PRINT TUNING // SMALL FEATURES
Tiny Text, Big Problems: What Finally Made It Printable
The CAD looked fine. The slicer had other ideas. The plastic had even more.
Small text on an FDM print looks deceptively easy in CAD.
You type the words. Extrude them. Slice the model. Done.
Then the printer politely informs you that physics has some feedback.
I ran into this while making some small thank-you pieces using the Kokoro Judo Club Hobart logo. The larger KOKORO lettering wasn't the problem. The tiny JUDO CLUB and HOBART text underneath it absolutely was.
Letters distorted. Some strokes almost disappeared. The H in HOBART seemed determined to become its own experimental font.
What finally worked
The successful print ended up using settings I wouldn't normally apply to an entire model:
0.10 mm layer height · 0.35 mm line width · one wall loop · 40 mm/s outer-wall speed · ironing disabled.
The important part was treating the lettering as a different manufacturing problem from the rest of the object.
The body of the model didn't need those settings. The tiny lettering did.
Smart walls aren't magic
Modern wall-generation engines are incredibly clever, particularly around thin geometry, but “clever” doesn't automatically mean “best for every glyph in a tiny font.”
Sometimes the slicer tries so hard to preserve a shape that the resulting toolpath isn't what you expected at all.
Previewing the sliced toolpath became just as important as looking at the CAD model.
If a letter already looks questionable in preview, sending it to the printer and hoping molten plastic will somehow improve it is probably not a serious engineering strategy.
Although I'll admit I've tried it. More than once.
PROJECT SOUNDTRACK
Clear That Nozzle
There are technical troubleshooting guides for printer frustration. Then there is turning it into a song.
Listen on Suno ↗The lesson
There are really three different designs involved in a 3D print: the geometry you create, the toolpath the slicer creates from it, and the physical object the printer manages to produce.
They are related. They are not identical.
A feature being valid in CAD doesn't mean it's printable. A feature appearing in the slicer doesn't guarantee it'll print cleanly. And a setting that's perfect for 95% of a model can still destroy the remaining 5%.
The final piece wasn't fixed by one massive change. It was fixed by several small ones.
That's probably a fairly good description of 3D printing in general.
← Back to stories