Building a prototype retro arcade machine

The aspirational goal is to build a full size arcade machine. But first things first.

A friend of mine gave me the idea to try this and he's got all the software set up for RetroPie.

Buying the important stuff

I bought a Raspberry Pi, a monitor, some USB speakers and some cables.

Then, as our Melbourne, Victoria, Australia, 2020 COVID lockdown lockdown was ending the, things stalled.

Planning a box to house the controls

In 2021, I started thinking about picking this up again. This time I bought:

• Xin-Mo PS3/PC DUAL Player Controller since I'd read that it should work with RetroPie. And it's not a keyboard encoder so I figured that's probably a plus (if I can get it working of course). I bought it here.
• A Sanwa arcade joystick.
• An assortment of buttons.

The plan is to build a smallish wooden box to house the joystick and buttons and then go from there. And one of the trickiest things seems to be how to mount the joystick in a way that is solid and robust enough to actually play with and not have nasty bolts on the playing surface. 

There is a lot of good information to be found. Some good ideas here.

Mounting the joystick

The mounting plate on the joystick does not overlap all the circuitry, so that makes things tricky as I didn't want to have to cut out a huge hole in the wood to get it through. I was only going to use 12mm MDF, which is not a particularly strong wood so mounting the joystick from the top was not going to be feasible without some kind of reinforcement.

Mounting from the bottom seemed like it could work but I didn't want to lose too much joystick height. So to deal with that I'd need to chisel out the size of the mounting plate both on the top and bottom. With only around 12mm of wood to play with, removing 3-5mm on the top and bottom meant this was not going to be strong enough if bolts or screws went straight into the wood.

I found this at Bunnings (76¢) which was roughly the size of the mounting plate and had holes in the perfect positions. And it was only 1mm thick so it could be used to reinforce the top of the mounting area.

I'd bought a piece of 600x900 12mm MDF and I cut off a 25cm x 45cm piece for the top and started drawing lines. I drilled the holes for the joysticks early on. I only have one joystick for this prototype but I thought I'd make room for another so I can actually use it to play double-joystick games.

You can also see I removed the middle connecting part of the metal plate (with pliers) since that was right where the joystick handle needs to be.

The wood was quite soft, so chiseling it out was easy. Getting it relatively smooth and level was more difficult since I was doing it all by hand.

It ended up like this, looking at the top, but I'd chiseled out more than I needed to.

So I flipped it over and made it the bottom instead. That way I'd get more joystick height.

There is only about 3mm of wood left between the metal plates. Nevertheless it feels strong enough to take some rough handling. 

The hole for the joystick is just the right size and joystick moves quite well. Eventually the top around the joystick will be filled in and smoothed out.

Building a wooden box

600 x 900 x 12mm MDF

2 hinges

2 magnetic clasps

Rubik's Cube Simple Algorithms

Here's another Rubik's Cube cheat sheet I put together for the beginner's solving method.

This one also contains miscellaneous algorithms such as:

• Swapping two adjacent edges and two corners at the same time (see (6))
• Swapping opposite edges (or diagonally opposite corners depending on how you look at it) (see (7))
• Rotating two diagonally opposite corners without disturbing anything else (see (8))
• Flipping two opposite edges without disturbing anything else (see (9))

Solving a 3x3x3 Rubik's Cube with Minimal Memorisation

Back in the 80's I bought a Rubik's Cube and a booklet on how to solve it. I've long since lost the originals but I still own a cube. I got interested in learning to solve the cube again recently because my nephew impressed me with how fast he could do it. I thought if I practiced, maybe we could compete.

So I searched for a simple method that wouldn't take too much memorisation. I found a web page (which I since lost track of) that listed some rules for solving the cube in about 7 stages and it had only one algorithm to memorise for each stage. This method is nothing new and I don't take any credit for it. I've since found that the method is also described at rubiks.com.

The method builds up the cube layer by layer and does the final layer in two distinct stages. In the first stage the final face is put in place without worrying about the final positioning of each cube. That means that if you looked at the final layer from the top, it would look like it was done but its edges would be all over the place. The final stage involves keeping the colour of the face correct but moving the cubes around to put the edges and corners into the right places and thus finish the cube.

The web page I originally found didn't have any information about how to apply the algorithm at each stage; it just had some pictures that said "follow the steps and match the picture". Without any guidance I found it quite hard to solve the cube consistently.

So I investigated what each algorithm does and came up with the following instructions for solving the 3x3x3 cube. The little diagrams I use to describe the moves depict a side-on view of the front side of the cube and might take some getting used to. The cube pictures show the view from the top of the cube with the bottom of the picture representing the front side of the cube. The only exception is in step 3 which shows a side-on view of the front of the cube.

There are no instructions included for solving the first layer but it's not too hard to work it out.

So, no new information here but I had fun making this guide. Maybe it will be useful to someone.