Flatkeys60
flatkeys: the logbook
total time spent: ~50 hours of CAD pain, keyboard gain, and very little actual typing
summary: the dream? a clean, minimal keyboard that looks like a boutique board but was made in my bedroom with too much caffeine and questionable precision.
📋 Progress Table
| Date | Title | Time Spent | Summary |
|---|---|---|---|
| June 8 | The Spark | 2 hrs | Sketched layouts, landed on a TKL-ish vibe |
| June 9 | CAD Panic Begins | 3.5 hrs | Fusion 360 and my patience fought hard |
| June 10 | Plate Work but Make It Precise | 3 hrs | Flex cuts, stabs, and scissors got involved |
| June 11 | Wall Thickness Dilemma | 3 hrs | From 2mm noodles to 4mm tanks |
| June 12 | YouTube Research = Progress (kinda) | 2 hrs | Watched 5 builds and learned about magnets |
| June 13 | Layered Plate Experiments | 4 hrs | Laser cut prep, kerf chaos |
| June 14 | Mounting Mayhem | 3 hrs | Switch fit tests, low-pro plate abandoned |
| June 15 | Material Wars | 1.5 hrs | Reddit fights over POM vs FR4 vs acrylic |
| June 16 | Snap Fit Adventures | 4 hrs | Designed screwless base, alignment trials |
| June 17 | Overengineering Top Plate | 3 hrs | Rethought magnets, stashed the idea for later |
| June 19 | The Bom Bom | 2.5 hrs | Connector datasheet deep dives |
| June 20 | Case Redesign ft. Insecurities | 3.5 hrs | Fixed angles, tolerances, ego |
| June 21 | Final Stretch (lowkey tired) | 3.5 hrs | Final exports, last Fusion cleanup |
| June 22 | Submission Chaos | 3 hrs | Everything zipped, clicked submit half-asleep |
june 8: the spark
2 hours
Was just chilling and saw a low-profile keyboard on Instagram that looked clean af.
One Pinterest scroll later, I had 9 screenshots, 2 moodboards, and a burning desire to design one.
Opened a new folder called flat_af_keys because naming is everything.
Sketched some cursed layouts some with knobs, some with too many keys, one was just arrows and a smiley face.
Settled on a TKL-ish layout without the numpad. Compact but not annoying.
june 9: cad panic begins
3.5 hours
Jumped into Fusion 360 like I knew what I was doing. I did not.
Modeled the top plate and a basic bottom shell. Key spacing was off by a few mm at first and it triggered me deeply.
Fixed it, learned how to do fillets properly now the curves are curving.
Printed a paper layout to check spacing and mock-typed on it like a nerd. Felt real.
june 10: plate work but make it precise
3 hours
Worked on the plate design. Tried adding flex cuts but it just made it look like swiss cheese.
Added standoff holes and support for screw-in stabs (finally).
Printed another paper layout and cut it out like it was arts and crafts time.
Felt way better than expected.
june 11: wall thickness dilemma
3 hours
Tried 2mm, felt flimsy. 3mm, decent. 4mm, chef’s kiss.
Modeled in the USB Type-C breakout barely fit, but it fits.
Lots of zooming and squinting. Changed wall profiles thrice.
Slight redesign of the outer shape to feel a bit sleeker.
june 12: youtube research = progress (kinda)
2 hours
Watched keyboard teardown videos while eating lunch. Called it research.
Learned about keyboard tilt, gasket mounts, and why magnets are scary but cool.
Added a slight tilt to the bottom shell. Contemplated making it magnetic.
Decided not to future Milan can deal with that.
june 13: layered plate experiments
4 hours
Redesigned plate for laser cutting: layered style.
Kerf offset math hurt my brain. My first SVG export looked like pixel soup.
Adjusted and re-exported everything again.
Finally got clean curves and spacing. Feels sharp now.
june 14: mounting mayhem
3 hours
Switch mounting tolerances are no joke.
Clip-in mounts worked okay-ish, plate-mount was a disaster.
Added mini tabs for support not pretty, but functional.
Low-profile version? Bad idea. Deleted it. Good riddance.
june 15: material wars
1.5 hours
Reddit thread got me deep into FR4 vs POM vs Acrylic vs “just vibes.”
Decided on acrylic for now cheap, easy, and good RGB diffusion.
Even if the board sucks, it’ll glow.
june 16: snap fit adventures
4 hours
Reworked the bottom case to be completely screwless.
Snap fits and alignment pins galore.
Modeled a dummy PCB just to make sure nothing collides.
Debated doing a daughterboard. Didn’t. Not my circus (yet).
june 17: overengineering top plate
3 hours
Felt like doing something dramatic so I redid the top half.
Redesigned curves, internal supports, and added a magnet mount idea (again).
Decided to keep it screw-fit for now but left magnets in the CAD for future me.
june 19: the bom bom
2.5 hours
Started making the BOM. JST connectors are a different kind of pain.
Compared like 5 datasheets before picking one that fits.
This felt more like rocket science than keyboard making.
june 20: case redesign ft. insecurities
3.5 hours
Showed renders to a friend who roasted the side profile.
Redesigned the case that night flatter, cleaner, less chonk.
Fixed tolerances and the seam line. Case went from “meh” to “maybe I’d buy this.”
june 21: final stretch (lowkey tired)
3.5 hours
Cleaned up the Fusion project like, properly.
Every unused sketch got deleted. Every body named.
Final render time. It felt... done. Almost.
june 22: submission chaos
3 hours
Zipped the folder, double checked every STL and DXF.
Filled the Hack Club Highway form.
Was half-asleep when I hit submit.
Immediately regretted not proofreading but whatever, we ball.
final vibes
This project made me fall in love with design and hate tolerances.
Would I do it again? Absolutely. But next time, I’m getting laser eyes and a CNC machine.
flatkeys started as a meme name and now it’s kind of iconic in my brain.
Let’s see if it types.
total time spent: 50 hours