Build a small box that only opens on weekends. It uses:
- A Raspberry Pi Pico for logic
- A DS3231 RTC to keep the day/time
- A 128×64 I²C OLED (SSD1306) to show status
- A 5 V cabinet latch/solenoid lock
- A 3.3 V logic-compatible relay module (optocoupled) to switch the lock’s 5 V power
- Pico reads day/time from the RTC.
- If it’s Saturday or Sunday (I'll define the open window), the Xiao and battery energizes the relay, powering the 5 V lock to open.
- OLED shows status: day/time and “Locked/Unlocked/Next open”.
- Weekdays: relay stays off, lock remains closed.
- Seeed Xiao esp32c6
- DS3231 RTC module (I²C, CR2032 coin cell installed)
- OLED 0.96" 128×64 SSD1306 (I²C)
- Relay module (3.3 V input), 1-channel, opto-isolated, low-trigger OK
- Meccanixity 5 V cabinet lock / electric latch
- 5 V 3 A DC power supply (barrel jack or screw terminals)
- 1N4007 diode for the lock (flyback protection)
- USB-C Trigger Board Module 5V-20V
- Wires, breadboard, productboard, small inline fuse (e.g., 1–2 A), enclosure
- Battery
Why 5 V / 3 A? Gives headroom for the lock’s inrush + relay + Pico. The Pico can also be powered from this same 5 V supply.
- Single 5 V PSU powers everything.
- Pico power: from 5 V via USB or VSYS.
- Relay + lock: directly from 5 V rail (separate path).
- Common ground: tie GND (Pico), GND (relay), and GND (5 V PSU) together.
- Flyback diode across the lock terminals (stripe/band to +5 V).
The relay board’s diode protects the relay coil, not your lock—you still add the flyback diode across the lock.
| Function | Module | Pico Pin (example) | Notes |
|---|---|---|---|
| I²C SDA | DS3231 + OLED (shared bus) | GP4 (SDA) | Both modules on same I²C bus |
| I²C SCL | DS3231 + OLED (shared bus) | GP5 (SCL) | Add 4.7 k pull-ups only if your boards don’t already have them |
| 3.3 V power | DS3231 + OLED | 3V3 | Check OLED VCC is 3.3–5 V tolerant (most are) |
| GND | All modules | GND | Common ground everywhere |
| Relay IN | Relay module | GP15 (any GPIO) | Many opto modules are active-LOW (IN=LOW → ON) |
| Relay VCC/GND | Relay module | 5 V/GND | Relay powers from 5 V rail |
| Relay contacts | Relay NO/COM → Lock + | NO→+5 V to lock, COM→+5 V rail | When ON, lock gets +5 V via NO |
| Lock – (return) | Lock negative | GND | Return to PSU ground |
| Flyback diode | Across lock | Stripe to +5 V | Protects relay & Pico |
Alternate contact wiring is fine (COM to +5 V rail, NO to lock +). Keep it consistent.
- Simple: Unlock all day Saturday & Sunday.
- Windowed: Unlock only Sat 9:00–Sun 21:00.
- Timed pulse: At allowed times, pulse the lock for e.g. 1–2 s (some latches only need a momentary energize).
- Init I²C, RTC, OLED; set GPIO for relay (OUTPUT, default OFF).
- Read RTC → day-of-week + time.
- Decide
allowed = isWeekend(now)(and within open window if you set one). - If
allowed: set relay ON (remember if active-LOW), else OFF. - Update OLED: line1 time/date, line2 state (LOCKED/UNLOCKED), line3 next open time.
- Sleep ~500 ms; loop.
Libraries: SSD1306 driver + DS3231 RTC (Arduino-core for Pico or MicroPython). In Arduino, use Earle Philhower Pico core; in MicroPython, use
ssd1306.pyand simple DS3231 helpers.
- Blink an LED (prove toolchain).
- RTC read time to serial.
- OLED shows a “hello + time”.
- Relay click with test sketch (no lock yet).
- Add flyback diode and connect lock; test momentary energize.
- Implement weekend logic.
- Enclosure build: mount lock & strike plate; add fuse, strain relief.
- Burn-in: 24–48 h test with PSU.
- Diode across lock is mandatory.
- Watch relay trigger polarity (active-LOW is common).
- Keep 5 V path to lock short & thick; separate from signal wires.
- Secure PSU input with an inline fuse and strain relief.
- If the lock heats up, use timed pulses instead of continuous power.
- Manual override 2 phisical keys that will open the box
- Buzzer/LED when open window starts.
- Settings menu on OLED (set open hours, show next open).
- EEPROM/Flash to store user settings.
- RPi Pico, DS3231 module, 0.96" SSD1306 I²C OLED
- 1-ch 3.3 V opto relay module (low-trigger OK)
- 5 V cabinet lock
- 5 V 3 A PSU + barrel jack breakout
- 1N4007 (or 1N5817) diode, fuse holder + 1–2 A fuse
- Wires, breadboard, enclosure & hardware
- Order parts above (if not already).
- Breadboard the I²C + relay without the lock; verify OLED/RTC.
- [x]Add flyback diode, connect lock, and test 1–2 s pulse.
- Implement weekend logic; adjust unlock window to taste.
- Move into enclosure; add fuse and label wiring.
I have completed a lot from the past couple days
- Worked out how to wire and test with a multimeter
- I got the core functionality tested, meaning I was able to hook up my pico, relay, and 5v lock and unlock the lock from a program on the Pico.
- I soldered the Pico, a led, a USB-C Trigger Board, to a prototype board and have it being powered by the usbc and blinking the led.
