Tamagotchi Breadboard Journal

Session 1 | March 21 | ~45 min

Getting Started: Why a Tamagotchi?

I want to build a polished breadboard prototype of my Tamagotchi project. It's a project I already care about, so it felt like a natural fit for making something real and interactive rather than just a basic LED circuit. The kit gives me enough components to actually replicate the core Tamagotchi experience: a display, input buttons, a buzzer, and some sensors for ambient awareness.

ESP32 Schematic Symbol

First up: getting my ESP32 into KiCad. I couldn't find the exact schematic symbol for my specific board on SnapMagic or Ultra Librarian, and Tinkercad doesn't support the components I need, so I had to make my own symbol from scratch using a reference image of the board pinout.

Then disaster struck. I lost all my progress after a KiCad crash and had to restart the symbol from scratch.

After the redo and referencing the pinout image again, I finally finished it:

Time: ~45 min (including the redo after the crash)

Session 2 | March 21 | ~30 min

Adding the OLED Display

I want to be able to see my pet, so the next step is wiring up an OLED. I found my specific display via reverse image search: a 1.29" 128x64 I2C OLED.

I checked the datasheet to figure out pull-up resistors. The datasheet doesn't specify internal pull-ups, which means I need external pull-up resistors on the I2C lines (SDA and SCL).

My available resistor values were 220 ohm, 1k, and 10k. The recommended value for I2C pull-ups is typically around 4.7k, so 1k is the closest I have and will work fine.

I later confirmed via this post that the ESP32's internal pull-ups are around 45k, which is way too high for I2C. Such high resistance limits current too much for reliable signaling, so external pull-ups are definitely the right call here.

Per the ESP32 Arduino pin definitions, the default I2C pins are:

• SDA: GPIO 21
• SCL: GPIO 22

Time: ~30 min

Session 3 | March 21 | ~5 min

Adding Buttons

Tamagotchis have buttons, so I need those. I went with an active-low layout, which is the most common approach:

• One side of each button connects to GND
• The other side connects to a GPIO pin with the ESP32's internal pull-up enabled
• This means the pin reads HIGH at rest and LOW when pressed

Why active-low?

Active-low means the button is "active" (pressed) when the signal is LOW (0V / GND). The internal pull-up resistor holds the pin HIGH when idle. This is clean and simple, no extra resistors needed, just the button and ground.

Time: ~5 min

Session 4 | March 21 | ~5 min

Adding a Buzzer

Tamagotchis make sounds, and I have a buzzer available. I added one from a similar LCSC listing. Wiring is simple: one pin to a GPIO output, the other to GND. The GPIO will drive it with a square-wave signal to produce tones.

Time: ~5 min

Session 5 | March 21 | ~20 min

3D Assembly in Onshape

I wanted a visual of how all these components actually fit on the breadboard before physically placing things. I found an 830-hole breadboard model on GrabCAD, imported it into Onshape, then did the same for the other components and assembled everything together.

This was useful for two reasons: it helped me see whether component spacing would be an issue, and it surfaced the interesting question of whether an ESP32 even fits on a standard breadboard. After testing, a bent wire bridges the ESP32 across the gap perfectly and still makes contact on both rails.

Time: ~20 min

Session 6 | March 21 | ~1 hr

Adding the Photosensitive Resistor Module

I thought it'd be a fun mechanic if the Tamagotchi gets tired or sleepy when it's dark, and the photosensitive resistor module is perfect for this. Finding usable documentation took surprisingly long. I finally found the wiring info here.

The wiring itself was simple. The KiCad symbol was not.

I ran into a really frustrating bug: the parent symbol was named photomod, but the sub-symbols were named TFT_MSP1803_0_1 and TFT_MSP1803_1_1. In KiCad, sub-symbols must match the parent name exactly, which I had to discover through trial and error. This caused errors that nuked my schematic progress again.

My workaround: instead of creating a new schematic from scratch (which kept breaking), I imported a working schematic as a base and built from there.

Time: ~1 hr

Total time so far: ~2h 45min