Part 3: Final Testing and Demolition

This first bit probably should have been in the last video, but it was getting a little long. All that really did was make this section longer. Starting off, I put together all the components from the last video. The only thing I didn't wire up was the power switch which would have made continued testing harder.

Final Testing

I wasn't quite sure how the auto-disconnect headphone jack was intended to work. The datasheet isn't that helpful and slightly cryptic, but that's what breadboards are for, I guess. The audio amplifier is connected to power via the Power Boost Charger, gets audio input from the RPi board. Output is connected to the headphone jack which is also connected to the small speaker. The test controller is plugged in via USB and the RPi is powered off the Boost Charger. As a last test, the 3.7v LIPO battery is connected and main power disconnected.
The only missing feature here is the power switch with an orderly shutdown, which was one of the last things I figured out.
Here you can see the two points on the back of the RPi board where the audio connections are made. On the other side of the board, below the red wire, is the headphone jack. It was simpler, and saved more space, to just tack on these wires instead of using a 3.5mm jack wire tail.

Assembling all the test components together was pretty easy. Originally, I had too many grounds connecting everything which produced some static coming out of the speaker. The audio amplifier has a small adjustment pot for volume. I tweaked it until the rest of the static went away, and later adjusted the system volume within Emulation Station. This was also the first time I tried running it all off of the 3.7v LIPO battery. After some fiddling, everything worked as expected.

Demolition

Cutting into the case is where this project gets serious. I put off doing this until I had everything and a better idea of how it would all fit together. Gutting the back of the case was an easy decision. There were several screw posts and ridges for supporting the logic board of the GAMEBOY, most, if not all, were removed. Another key piece to remove was the backing for the game cartridge area. This was in the direct way of the RPi board, and it still didn't completely fit with it removed. I made sure to conserve the supporting plastic for the power switch and leave a little bit of a rim around the cartridge area to attach the acrylic cover I made. Later on, I cut away some of the battery compartment near where the dremel tool is in the picture above. I put this off until I knew it had to happen. The battery compartment on the GAMEBOY case is actually pretty important structurally. Two of the case screws are attached through it, as well as reinforcement support for the gamepad area. These supports can be seen in the picture above to the left and right of my thumb. They keep the PCB from flexing and eventually breaking from jamming on the buttons.
Around this time in recording, the PCBs arrived from the fab house. They weren't perfect to, but work great. Some of the mounting holes required minor drilling or filing out. Next up was drilling holes for the x/y buttons. I used the DXF file from the previous section to create a template for where to drill the holes. Again, I was really nervous about this because it was the front of the GAMEBOY and didn't want to mess it up. The template lined up perfectly with the PCB. I under-drilled the button holes, routed them out with the dremel most of the way, and then finished with files and sandpaper until the buttons moved in them easily.

Installing the Buttons

I spent a lot of time on mounting the new buttons because they are probably one of the most important features of the build. To both retain the x/y buttons and ensure they moved only up and down in the slot, I cut out the supports from the SNES controller. Cutting off the excess plastic and supergluing them in place was the next step. I only had a slight issue with alignment on one of the buttons which was fixed by running a tube of sandpaper through the hole until the button moved freely. Then I reinforced all of the supports with super glue. The end result almost looks like it was made that way.
There's a molded depression around the A/B buttons on the case. I couldn't think of any reasonable way to reproduce this without messing it up, so I didn't bother. The end result is very natural and has thrown off a few people I've shown the finished product to.