds”, which were another genre of Cyberdeck. I was interested in these devices, because their smaller size made them easier to transport, and/or recreate on a tight budget. Some hobbyists were also debuting devices that were simply just the devices themselves - the function or use case was determined by the user. Some of the inspirations for my project were the YARH.IO handheld, which had a full keyboard, a trackpad, and a screen, the Blackberry Pi, the Brick Pi, and the SmallPi.

The YARH.IO, a handheld Cyberdeck gadget with a full keyboard and mouse pad.

The Blackberry Pi, a handheld with a BBQ20 Blackberry Keyboard, created by Taylor Hay

*The Deckility, created by Ken Van Hoeylandt

Most of these projects were small, practical, and inexpensive to reproduce, and in all honesty, were extremely smart designs that are still more practical than the modern iteration of my Penguin Pi project.

Even though it would have been easier and far more practical to modify one of these existing designs, (Multiple are still open-source) I wanted to take it upon myself to create a Cyberdeck of my own, through reverse engineering and my own spin on the idea.

The project had no clear purpose when it started. The entire goal was simply to create a mobile interface for the Raspberry Pi. I had multiple ideas for possible use cases, such as a Wikipedia Backup, a digital library, and even a hacking multi-tool.

Looking back, I feel that I was focusing too much on trying to formulate a solution to a problem that did not exist.

I still felt that the design needed an overall purpose behind it, and so the early iterations gained a working name: ERIC. ERIC stood for Emergency Recovery of Information Computer. My Raspberry Pi OS was already filled to the brim with special software that I had downloaded, including local file converters, Wikipedia backups, local copies of software, manuals to different computers, hardware documentation, and all of my personal data. I felt that it was important to have a computer that could host all of the important local files and software that I cared about, that could operate without Internet access for long periods of time.

ERIC became synonymous with the handheld device itself, (Now referred to as Penguin Pi) and my Raspberry Pi OS. The project was centered around building an interface that was simple to build, repair, and modify, as well as host all of the necessary software to do basic computer tasks off-grid.

Once I had a clear understanding of what I wanted to build, I made a list of criteria for the hardware.

The criteria for the initial design was as follows:

• Simple assembly of the case using M2.5 through-hole bolts and nuts (Standard for Raspberry Pi and similar boards)
• A joystick, two buttons, and screen to function as the interface.
• A 2.5 Inch or larger screen for the Raspberry Pi

The constraints were as follows:

• Total cost of parts should not exceed a $350 maximum budget.
• The device needs to have an interface that is comfortable to operate while being held.
• The components should be plug-and-play, with minimal soldering required. (This excludes the drivers for the screen and the programming for the joystick as well as buttons)
• The case needs to be simple to print and assemble.
• The device needs to have a durable case that can protect the Raspberry Pi and other components from damage, should the device fall to the ground.
• The device should be small enough to be held firmly in the hand without slipping, not too thin, and not too thick.
• The screen needs to be durable and resistant to damage (Can be accomplished with a simple screen protector)

The first step in the project’s development was to elaborate upon the basic sketches that I had made in my notebooks and scrap paper. The initial designs that I had drawn had potential, but I needed to assess whether or not they were practical and/or if they would fit the criteria and constraints that I had made for myself.

I started to draw designs with rough dimensions, shading, and diagrams in order to visualize my ideas, which I did in my spare time. I drew the majority of these pictures in the small notebook that I carried in my pocket wherever I went, so I frequently would jot down ideas whenever I had the opportunity.

I ended up filling up around 3 notebooks with very rough sketches and notes, but despite the abundance of ideas, the project didn’t get anywhere. I stopped drawing as frequently, and took a break that lasted a few months.

JANUARY 2023

In January of 2023, I was cleaning out my desk and stumbled upon my notebooks from a few months prior, which inspired me to keep the project going. This time, I didn’t want to waste time and let the project fizzle out, so I made a few more refined sketches for the design, and started 3D modelling a basic case.

For the 3D modelling, I opted to use TinkerCAD, which is considered to be extremely simplistic and limited CAD (Computer-Aided-Design) software. At the time, I did not know how to use more powerful software such as Fusion 360 and/or Autodesk Inventor, and TinkerCAD was my go-to. In all honesty, I still find TinkerCAD to be useful for sketching out shapes with simple geometry, before redoing it in Autodesk Inventor for more detailed work and for the tools it offers.

