A custom-built clock using Nixie tubes — the old-school glowing vacuum-tube digit displays from the mid-20th century — paired with modern electronics (a microcontroller and a precision real-time clock chip) and a PCB designed from scratch. Started June 2025, completed.

Project repo: github.com/fangmike/Nixie-Tube-Clock

What It Is

Nixie tubes are vintage cold-cathode displays that light up individual digits using a glowing orange filament, popular in calculators and lab equipment before LEDs took over. This project takes that old technology and builds a modern, reliable clock around it: a microcontroller keeps time and drives the tubes, a dedicated high-voltage power supply lights them (they need ~170V to operate, far more than typical electronics), and a custom PCB ties it all together into a finished, enclosed product rather than a breadboard experiment.

Why I’m Building It

I’d built a simpler nixie tube display once before and it worked reasonably well, but this time the goal is to do it properly — real research, real engineering decisions, and a finished product rather than a pile of parts from eBay. It’s not a project with some grand mission behind it; it’s about sharpening my ability to design and build things end-to-end, and getting a genuinely old, beautiful piece of technology to work well in a modern build.

Goals

• Reliably display the time
• Set itself without needing a phone, app, or Wi-Fi connection
• Come in a proper enclosure — a finished object, not exposed circuitry
• Stretch goals: swappable tubes, dimmable digits, and eventually maybe internet connectivity, an alarm, or a Pomodoro timer mode

How It Works, at a High Level

• An Arduino microcontroller acts as the brain, keeping track of button input and driving the display logic
• A ** real-time clock chip** keeps accurate time independently, so the clock doesn’t drift or need to be reset constantly (accurate to about a minute a year, with a battery backup so it keeps time even when unplugged)
• A small high-voltage power module steps up standard low voltage to the ~170V the nixie tubes need to glow
• A driver circuit translates the microcontroller’s digital signals into the higher voltages needed to switch individual digits on the tubes
• Everything is being brought together on a custom PCB, designed in KiCAD, rather than a rat’s nest of wires

The Journey So Far

Stage 1 — Research (complete): Researched how nixie tubes are actually driven, evaluated real-time clock options, compared commercial high-voltage power supplies, and worked out a bill of materials. Decided early on not to design a custom power supply — that’s a whole project in itself and good commercial options already exist, so buying one was the smarter use of time.

Stage 2 — Testing (complete): Bench-tested individual pieces of the design before committing to a PCB: confirmed the tubes light up reliably at the right voltage and current, verified the driver transistors work off the microcontroller’s low-voltage output, and learned ESP32-specific PCB design. Also spent time leveling up in KiCAD — learning schematic and layout best practices, and picking up on common beginner PCB mistakes along the way.

Stage 3 — Design (complete): Turned the validated sub-circuits into a finished schematic and PCB layout.

Stage 4 — Build (complete): Assembled, tested, troubleshot, and finished with an enclosure.

Challenges & Lessons Learned

This project has been as much a lesson in how to run a project as it has been about electronics:

• Presentation vs. substance. After finishing an early PCB revision, I realized I’d been spending more energy making things look polished than actually making progress — a good way to burn out. The lesson: build the substance first, polish later.
• Define things before building. Skipping over basic decisions, like the physical dimensions of the tubes, early on made later design work harder than it needed to be. Nailing down the fundamentals up front saves a lot of rework.
• Not every rabbit hole is worth going down. It was tempting to design a custom high-voltage power supply from scratch, but that would have become its own multi-week project and added little value over a proven commercial part — so I let it go.
• It’s okay to slow down. I hit a point where I paused the project for a while after overestimating how comfortable I was with KiCAD. Rather than pushing through in frustration, I stepped back to rebuild the fundamentals properly — a more sustainable approach than brute-forcing it.

Status

Complete. The clock went from research and bench-testing through to a finished schematic, PCB, and enclosure — including a pause along the way to reset expectations and shore up KiCAD fundamentals before pushing through the design stage. Full project files and details are on GitHub: github.com/fangmike/Nixie-Tube-Clock