CNC Tech for Phone Buttons

• 1
  What the button actually has to do
• 2
  Which button shapes suit CNC, and which do not
• 3
  Where the tolerance budget goes
• 4
  Surface finish and how it changes the fit
• 5
  Where CNC stops making sense for buttons
• 6
  Machining sequence for a side key
• 7
  Material and process fit for phone buttons
• 8
  Choosing between CNC and other processes

A side key, a volume rocker, a power button or a camera shutter key is a small cantilever. The user presses one end, the part pivots, and a dome or tact switch underneath closes. Three things decide whether that feels right: the gap around the key, the travel distance, and the force curve as the dome collapses.

CNC tech for phone buttons starts from that mechanism, not from the drawing outline. If the key sits 0.05 mm too deep in its frame, the press feels mushy. If it sits 0.05 mm proud, it catches on pockets and cases. That is why the tolerance budget is written around the gap, not around the part's overall length.

The frame is usually the harder part. On a unibody aluminum or titanium housing, the button slot is cut into a wall that may be only 1.2 mm thick. A thin wall moves under cutting load. So the slot dimensions and the key dimensions have to be planned as one pair, even when they are machined in two different setups or at two different suppliers.

• 1
  Gap firstSet the nominal gap before choosing the key tolerance.
• 2
  Wall thicknessBelow 1.5 mm wall, expect to slow the finishing pass.
• 3
  Travel0.25–0.40 mm is a common press travel band.

CNC handles the shapes a phone button actually needs. A raised cap with a 0.3 mm chamfer, a concave finger dish, an internal pocket for the dome, a small boss for a retaining clip, a light pipe channel beside the key. All of that can come off one 5-axis setup with no secondary stamping die.

Where it gets awkward is very thin, very long, flat parts. A 40 mm long rocker that is 0.35 mm thick will chatter during the finishing pass. It can be machined, but you need soft jaws or a vacuum fixture, light radial engagement, and a slower spindle ramp. The cost per part creeps up fast at that point.

Shapes with undercuts on three sides also force a decision. Either you pay for a 5-axis cycle with a small ball nose tool, or you split the part into two pieces and assemble. For a volume key, splitting often wins on cost once the annual quantity passes a few thousand. For a low-volume or prototype key, one 5-axis cycle is cheaper and keeps the geometry in one piece.

• 1
  Good fitCaps with chamfers, pockets, dishes and bosses.
• 2
  Watch outLong, thin flats under 0.5 mm thick.
• 3
  UndercutsOne 5-axis cycle below ~3,000 pcs; split above.

On a typical handset key, the total gap between key and frame runs 0.10–0.20 mm. Half of that budget disappears into plating or anodizing thickness, paint, and assembly stack-up. What is left for machining is usually ±0.03 mm on the key outline and ±0.02 mm on the frame slot.

GreatLight machines to ±0.005 mm on critical features, which sounds like a lot of headroom. It is not wasted. That headroom absorbs thermal drift across a run, tool wear between offsets, and fixture reset between the first article and the production batch. A shop that can only hold ±0.03 mm has no margin left for anything.

The features that matter most are the ones you cannot measure with calipers on the bench. The depth of the dome pocket, the flatness of the key underside, the position of the clip boss. Those three control the press feel. Put them in the drawing as datum-referenced features, not as general notes, and ask for a first article report with the actual numbers.

• 1
  Outline±0.03 mm is often enough for the key body.
• 2
  Frame slotTighter than the key, since the gap is shared.
• 3
  Dome pocketDepth controls click force more than any other surface.

Anodizing, plating and powder coating all add thickness. Type II anodize typically grows 10–25 μm per surface depending on the alloy and bath. Hardcoat can add more. If you machine the key to the final nominal size and then anodize, the key grows and the gap closes. Every dimension that touches another part has to be pre-compensated in the CAM model.

Finish also changes the feel. Bead blasting at Ra 0.8–1.6 μm gives a matte, fingerprint-resistant cap. Polishing to Ra 0.2–0.8 μm gives the glossy look but shows scratches quickly on aluminum. For a key that rubs against a case or a pocket lining, a hardcoat or an electroless nickel layer holds up far longer than bare anodize.

Laser marking is the last step and it has a floor. Minimum character height is 1.5 mm. On a 3 mm wide key, that leaves room for roughly two characters or one small icon. Plan the marking area in the CAD model before the toolpath, not after the parts come back.

• 1
  Pre-compensateSubtract coating thickness from machined dimensions.
• 2
  FeelBead blast for matte, polish for gloss.
• 3
  Marking1.5 mm minimum character height.

CNC stops making sense when the part is flat, thin and needed in the millions. A stamped steel key with a plated finish costs a fraction of a machined one at that volume. The trade is gap control: stamping cannot hold the slot-to-key relationship as tightly, so the design has to open up the gap.

It also stops making sense when the geometry is mostly cosmetic. If the key is a plastic cap over a membrane switch and no one will ever feel the edge, injection molding wins on every axis. CNC is for the parts where a person touches metal and notices the difference.

There is a middle ground worth knowing. Cast the key body and machine only the fit surfaces: the outline, the dome pocket, the pivot hole. That keeps tight control where it matters and drops the cycle time to a couple of minutes. GreatLight runs this combination for handset hardware and for automotive switchgear with the same tolerance targets.

• 1
  Millions of flat keysStamping or molding, not CNC.
• 2
  Hybrid routeCast the blank, machine the fit surfaces.
• 3
  Prototype routeOne 5-axis cycle, no tooling cost.