Custom Silicone Keypads for Legacy Equipment: How to Update or Replicate Obsolete Designs
- Share
- publisher
- Suey
- Issue Time
- Nov 6,2025
Summary
How to rebuild custom silicone keypads for legacy equipment when the drawing is gone: reverse engineering from a worn sample, the six interface details to capture, design changes that cost nothing on a new tool, and how to prove the replacement matches the original.
The machine is still in production. The keypad is delaminating. The OEM stopped supporting it years ago, the drawing was on a server that no longer exists, and the only remaining specification is the worn part in your hand.
That is a workable starting point. Custom silicone keypads for legacy equipment are rebuilt from physical samples every week, and the process is far more repeatable than most maintenance engineers expect, provided the right measurements are taken before the sample is handled to death.
How Do You Rebuild Custom Silicone Keypads When the Drawing No Longer Exists?
The direct answer: measure the sample, reconstruct the geometry as a drawing, then cut a new tool from that drawing. The sample supplies dimensions and layout; the drawing restores what the sample cannot show, which is the original nominal size before wear, the material grade and the legend method.
The mistake is to send the old part to a molder and ask for "the same thing". A worn part is not a specification. Sections are thinner than they were, domes have taken a compression set, and edges are rounded by handling. Copied literally, the new tool inherits the wear.
The right instruction is to measure the sample, compare it against the flat surfaces and untouched features that still carry the original dimension, and then nominalise: rebuild the drawing at the design intent rather than at the current condition.
From Worn Sample to Tooling: The Reverse Engineering Workflow
Five steps, in this order. Skipping the second one is what causes a second tool.
- Document the sample in place. Photograph the installed panel with a scale in frame, including which keys are worn, so the wear pattern is recorded before disassembly.
- Capture the interface, not just the part. The PCB pad pattern, the connector tail position, the light guide and the bezel opening all constrain the new part as much as the keypad outline does.
- Build the drawing. Panel outline, key pitch, key height, travel, web thickness, tail length and legend layout, each with a tolerance.
- Tool and first article. Cut the cavity, mould a first sample, and compare it dimensionally against the drawing, not against the worn part.
- Fit and function test. Install one panel in the machine and run a full shift before releasing the batch.
What to Capture From the Old Keypad Before It Is Lost
Most legacy rebuilds stall not on molding, but on missing information. Six items cover almost all of it, and each one is easier to obtain while the machine is still running.
| Item | How to capture it | Why it matters |
|---|---|---|
| Panel outline and key pitch | Calipers plus a flatbed scan at known scale | Fixes the tool size and the key positions in one step |
| Key height and travel | Height gauge on an unworn key, plus feeler measurement of travel | Sets the dome geometry and the tactile feel |
| Switch type | Look at the PCB under the key: carbon pill, metal dome or shorting pad | Determines whether the new part carries pills, domes or neither |
| Legend method | Scratch test on an inconspicuous key; screen print lifts, laser etch does not | Drives cost and chemical resistance more than the material does |
| Tail and connector | Measure pitch, length and orientation of the flex tail | A 0.5 mm pitch error makes the part unusable |
| Service environment | Cleaning agents, temperature at the panel, UV exposure | Selects the silicone grade rather than defaulting to a general purpose one |
Where Legacy Keypads Fail, and What to Change in the New Design
A rebuild is the only chance to fix the original weakness at no extra tooling cost. Three changes cost nothing on a new cavity and eliminate the most common repeat failures.
1. Thicken the Web Under High-Use Keys
On most legacy panels, every key shares one web thickness. The enter key and the numeric keys take far more actuations than the rest, and they are the first to tear. Modulating web thickness by key usage is a drawing change, not a tooling cost.
2. Move the Legend Off the Wear Path
If the original legend sat exactly where the thumb lands, move it up the key face a fraction and reduce the print area on the highest-use keys. On a rebuild this is free; on a running tool it is not.
3. Specify the Finish and the Grade Explicitly
Legacy drawings often say "silicone, black". That is not a grade. Writing the durometer, the temperature range and the cleaning chemical into the new drawing is what prevents the second failure mode, chemical attack, from replacing the first.
Proving the New Custom Silicone Keypads Match the Original
Acceptance should be a document, not an opinion. Three checks are enough for most legacy rebuilds: a dimensional report against the new drawing, a force-displacement curve compared with a curve taken from an unworn original key, and a full-shift functional run in the machine.
Where several machines share the same panel, keep one new keypad as a sealed reference sample with its measurement sheet. Ten years from now that sample and sheet are the drawing, and the next rebuild becomes a two-week job instead of a six-week investigation. The same preservation logic applies to instrument panels, which is why we document builds on the instruments and meters keymat page.
If the original tool still exists and is simply worn, the alternative is a cavity rebuild rather than a new design, which shortens validation but keeps the original weaknesses. Our guide to RTV and LSR keypad manufacturing explains why the process choice can change the achievable tolerance on a rebuild.
Frequently Asked Questions
Can you make custom silicone keypads from a worn sample with no drawing?
Yes, provided the panel interface is also documented. The sample supplies the geometry and the PCB supplies the switch pattern; together they are enough to reconstruct a drawing, which is what the tool is cut from.
What if the original keypad has no markings at all?
Then the legends become a design decision. Match the machine's existing functions, keep the layout identical for operator familiarity, and confirm the text with the people who run the machine rather than with the manual, which is often out of date.
How many legacy keypads should be ordered at once?
Enough to cover the remaining service life of the installed base plus one reference sample per machine model. Tooling usually has a long shelf life, but the sealing and legend materials should be re-confirmed if the batch is held for years.