Imagine sealing a timer today and trusting it to fire a single blip around the year 2075. That is the rough pitch behind a quirky maker project from philipzucker.com: a dodecahedron-shaped object engineered to sit quietly for half a century and then, just once, beep.
The humor is in the anticlimax—a blip, not a fanfare. But getting a cheap microcontroller and real-time clock to sip power for that long is a genuine engineering puzzle. The project leans on low-power tricks that any hobbyist building long-life gadgets should know.
How the build actually works
The core is a stock DS3231 RTC module, the kind available on Amazon for a few dollars. The on-board power LED alone draws around 1 mA, and the goal is microamps. Snip that LED and the module drops to roughly 100 µA in standby—tenfold reduction that turns "dies in months" into "might outlive me."
A QT Py board (SAMD-based Arduino) handles waking up, processing, and driving the beeper. The same LED-removal trick applies. A small custom PCB ties everything together, and lithium chloride cells feed the chips directly without a regulator, since the cells' flat discharge curve stays close to the working voltage of both components.
One bug nearly derailed the project. The RTC wakes the SAMD chip to reset the alarm, but the timing of that handshake was off, preventing the beep from ever firing. When Will exclaimed "what is the point of your formal method bullshit if it can't do this!" he had a point. The team was thrashing on timing constants without a real model of the system—how a lot of hobby code gets written. Even a rough mental model of the wakeup sequence would have caught it.
The 50-Year Beep Playbook
| Technique | Before | After |
|---|---|---|
| RTC module standby | ~1 mA | ~100 µA |
| Remove LEDs | Required | First step |
| Voltage regulator | Used | Bypassed |
| Battery selection | Standard coin cell | Lithium chloride (high capacity) |
Who should care about low-power design
- Remote sensor builders** — garden monitors, wildlife trackers, structural sensors that need multi-year unattended operation
- Single-use device makers** — shipping loggers, warranty indicators, one-time-use actuators
- Anyone sick of changing batteries** — smart home devices, desk gadgets, anything annoying to access
Why this matters for tinkerers
Long-life gadgets force a mindset shift. Instead of milliamps, you think in microamps. Instead of "will it work," you ask "will it outlast the battery?" You start caring about pin states, floating inputs, and whether a voltage regulator's quiescent current matters more than your actual circuit.
The bug story hides a softer lesson. The makers admit they had no model of the timing variables involved—wakeup latency, communication delay, processing time. Thrashing with flags until it works is fine for prototypes, but these are exactly the edge cases that bite you in real deployed systems. Sketching the wakeup sequence before coding it would have caught the no-beep bug.
What to watch next
The team estimates about an order of magnitude of headroom on the battery math for the one-year test beep. First checkpoint: roughly a year after sealing. If that early blip lands, the 50-year window starts looking plausible, assuming solder joints and battery seals hold.
The takeaway is less about building a dodecahedron time capsule and more about a reusable playbook: kill the LEDs, skip the regulator, pick a battery with a flat discharge curve, and treat your sleep current as the headline spec.
Key takeaways
- Desoldering status LEDs is one of the cheapest, biggest wins in any low-power build
- Bypassing a voltage regulator only works when your battery voltage already matches what your chips need
- Lithium chloride cells are worth the cost for projects that must sit idle for years
- Sketch the wakeup timing before you write the code, especially when more than one chip is involved
- A short proof-of-life test on a much shorter timescale validates a long-duration design
Frequently asked questions
What is the fifty year beep project?
A battery-powered dodecahedron device built to sit silently for around 50 years and then emit a single short beep, written up by philipzucker.com.
What hardware does it use?
A DS3231 real-time clock module and a SAMD21-based QT Py board, paired with lithium chloride batteries and a small custom PCB.
How did the makers cut the power draw so low?
They removed the on-board LEDs from both the RTC module and the QT Py, skipped the voltage regulator, and used batteries with a flat discharge curve that match the chips' working voltage.
Why did the prototype fail to beep at one point?
The timing of the wakeup handshake between the RTC and the microcontroller was off, so the alarm never fired. The team lacked a model of the various timing variables involved.
When will we know if the design really works?
The team planned a one-year test beep as a proof of life, with that first checkpoint expected roughly a year after sealing the device.