The case was a relatively simple layout, which took many design cues from the Nintendo Gameboy. I added brackets and mounting holes on the corners to allow for through-hole assembly for the different sections, as well as a keyring hole for a lanyard. The most notable difference between this early version and the modern versions of this project is the orientation, as this version was vertical, and not horizontal.

3D model in TinkerCAD: Top Plate, Bottom Plate, and Walls.

One of the primary reasons that I chose a vertically-oriented design was to create room for a BBQ20 Keyboard, which was manufactured by a seller on Tindie known as “Solder Party”. The BBQ20 is a special keyboard derived from the Blackberry Classic, which was a cellphone released by Blackberry in 2014, and discontinued in 2016. I chose this keyboard because of the quality of its construction, and its relative rarity.

To my knowledge, the keyboard is a custom board with a recycled Blackberry keyboard installed on top, which gives the device as much functionality as it would if it were installed in a regular Blackberry. The BBQ20 has a USB-C port, as well as I2C functionality, which allows makers to install it into their projects seamlessly. I’ve always adored the old Blackberry phones for their simplicity, and most notably, their physical keyboards.

*As of May 5th, 2025, the BBQ20 has been sold out, except for specific times when the seller releases a new batch!

In order to shrink the size of the device, I factored a ribbon cable into the design, which allowed me to connect the Raspberry Pi to a PiJuice battery, (Manufactured by Pi Supply) and still have room for the BBQ20 to fit snugly on top. Below is an image of the naked device (no case)

*The ribbon cable is visible in the image, which is connected to the battery board. (It is the pink board below the keyboard) The screen shown is an HDMI 3.5 Inch screen designed for the Raspberry Pi, and the keyboard is the aforementioned BBQ20.

I chose the PiJuice board because of its plug-and-play compatibility, and easy-to-install Shell interface, which comes with an optional GUI. It was originally installed between the Raspberry Pi and the screen, but the ribbon cable allowed me to move it underneath the keyboard, which made the device thinner.

*The Pijuice board comes in multiple variants, including one with an onboard battery. By modifying the hardware switch on the board, a larger battery can be accommodated.

To make the device easier to use, and to ensure that the interface was not limited to just the touchscreen and keyboard, I added mounting holes for an Adafruit 128x64 OLED Bonnet, which was designed for the Raspberry Pi Zero, but was compatible with the Raspberry Pi 4.

The bonnet, pictured above, includes a joystick, a mini screen, and two buttons. It can be programmed with a Python Library that can be found on their website and in the hardware documentation.

It is NOT plug-and-play, so I decided to create a basic Python script to convert movements into mouse inputs, and a step-by-step setup guide to install the screen drivers for the device.

APRIL 2023

After making some basic adjustments to the case design, and remodelling the backplate for the device, I organized the .STL files on my USB drive and took them to the San Diego Library in 4S Ranch, which, at the time, was the only library with an available printer. 

Luckily, the printer was open, and after contacting the librarians and the dedicated 3D printing manager, I was able to load the .STL files from my USB.

PLA was the only filament available for the printer, which did not affect the printing process, as I had already planned to print the prototypes in PLA. PLA is commonly used in most home 3D printing, as it can be printed at low temperatures, and does not require a heated printing bed. I ultimately chose to print the case using a Glow-In-The-Dark filament, which came out very nicely. The print took approximately 2 hours, and I picked it up the next morning.

The front plate, back plate, and main case walls disassembled. Barely visible in the top right corner is one of the friction-fit covers for the USB ports.

The assembled case, without electronics inside.

As soon as I got home, I cleaned up the waste material that had accumulated on the print, which was easier said than done, as the almost-clear material made it hard to see the straw-like fibers that were on its surface. I lightly rubbed sandpaper on the surface to smooth it down, and manually used an Exacto-Knife to clean off the majority.

Once most of the waste had been removed, I rinsed the PLA in water BRIEFLY, and dried it off as fast as possible. I left it to air-dry for about an hour before I moved onto the next step. Once I was confident that the case was dry, I brought the components as well as the case to my workspace, and began to assemble the device.

For the through-hole assembly, I used M2.5 bolts and nuts, which I ordered off of Amazon, as I could not find the proper size at my local Home Depot and Lowes. I also used a screwdriver to remove any remaining filament scraps from the holes.

I started by attaching the majority of the device directly to the backplate using brass standoffs, and securing it with bolts, and then mounted the keyboard on the inside of the top plate. I was pleased to see that the keyboard fit snugly into the case, and the keys were able to be pressed without flexing or friction.

