PISIC split keyboard blends ANSIC roots, Grin-inspired layout, low-cost PCB limits
PISIC turns a Grin-inspired split into a 102 mm PCB puzzle. The board’s real story is how layout nuance and fabrication limits meet.

A split that gets interesting in the last few millimeters
PISIC is the kind of board that catches experienced builders for a reason: it does not chase spectacle, it chases fit. Youngmin Lee, also known as yuburoll, has built a 42-44 key row-staggered split with slightly curvy rows, and that small move already puts it in the sweet spot where ergonomic keyboards become conversation pieces. In this corner of the hobby, the debate is rarely about whether split ergonomics work. It is about how much shape you can add before the board stops feeling familiar.
That is why PISIC lands so neatly in the middle of a very specific conversation. It is compact enough to feel deliberate, but not so experimental that it abandons the habits builders already trust. The result is a board that reads like a revision of something people already understand, which is exactly what makes it worth looking at closely.
What the Grin-inspired twist changes
The layout idea behind PISIC is a modification of ANSIC inspired by the Grin layout, and that matters more than the name suggests. In practical terms, this is not a total reinvention of how the keys sit. It is a familiar split geometry with a subtle twist in how the rows and positions are interpreted, the kind of change that experienced builders notice immediately when they compare it to other compact boards.
The name itself is part of the story. PISIC is described as meaning grin in Korean, and it is also framed as a Korean pun on the Grin layout of ANSIC. That is a very on-brand kind of hobby naming, because it signals that the project is part homage, part shorthand, and part inside joke. For people who followed the earlier Grin boards, the connection is immediate.
That earlier Grin lineage helps explain why the layout debate still has energy. The Grin Type-R and Grin RC boards showed that a single concept could support multiple keymap styles, including ANSI, ISO, and JIS, plus single-spacebar and split-spacebar options. PISIC inherits that same spirit of adaptation: use a recognizable base, then adjust it until it better serves the hand, the desk, and the build.
Why the 102 mm line matters so much
The most important number in PISIC may be the least glamorous one: under 102 mm. That is the PCB width ceiling the design is built around, and it is not there by accident. It is presented as the real size limit for a cheap JLCPCB fabrication run, which means the board’s geometry is shaped by the economics of actually getting boards made.
That is the story many hobby projects quietly live inside. A split keyboard can be elegant on paper, but if the PCB grows past the point where low-cost fabrication stays practical, the design stops being easy to prototype. JLCPCB’s own materials lean hard into fast, affordable PCB prototyping, so the under-102 mm target becomes a real-world constraint rather than an abstract spec line. PISIC turns that constraint into part of its identity.
For builders, that makes the project unusually useful. It shows how much ergonomic ambition can survive once manufacturing costs enter the room. Instead of treating production limits as a compromise hidden behind the scenes, PISIC puts them at the center of the design, which is exactly where they belong for anyone trying to keep a custom keyboard both clever and affordable.
Hardware choices that keep it approachable
PISIC does not just rely on layout to carry the concept. It uses MX switches and hotswap sockets, which keeps the board in the friendly, buildable lane that a lot of custom keyboard projects aim for but do not always reach. Add a Pro Micro footprint, and the board becomes even easier to place within a familiar ecosystem of parts and firmware expectations.
The PCB also uses reversible routing, a detail that matters more than it sounds like it should. Reversible routing helps simplify orientation and make the board more forgiving during assembly, which is a quiet quality-of-life win for anyone soldering or populating a compact split. Optional underglow adds the visual layer that many modern boards now treat as part of the experience, not just decoration.
The case design follows the same pragmatic logic. PISIC uses a 3D-printed case with a 3D-printed plate screwed into the top case, which keeps the build grounded in parts many hobbyists can produce or source without chasing exotic materials. There are also provisions for a wireless build, although that wireless variant has not yet been tested. That caveat matters, because it shows the project is prepared for true wireless experimentation without pretending the final word has already been written.
How PISIC fits the split-keyboard map
PISIC sits inside a split-keyboard scene that has become broad enough to support real comparison, not just novelty. ZMK defines split keyboards as setups with two or more physical parts, each with its own controller, communicating as one keyboard device. That framing helps explain why small boards like PISIC continue to matter: once you accept the split as a platform, the question becomes how far the layout can be refined without losing the practical advantages that made splits attractive in the first place.
That is also where the wireless conversation fits. nRFMicro describes itself as a DIY drop-in Pro Micro controller replacement for converting wired Pro Micro-based keyboards to true wireless, which makes PISIC’s wireless-ready provisions feel very current. It is the sort of detail that tells you the board is being designed with the next build path in mind, not just the first one.
In the end, PISIC is compelling because it understands the real tension in ergonomic keyboard design. The best ideas are rarely the most radical ones; they are the ones that survive layout debate, PCB limits, and assembly reality all at once. PISIC keeps the Grin inspiration visible, keeps the board small, and keeps the cost envelope believable, which is exactly how a compact split earns a place in the build queue.
Know something we missed? Have a correction or additional information?
Submit a Tip

