Disassembling toys since '91
Surgically implanting a 3D-Touch Capacitive Trackpad in place of the original Analog Joystick.
It’s 2020 and in the pandemic boredom it’s no easy job to survive from creativity suppression. But… when you realize that you own a toolbox with a plenty of electronics inside, it’s not so hard to see the same old things lying on the desk, in a different way. It’s the case of my Nintendo Switch: I never had the occasion to disassemble the Joy-Con controller before. What a wonderful occasion to put hands on its inside and start a new journey.
“Well, let me look at the internet if someone crazy like a horse had the insane idea to replace the good old Analog Joystick with a Capacitive Trackpad” — No results found.
I am used to start sketching ideas on a notepad and here is my first drawing about how I imagined the Joy-Con with a Capacitive Trackpad in place of the original Analog Joystick.
The best among the “Pros” bullet-points is the NO-DRIFT statement. If you are here, probably you already know that some Joy-Con controllers suffer from Joystick Drift. Think about that for a second: no Joystick = no drift.
First of all, you’re not a real nerd if you aren’t fascinated by transparent acid green stuff. So… a couple of keystrokes on Bezos’s website have been typed in order to achieve a new dress for the Joy-Con. Dismounting the shells isn’t complicated, the only difficult part here is actually gently handling a pair of tweezers if you have a big hand.
The Analog Joystick is not a complex entity — actually, it’s a 2-CH potentiometer plus a push-button — and it’s attached to the main board by a 5-positions FPC connector with 0.5mm pitch. I ordered a flat and flexible ribbon cable with the same specs in order to bring the signals out of the circuit.
Then, I re-assembled the Joy-Con and tested it directly with the Switch in order to check that everything worked fine.
Alright, I have to replace an analog potentiometer with a digital one.
I chose to test different kinds of digital potentiometers with different ranges of price and performance alongside with connectors for flat cables (one for the Analog Joystick input, one for the Capacitive Trackpad output). I ordered them online spending few bucks.
Within a week, I received the package from my favorite PCB dealer. With a solder paste spread and a hot-air gun job, the break-out boards were ready.
Now it’s time to perform some tests.
Inevitable to say that an Arduino, a breadboard and a bunch of jumper wires come in help in order to coordinate the ensemble.
The whole testing setup involved:
Among all the digital potentiometers I broke-out and tested, I chose the one that suited most, costed less and that, inevitably, got the job done. The first step was to check if all the components were correctly wired, alimented and responding. I flashed the standard i2c_scanner example into the board and the result was a success.
With few lines of code it was possible to initialize both the DACs and the Capacitive Trackpad on the I2C bus, then I focused on implementing an efficient strategy in order to explore how the digital potentiometers could emulate both the X and the Y axis of the original Analog Joystick.
Since the DACs have a resolution of 12 bits, they accept a value range that goes from 0 to 4095, so I passed it through a loop by steps of 10.
On the Switch console side, I fired up the the Control Stick Calibration utility from the main settings, in order to see if something was gonna happen.
The Capacitive Trackpad is an interesting piece of electronics hosted in a very small and versatile form factor. Surprisingly, it also offers an amazing resolution of 2048 x 1535 x 63 (X x Y x Z) — and yes it has 63 levels of pressure sensitivity (I’ll exploit this feature later in order to digitally emulate the Stick-Button press of the Analog Joystick).
It’s implicit to say that the second step was to implement a decent mapping in order to match the resolution of the potentiometers. Fortunately, the Arduino comes in help with the map method so it was quite immediate to achieve decent results in almost no-time.
About the pushStick(); method call instruction, since the corresponding pin on the Joy-Con side handles a simple close-circuit logic, I could emulate the button press via code.
Cool. The final result of the first prototype looked to me very promising.
Alright. We have a plane here, now we’d let it fly.
The following step I had in mind, was to create an embedded and more portable solution of the first prototype.
I decided to switch from UNO to a PRO MINI with its smaller package and to build a circuit board that could host it alongside with all the connectors and the digital potentiometers. In order to keep the optimization level high, I found that I could substitute the two single-channel DACs with a double-channel one of the same family, saving some space on the PCB.
I fired up my faithful KiCad, I drew the schematics and then placed all the components at their place in a quite obsessive-compulsive (dis)order.
Punctual as usual, the courier brought the green gold alongside with the other components.
Testing with the digital multimeter.
The PRO MINI needed to be alimented: I stole all the milliamps needed directly from the main board, siphoning them from an uncovered solder pad.
The main idea was to stick the board directly to the Joy-Con’s back with a professional double-sided tape, make a hole in the plastic shell and then pass all the cables through it.
The Capacitive Trackpad came out naked from its original package, so I had to design and develop a 3D enclosure by myself. Tinkercad is very handy in this kind of situations: by adding and subtracting a couple of primitives shapes, the job is well done without complications.
After a quick slicing with Cura, a few minutes of waiting time later and the 3D printer gave birth to the new born.
Super. It suited in perfectly at the first shot. I must say that the filament used preserved the mechanical properties very well. Nice job FILOALFA.
Since the power implied in alimenting the PRO MINI relies directly on the Joy-Con’s battery, I decided to add a small switch in order to turn off the implant while not in use, avoiding extra drains.
Safety first: I covered all the exposed electronics with a good piece of Kapton tape, in order to don’t short-circuit anything while your hand’s sweating.
And here we go. That’s how it looks in the end. It’s still portable, so it is possible to use it attached or in wireless mode.
Let’s see it in action.
Enjoy the world’s first Nintendo Switch Joy-Con controller with a 3D Capacitive Trackpad.
Doing things more seriously: in my plans there’s the idea to miniaturize the whole circuitry in order to create a smaller PCB that could suit directly inside the Joy-Con controller in place of the Analog Joystick small box that’s approximately 19 x 16 x 4 mm in size.
As you can see, the job is already ongoing.
I am aware that this could arouse interest in a small niche of fanatics, modders and gamers but its HARDWARE stuff, and it costs. This project implies a good amount of work that I’ll be happy to do with your support.
If it’s something that market asks, I would be happy to co-finance the development of the first small batch that you could potentially buy on Tindie in the incoming future.
This is explicitly a call to action. If you want to make this Joy-Con Touch Mod-Kit real, donate.
Previously published at https://email@example.com/how-i-hacked-nintendo-joy-con-controller-8ac22d75b0b8
Level up your reading game by joining Hacker Noon now!