Once everything was installed, I placed the top plate on top to see how the device would look.

The top view of the device looked exactly as I imagined it, but once I looked at the sides, I noticed a few problems right away.

As seen in the above images, the boards were far too close to the walls of the print. I printed the case with extremely thick walls in order to make the case more durable. Unfortunately, I placed the mounting holes for the Pi and the battery too close, and therefore, the walls would not fit. The case could not be completely assembled without the walls, which provided a mount for the top plate and backplate.

On top of this, there was also no room for the joystick or buttons! (The holes can be seen between the screen and the keyboard)

The final nail in the coffin for this design occurred when I opened the file for the case. Unfortunately, there was a saving issue with the .STL file, and I lost the original copy, with the edits that I had made. The only model that remained was the completed device, but all of the components in the file were fused together, which made editing it extremely difficult if not impossible. At that point, I realized that it was easier to just start from scratch.

About a week after this, I stopped working on the project and took a month-long break.

MAY 2023

Even though I was not actively working on the project, I still wanted to see it completed. I had also been sketching out various designs in my spare time, which closely resembled the original. The vertical design created a size issue that was difficult to work around. The components simply needed more room without making the device thicker or larger.

But eventually, a new design began to emerge:

In these new drawings, I re-oriented the device to be landscape, which allowed me to downsize the device significantly.

The new landscape-oriented design came with many benefits, as well as a few drawbacks:

Pros:

• The design allows access to FIVE extra ports on the Raspberry Pi, while the previous design only allowed access to the USB-A ports.
• The previous design needed an extender strip inside the case in order to add extra USB ports. Now that the native ports for the Pi are exposed, the device doesn’t need one.
• The screen is much easier to factor into the design, and is easier to utilize as a touchscreen. The screen can also be upgraded to be much larger than a 3.5 Inch screen, and GPIO support is available.
• The device can have a far larger battery, with support for up to 12000 mAH batteries. This can run the device for up to 30 hours realistically, which is a massive performance leap.
• The device has more room for modular components.

Cons:

• The device needs to be held with two hands, as opposed to one (The previous iteration was uncomfortable to operate with one hand regardless)
• The keyboard has to be removed to make room for more components.
• The device may become slightly thicker.
• The existing Pijuice board would need to be swapped for a smaller variant (Does not affect battery size)

Even with these drawbacks, the new design would allow for greater flexibility and modularity, and was easier to customize. Starting in mid-May, I modelled the second iteration of the device:

As one can see in the image, the left side of the device has 4 mounting holes, designed to mount the joystick and button board to the underside of the top plate. I also added a black “mock screen” to show where the screen was located. The bottom plate has also been modified to instead serve as a battery housing for a 12000 mAH battery. The center part of the “sandwich” now consists of the walls, as well as an integrated bottom plate. It is still missing holes for the ports, and cutouts for the joystick and buttons.

SEPTEMBER 2023

By September of 2023, I had finished modelling the vast majority of the new case. The design now had proper cutouts for the joystick and buttons, a debug screen, and the main screen. Along the sides, there were slots for the USB-A ports, as well as Ethernet. On the adjacent side, five new slots were added for the USB-C power connector, the two Micro-HDMI connectors, and a slot for an optional ribbon cable, should one decide to install a camera. I also added a hole for the 3.5 MM audio jack for the Raspberry Pi 4 and 3 models.

The new side USB port slots.

The device gained its new working name around this time: "Penguin Pi", which is named after Tux the Penguin, the mascot for Linux, and the Raspberry Pi.

To commemorate the new name, I created the 3D "stamp" that I would add to all future models of the Penguin Pi:

Rotating the logo makes the hidden Wi-Fi symbol more visible!

Unfortunately, it was hard to find time to work on the project, schoolwork, and my extra-curricular classes at the same time, and so the finished model would sit on my laptop until February.

FEBRUARY 2024

Once again, I restarted the project, and the design was now ready to move into the prototyping phase. After a short period of time, I downloaded each component in the case as a separate .STL file, and loaded them onto my USB.

This time, I did not print my case at the library, because I was afraid that the prints would fail and/or the printer would not be able to handle the overhangs in the design. Luckily, my robotics team, Team Spyder, had just received new X-1 Carbon 3D Printers, made by Bambu Labs. The X-1 Carbons are very versatile printers that can print quickly, reliably, and with a variety of materials. Since this design contained many overhangs and complex shapes, these were the perfect printers for the job. After consulting the team and the donor of the printers, I was given the green light to print my case out, so long as I paid for the used filament.