My USB C is returning 5 volts and after the diode its about 4.75, I don't know if the 5v is "diluted" when being split between the relay and Pico. I'm wondering if I should increase the voltage and have 2 resistors making it 5v for the board and relay.
I am still considering adding a battery so the user can turn it on and off when its time to unlock
I am also dreading creating a 3d print to hold all of this
I was thinking of having multiple physical locks that can unlock the box in the case of an emergency, I think that would be fairly simple to do just have a "false top" that is held on top by 2 locks
- Solder the relay to a protype board
- Look into how to create a 3d print to house all of these components
Happy with my progress from yesterday. I didn't complete a lot so I will have to explain as to why I'm excited.
- Moved my GRD wire on my Pico's prof board to make some room for some header pins
- I'm doing this so I can easily connect some io to the board like my relay, and OLED
- Created a Test case for my Pico and it actually fit
- I never created a 3d model before, I've attempted in the past but who knew designing parts that need to be precise to the mm would be hard.
It took be about 5 hours to just make a plastic bed which has 4 pillars to be inserted into the mount holes on my perf board. The exact design of my perf board was not available so I had to use my "measuring clamps" to get my dimensions and create a model to match.
I then used that model to design a mount to hold it, while doing this I knew that if mis-measured anything it wouldn't work but luckily that didn't seem to happen and I do say I'm pretty proud of my first go.
- What is the purpose of the different types of capacitors
- While researching it is recommended that I add capacitors so my electronics have consistent supply of energy. This makes sense because I did notice flickering for my LED / OLED screen when I did anything like press the button.
- How to safely add a battery, or should I even bother.
- I really really want to add a battery but I don't want to bite off more than I can chew and it could be dangerous if I don't do it correctly.
- I've been learning as I go and that usually works for me but I like my house undercooked so I may want to research before I mess with battery's. I was working on attempting to use another voltage setting on my Trigger Board and intake 9v instead of 5v so I wasn't splitting 5v => 5v & 5v, so I wanted to use a resistor to turn that 9v into 5v and the resistor started to smoke so I'm taking the power a bit more seriously.
- Adjust my 3d print to test thicker mounts
- Adjusted a old 3D print to use on my pegboard
While I'm waiting for my transistors which should arrive today I had time to kill so I decided to practice my 3D modeling and adjust my test mount for my Pico and adjust a old 3d print for my pegboard.
The test mount was better, I increased the base by x1.5 and shortened the top bit which is what goes in the mount hole itself. It was a good thing too because while I was removing it from the previous test print it broke. So I tested if it would break on the new test and it didn't.
The old pegboard print was funny to say the least. I was so focused on getting the hooks the correct distance apart that I didn't focus on the size of the bucket at all which lead to a very adorable little box.
This new one I increased the size of the hook and the bucket, made sure the distance of the hooks were right. This time I actually made it a little too big but that's fine. While i mounted it, one of the hooks cracked so I might look up the best print dimension for that type of stress. Also While I'm thinking about it I set the infill to 12% which is lower then I normally do it so maybe that has something to do with it cracking.
Autodesk is very confusing to me, especially its history aspect. The file that I created months ago was terrible, all of the sketches didn't make sense and didn't match the 3D body at all. When I tried to delete the sketches it would delete parts of the body that I assume were once created by that said sketch. When I work on it again Im going to export it as a model and import it for a fresh file with just the body.
I have updates for my skill set not so much the box project. I've been busy and have been driving 4 hours back and forth from my family's location.
During these development breaks I was bored and looked up 3d prints that were designed to house my components, specifically my SSD1306 because I want to basically copy the mounts and use it in my custom 3d print and why reinvent the wheel.
I came across this Octoprint 3D Printer Status Monitor and thought I want to try this out.
Along the same time and doing research on how to attach a battery without blowing up my board I can across another small SOC board, the XIAO ESP32C6.
This board is small, a lot smaller than the PICO, cheap, and it has a built-in a Li-ion/LiPo charging chip and battery measurement. I bought 3 for around 20 bucks.
So I built this contraption.
This is a SSD1306 and XIAO ESP32C6 encased in that 3d Print.
I was trying to learn how to actually display things other than text, which made me think of sprites. So I got an image of a poke ball, scaled it to 20x20px which turned into.
const uint8_t PROGMEM ball_bitmap[] = {
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x03, 0xfc, 0x00, 0x07, 0x9e, 0x00, 0x0c, 0x03, 0x00, 0x18,
0x01, 0x80, 0x30, 0x00, 0xc0, 0x30, 0xf0, 0xc0, 0x31, 0x98, 0xc0, 0x3f, 0x0f, 0xc0, 0x3f, 0x0f,
0xc0, 0x31, 0x98, 0xc0, 0x30, 0xf0, 0xc0, 0x30, 0x00, 0xc0, 0x18, 0x01, 0x80, 0x0c, 0x03, 0x00,
0x07, 0x9e, 0x00, 0x03, 0xfc, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00
};They represent the pixel data of a bitmap. I will explain it better when I adjust this.
I programmed the Wi-Fi to connect and grab the current time from an NTP server. Which I learned is a server, that is designed to give precise time. After it grabs the time, it will display it and be able to keep time itself.
Giving me a worthless clock that bounces around a pokeball like those old DVD players. Its too small to be useful but was a good project for me.
I learned a couple of things from this side quest.
- How to turn bitmaps into images on the screen.
- Errors are harder to find. Some errors, like connecting it to Wi-Fi, don't spit out errors because they are tied to IO ports.
- Soldering components for ann enclosed space can be frustrating.