The Idea #
Running the 'Coding Club' at school, our mentor Mr. Bush suggested creating a "Videowall" out of Chromebooks the school was going to throw away. We thought it was just another teacher with a fanciful tech idea with no idea how hard it would be to execute.
This project was a collaboration with my friend Varun Biniwale. I was responsible for the hardware, and he designed the incredible software, see his blog to learn about the unexpectedly complex software needed to run these dismantled computers.
We got started on the project in late 2021 anyway, mainly because our mentor was one of our favorite teachers.
The first part we started on was the software, as it seemed like the easiest, and frankly like the only possible part of the project at the time. What we wanted to do was have one long video playing across multiple screens. Putting together a line of the Chromebook screens and connecting them to one computer was an idea, but we had no clue how to interface with the Chromebook screens and so we decided against this idea.
The first few weeks, we spent our Monday afternoons trying to find existing software to synchronize videos across different computers. While we wanted the final product to look like a single, continuous video, in reality we would take a long video and split it into smaller parts of equal length to create the illusion. We experimented with Syncplay using VLC and looked at paid software such as Navori, but none worked for our use case.
Giving up on a existing software, Varun got to work on a software to synchronize the video across multiple computers. He called it C-Sync after other synchronization technologies such as G-Sync, FreeSync, VSync, etc.
Comic Sans forever
Again, read about the insanely difficult task Varun had to actually make it run on the Chromebooks here.
The Physical Part #
Now comes the part which was actually figuring out how to make the physical videowall.
We’d already taken apart one of the Chromebooks, which was a fun process if not a little scary with the batteries that could burst into flames. Good thing we wouldn’t be messing with power cables or anything like that…
For the first iteration, we just wanted to get a good digital layout of the motherboard. A neat trick I used was to scan the motherboard with a ruler. The scan should be to scale, but just in case the ruler lets you resize it easily.
For this version, using our school’s laser cutter was the easiest method, making a very clean piece.
Thinking everything was going smoothly, I quickly realized that this would not work at all with the screens, as the motherboards are longer than the screen itself, meaning that they would clash into one another when placed in a line.
This was the first major setback for the project. Using acrylic was out of the question as even if we made standoffs so the motherboards didn't touch, how would we attach the screens in front?
Computer Aided Discomfort (CAD) #
My experience with 3D modelling was Tinkercad, a simplistic browser-based modeler that I’d used for small projects, and Blender, a more artistic focused 3D modelling software.
For this more complicated task, I needed to learn a new program.
While Fusion360 by Autodesk was confusing to use at first, it was leagues better than attempting to make the backplate in Tinkercad or Blender.
Thankfully, importing the outline I had scanned meant that it was a lot easier to create the first prototype backplate.
After printing it off on a 3D printer, I slotted two motherboards through the pegs...
Unfortunately, the printer I was using couldn’t print a backplate long enough to support the entirety of the two motherboards, leading the ends to sag heavily.
I titled this section Computer Aided Discomfort because while it seemed straightforward what I had to do, I spent months trying to get Fusion360 what I wanted to do.
I'd print a new version to fix a problem, realize another issue, and have to print a whole new version, only to realize that something else was broken etc.
Initially unsure on how I could proceed, a thought came to me while on the plane and I pulled out Fusion and made a new version with add-on pieces that would be printed separately and could then be attached to support the ends of the motherboard.
Later, I changed the add-on pieces to attach via a clip system so I could remove pieces if need be.
After many frustrating setbacks where the motherboards just barely didn't fit, I finally settled on a robust design to hold both motherboards. Note, this was 14 major revisions in.
At this point, while it may seem like this whole process was pretty straightforward, at this point we were a solid two years into the project. While a lot of it was downtime due to Varun and I having many commitments, there were many setbacks that took weeks to solve.
The Screens #
Now onto the actual screens themselves, as having a row of motherboards on their own while pretty, isn't what we want.
Following the same scan method I used for the backplate, I recreated the screens inside of Fusion360, allowing me to build an accurate retaining frame to hold it to backplate.
I used the same clips as I had for the additonal pieces to attach the screen holder, which led to many frustrating attempts to try and put the pieces together only for them to snap.
Instead of making a new clip that would be easier to attach without the risk of breaking, I instead ignored the issue and continued to make the final touches to the design, adding cable channels to clean up the overall look of the backplate. Even if it was being hidden away, I wanted it to look as good as possible.
For attaching it to the wall, we used a French cleat system which allowed the individual parts to be easily removed and serviced if needed. Doing some janky tests with tape, the project held well.
Finally, after many many iterations in Fusion360 (120+!), the final CAD file was finished.
Doing some sketchy cable splicing (first time soldering and it's power cables, whoo!), we could power two motherboard screen pairs with one cable.
The final stretch #
Throughout this project, the absolute workhorse that is the Bambulab P1S printed over 10kg of plastic with a total print time over 300 hours for this project.
Printing everything off we now just had to assemble everything.
Taking some picturesque photos...
We hung all the motherboards and their screens onto the wall.
After ensuring everything worked and replacing a few scratched screens, we cleaned up the cables and added a border.
Over two years after starting this project, we were finally finished.
Thank you very much to Varun, without him this project would never have been possible and I would have never wanted to do it. Our Design teacher Mr. Bush also played a massive role for pushing us to do the project in the first place, and I can't thank him enough for being an inspiration to do design.
If you would like any of the files, feel free to reach out.