Before I could print the case components, I needed to modify the .STL files to prepare them for printing. I loaded the files onto my school desktop in our computer lab, and imported the files into “Bambu Labs”, the official printing application for the Bambu Labs printers. Once I oriented the parts and sliced the layers, I transferred the files onto an SD card, and then inserted them into the printer.

The print took approximately 3 hours to complete, and I picked up the completed prints the next morning. I was still modifying the battery compartment, and I did not need the device to have its own power supply for prototyping, so I opted to only print out the top plate, as well as the main body. The results exceeded my expectations:

The print was extremely crisp and smooth. The surfaces had little to no scrap fibers, the edges were clean, and the overhangs had no sagging whatsoever. This print was far better than the previous version in terms of print quality, and the difference was obvious. The printer handled the larger overhangs as well, and unlike the previous print, the mounting holes were intact and had no signs of warping.

Needless to say, I was very happy with the design. With the new integrated bottom plate, (Bottom plate directly fused to walls) the design was extremely sturdy and tough. I even felt confident enough to try to bend the main body, (not the top plate) and it barely flexed. Although the case had exceeded my expectations, it still needed to be assembled around the Raspberry Pi itself. Once again, I decided to use M2.5 screws, brass standoffs, and my screwdriver to mount the Pi and its internals inside of the case. 

Unfortunately, I came to the conclusion that my dimensions were off.

The main body and top plate of the case were aligned perfectly, as were the USB, HDMI, and audio ports for the Raspberry Pi, but the Ethernet and USB-A port holes in the case were misaligned with the actual Raspberry Pi. This meant that the Raspberry Pi could not be mounted correctly without colliding with the plastic, which was a few millimeters off.

I ultimately decided that testing the Pi was more important than having perfect USB shields, so I used pliers and a knife to break off the USB shield. Doing so allowed me to mount the Pi properly and have access to all of the USB ports, so I was able to start testing the Pi itself. It is important to note that the case was not designed for the screen that I already had installed on top of the Raspberry Pi, and so I had to remove the top plate in order to mount it properly. This was a necessary trade-off in order to test the Pi.

But by the time I had everything assembled, it still required a mouse, keyboard, and a connection to my desk’s monitor. The only difference was that the Pi now had no internal power source and sat in a plastic shell with no cover.

Around this time, once again, I became extremely busy with schoolwork and extracurricular activities, and the project was halted yet again.

JUNE 2024

Even though I was busy with schoolwork and life, I was still making incremental progress on the project. Around this time, the Raspberry Pi 5 became available, and along with adjustments to the case dimensions and mounting hole orientation, I changed the USB ports to accommodate the Raspberry Pi 5 as opposed to the Raspberry Pi 4.

This change was made because the Raspberry Pi 5 was the first Raspberry Pi model to have a dedicated port for a fan. The Raspberry Pi Foundation added a fan port because the Pi 4 and Pi 3B+ were notorious for heating up significantly due to their lack of a native fan. Installing a fan on any Raspberry Pi model prior to the Model 5 required sacrificing an entire GPIO strip just to connect two to three wires to the 5V and GND pins. Doing so made the 40-pin GPIO strip incompatible with a HAT. (A board that directly sits on top of the Pi, and requires all 40 pins)

A Raspberry Pi user installing a fan on the Pi 4B.

During the Summer of 2024, I did not work on the project as much as I would have liked, but during the Fall when I returned to school for my Junior year, I dusted off the project and gave it a major overhaul.

DECEMBER 2024

I realized that adapting the device specifically for the Raspberry Pi 5 violated my original vision for the project, which necessitated modularity and flexibility. Making multiple cases for different Raspberry Pi models was time consuming and not practical, and would be extremely tedious.

Instead, I settled on a more unique approach. In the 3D model for the case, I strippd out the USB port shields entirely, both for the USB-A/Ethernet side, and the side-facing USB-C, HDMI, and Audio Jack. I then backtracked through the extrusion flowchart for the 3D model and increased the height of the device (From bottom of front to top) by 10 millimeters. This created more room for the 90 degree expansion headers for the ribbon cable, and created room for springs to be added behind the power button. (More on that later)

With the shields removed and space made within the case, I turned the main body into a one-size-fits-all design, which could now adapt most models of the Raspberry Pi! (excluding the original Raspberry Pi, and Pi Pico microcontroller)

More info on modularity and Raspberry Pi adaptability in May 2025 Documentation (Down below)

Main body of the case, without shields

I also added tabs on the edges of the case walls, so that when the top plate, main case body, and battery compartment are fitted together, the tabs will lock together and create a more solid connection.

