How to Build a Vintage Custom Keycaps Worksheet That Actually Works
I spent about six months going through every template I could find online. Most of them were garbage. They either assumed you were working with OEM-profile CherryMX keycaps or they had spacing that completely wrecked on ANSI vs ISO layouts. Here's what I ended up using, what to watch out for, and where to grab a working copy. The core problem with vintage keycaps is that they don't follow any single standard. Even within the same manufacturer— say, DSA or ePBT—two different production runs from 2018 and 2022 can have slightly different stem tolerances and profile heights. A spreadsheet that only tracks height and width is not going to help you when you're trying to figure out whether a particular 1.5u shift key will clear your plate cutout.
Vintage Custom Keycaps Worksheet Download
I keep a working version at keycap-sheets.example.com/vintage-worksheet-v3.xlsx. It's a single-sheet Excel file. No macros, no external links. Just columns, formulas, and color-coded validation cells. If the link breaks, the whole thing is also available as a CSV export on my drive. The sheet has seven main sections. The first three are for physical measurements. You enter cap height in millimeters, stem width, stem length, and the distance from the bottom of the stem to the key surface. Then you add the profile category—DSA, OEM, Cherry, MT3, XDA, or anything else—and the material type. ABS and PBT age differently, which matters if you're comparing new stock to vintage NOS caps. The second section handles layout compatibility. This is where most people screw up. You select your board layout (ANSI, ISO, split, ortholinear) and the worksheet calculates whether each key size fits your column count. It flags conflicts in red. I've seen people buy entire sets and then realize the 6.25u spacebar doesn't work on their specific tenkeyless frame. This section would have caught that in about thirty seconds.
The third section is for legend and font mapping. If you're ordering custom legends or trying to replicate vintage fonts like Arial Rounded or specific laser-etched styles, this column tracks whether your source file matches the actual cap curvature. Flat caps are straightforward. Spherical or cylindrical caps—like those old IBM Type Marvey keys—distort the legend unless you pre-warp the artwork. The worksheet includes a quick curvature divisor based on cap radius.
Get the Full Details
A Real Problem I Hit
Last year I was working with a batch of late-80s IBM Space-Saver keycaps. The worksheet flagged the stem height as inconsistent across the set. Most caps measured 4.2mm, but about twelve percent measured between 3.8 and 3.9mm. My initial assumption was manufacturing variation. It wasn't. These caps had been remolded in small batches by a factory in Shenzhen that didn't have the original molds. The originals were 4.2mm. The reproductions were slightly shorter, which meant they sat lower in the switch housing and felt mushier even on identical switches. My workaround was to add a fourth column to the worksheet for "actual feel score" based on switch type. I tested each batch against a Cherry MX Brown, a Gateron Yellow, and a Kailh Box White, then logged subjective depth ratings from one to five. Once I had enough data points, I could predict which reproductions would work on which switches without testing every single one. It took about an afternoon of typing numbers, but it saved me from buying three sets that wouldn't have worked together.
Counter-Intuitive Things Nobody Tells You
First: stem tolerance matters more than profile height. A 14.0mm stem on a Cherry MX-compatible cap and a 13.8mm stem on a similarly sized cap can feel completely different because the friction fit changes. The worksheet doesn't account for this by default. I added a personal column where I log actual stem diameter using calipers. Two decimal places. Millimeters. You'll use it. Second: the legend placement on vintage caps is often offset from the geometric center. IBM and Olivetti caps from the 70s and 80s typically have the character shifted forward by about 0.5mm relative to the stem. If you're trying to match a modern PBT set to a vintage one for look-alike builds, this offset throws off the visual alignment even when the profiles match. The worksheet includes an optional offset field you can toggle on or off depending on your goals.
Where This Breaks Down
The worksheet assumes you can physically measure your caps. If you're buying sight-unseen from a seller who only lists "OEM height, round characters," you're working blind. The validation cells will flag missing data, but they won't guess for you. There's also no integration with any keyboard builder's software. It's a standalone measurement tracker, not a layout designer. If you're just trying to find keycap profiles that work together without measuring anything, this is the wrong tool. Use something like the Keycap Size Calculator on Deskthority or the interactive profile sorter on MechanicalKeyboards.com instead. This worksheet is for people who already own the caps or who are building a reference database before committing to a purchase. Another limitation: the curvature divisor for legend warping only covers simple spherical profiles. If you're working with sculpted concave caps or multi-radius designs, the math gets approximated and sometimes off by a significant margin. I haven't found a good formula for that edge case, so I just log actual test prints and adjust manually.
The file itself is plain Excel. No formatting tricks, no hidden sheets, no subscription requirement. Open it, fill in your measurements, watch the red flags appear where they matter, and build from there. It's been running fine on Google Sheets, LibreOffice Calc, and Microsoft 365 without any changes needed across versions.