This new iteration was now far more customizable and adaptable than the previous, but it still required shields for the USB ports. This new model was designed to have separately-printed shields installed, which could be swapped, depending on which model of the Raspberry Pi was being installed into the case.

Optional shields for the case. The red shield card can adapt both Pi 5 and 4; blue and green shield adapts 5 and 4 respectively.

EXAMPLE: Case with Pi 5 shield installed to fit Pi 5 USB and Ethernet ports.

The shields are all friction-fit, meaning that installation and removal can be done by hand. 

The top plate also received an upgrade! The old prototyping cutout for the OLED bonnet was replaced with proper slots and button holes. I also created 8 variants of buttons for testing. The buttons simply fit into the holes (There are lips on the bottom of each that prevent them from falling out of the case) The joystick goes through the square slot, and the rubber cover already installed on the OLED bonnet keeps dust out by creating a small seal.

After a few more tweaks were made to the model, it was ready to print.

FEBRUARY 2025

Almost a year after the last 3D print, “Penguin Pi IV” was ready. After loading the files one by one onto my USB, I once again printed it on the X-1 Carbon. (After arranging to pay for the filament with my robotics team) This time, the initial print was made with ABS filament, but due to heating issues in the printer, the print failed and started to warp about an hour into the print.

Luckily, there was an available printer with black PETG filament, which was a filament that I was interested in printing with, and after about 20 minutes and some reslicing of the model, I was able to successfully print the full case out. The process took about 4 hours, and I picked it up at my high school the next morning:

This time, the case, top plate, buttons, and shields all printed without issue! After removing the tree supports and parts from the printing plate, I took the print home and installed the Pi inside.

All of the shields fit inside without issue, as well as the side card, and the Pi sat nice and snug inside. The only issue was that the side card and shield were a bit too tight when fitted, but that issue is one that can be solved using sandpaper. The replaceable shields worked perfectly, and all of the cutouts were now in the correct position.

The only issues that persisted with the case were that the mounting hole wells were inverted, (The wells where the head of the bolts sit were on the inside and not the outside of the case) and that the case needed to be approximately 5 Millimeters taller to accommodate the screen.

MARCH 2025

In mid-March, I started working on case dimension reworks for the Penguin Pi. The first thing that I did was create a new file titled: "Penguin Pi - Rework Extensive"

The first adjustment that I made was to adjust the height of the walls from 20 millimeters tall (2 centimeters) to 25 millimeters tall, (2.5 centimeters) in order to make the screen flush with the screen slot in the top plate. The second step was a difficult one, which involved disassembling the main body of the case, organizing each step into an exploded view, and adjusting the mounting hole  wells. I mentioned in the documentation for February that the mounting hole wells were inverted, and needed to be facing downwards, as opposed to upwards. This is so that the heads of the bolts stay flush with the bottom of the print, so that the 12000 mAH battery does not become damaged or scratched. I also made the decision to shift the mounting holes to the right by 1 centimeter, so that the USB ports of the Pi would now be flush with the side of the case.

The new USB port configuration. Notice that the USB and Ethernet ports are now flush with the outside of the case.

New shields were also created - Notice the "L" shape, which helps the shields stay locked inside the case, as well as the horizontal lines on the shields. Those are finger tabs, designed so that the shields can be removed using fingernails, a flathead screwdriver, or a knife.

The shields are also 5 millimeters taller than their predessors, which allowed me to add a "bridge" over the top, which makes the shields more resistant to snapping. The previous shields, as can be seen earlier in the documentation, had tall 2 millimeter-thick sticks, that were vulnerable to snapping.

The shields are designed to be easily swapped, and easily identifiable. The Pi 5 has the same side port configuration as the Model B+ and onward (Excluding the Pi 3A+, which has a smaller form factor, and a different configuration)

MAY 2025

Adjustments have been made to the power button, and new lanyard loops have been added. (Temporary, may not be the best design)

The screen slot has also been shrunken to fit a 2.8 Inch GPIO screen from Waveshare. This design change was made because moving the Pi to be flush with the side of the case made the old choice of screen, a Pimoroni Hyperpixel 4.0, impossible to fit within the case walls. The 2.8 Inch GPIO screen is sufficient as a monitor regardless, and the smaller size does not impact function.

The Penguin Pi name has been updated to "Penguin Pi V", and the "EXECUTE ORDER 66" splash text has had its font changed to Sans. This font will be much more legible for a 3D print compared to Multilanguage.

Device, with transparent top plate to make insides visible (Missing all electronics except for Raspberry Pi 5)

Aerial view with top plate and Raspberry Pi 5 removed. Note the mounting hole positions. There are five visible mounting holes! (For the Raspberry Pi, not the OLED Bonnet and Battery on left side of image) The extra mounting hole (Center of the farthest right row) allows for a Raspberry Pi Zero to be mounted inside. (And still have 3 points of connection to the case)

Example: A Raspberry Pi Zero mounted inside. With this configuration, it is best to remove the side shield, (Depicted in Yellow here) because the small size of the device does not conform to the sides of the case. You can remove the shield and run cables directly into the Pi instead if a USB or display connection is needed.

EXAMPLE: A Pi 4 mounted inside of the case.

The Pi now has proper vents for airflow. Previously, the fan would have had to push air out of the USB port shields, and into the battery compartment. This design will help the Pi stay cooler.

On May 15th, I had the opportunity to print my latest model, and so I downloaded all of the case components separately onto my school computer, and then onto an SD card. I then opened the file in Bambu Labs in the school computer lab, arranged the components, and exported the .gcode file for printing.

The file on the Bambu X1 Carbon.

This time, I configured the file to be printed in Black PLA, because PETG filament was in short supply, and using it was unnecessary regardless, because this was still a prototype. I configured the print with PLA tree supports to prevent sagging on the overhangs, and to keep the mounting brackets straight. I asked for help from other members of my robotics team, and once they checked my work, they gave me the green light to continue. Once it was set up, I started the print.

Right away, problems started to appear.

An image of the printing bed. This is just the first layer, and while it may not be visible in the image, the print had debris accumulation on one of the corners of the print, and the material had already begun to warp.

I ultimately decided to cancel the print, because it was clear that it was only going to get worse. The next morning, I came back to the robotics room and removed the magnetic printing plate so that I could clean it off and start a new print.

The printing plate, after removing it from the printer.

The removed prints, which were paper thin, next to an Expo marker and a pen for size reference.

The cancelled print in my hand. I will use it in the future for size comparison.

After I cleaned off the printing plate, I reattached it to the printing bed, set up the print again, and this time, I left the door closed when I printed it. Once it was running, I left it for an hour before checking on it again.

Unfortunately, the print failed AGAIN.

If you look closely at the print on the right side of the image, and then look at the bottom left corner, you can see the warping.

Unfortunately, the print warped halfway through the printing process, and so once again, I had to cancel it. 

The next day, I visited the printer in the morning and cleaned off the plate, and removed the failed print. I knew that the warping was being caused by poor surface adhesion to the printing plate, as well as uneven cooling. Some parts of the case were cooling off faster than others, which was causing the material to flex, break off of the printing plate, and warp.

I cleaned off the plate a third time, but this time, I altered some variables. This time, I left the printer door completely open in order to keep the cooling even, (as opposed to airflow flowing over one corner) and applied glue stick to the printing plate to keep the first layer attached, and to passively correct warping.

Halfway through the print- No warping, looks promising!

I let the printer print for the next four hours, (Into Friday afternoon) and I came back to pick up the finished result the following Monday.

This time, the device finally came out properly!

The trees snapped off nicely, without issue, and there was not any visible sagging! I did not account for tolerances on the interlocking surfaces, so some light sanding would be needed in order to create a proper fit.

The keyring holes also came out nicely. I was worried about them sagging or failing because of their overhangs, but the tree supports kept the overhangs up, allowing for the rings to be printed without issue.

While I was sanding and preparing the prototype case for assembly, I CADed a custom mount for a 5V 30MM Fan, which is the standard size for Raspberry Pi fans:

A mockup fan, pictured above, mounted on top of the bracket. The fan is centered between the Raspberry Pi's mounting holes, right over the RAM and the BroadCom CPU, which tend to get hot on a Raspberry Pi and throttle. The fan will reduce the heat buildup within the device.

This log will continue to be updated.