CNC Machining Tolerances: Standard vs Precision Explained

±0.1 mm and ±0.01 mm both read as tight on a drawing, yet they machine and cost like different worlds. The gap between a standard tolerance and a precision tolerance is not macho ambition; it marks a real shift in machine capability, inspection method, and scrap rate. Drawing to the right band keeps the part on time, while calling every feature to ±0.01 mm means paying for precision on dimensions that never needed it. This guide separates the two bands and shows which features genuinely require tight specs.

What a standard tolerance actually means

Standard tolerance forms the shop’s default band when the drawing states nothing specific. Most shops default to something like ISO 2768-m, which gives medium accuracy: on a 50 mm feature that lands near ±0.1 mm, and on a 500 mm feature closer to ±0.5 mm. This territory suits a well-maintained 3-axis mill or lathe.

Parts hold this band without temperature control, without CMM verification on every dimension, and without special fixturing. Brackets, housings, and covers live in this range. The ISO 2768 guide lays out the fine, medium, and coarse bands buyers can choose from.

What precision tolerance demands

Precision tolerance occupies the ±0.005 to ±0.025 mm band, and it carries overhead. It needs a machine holding repeatability well inside the asked tolerance, which usually means a compensated, thermally stable center. It requires inspection on a CMM rather than a caliper, because a handheld tool cannot resolve a 0.01 mm callout.

It often means the part measured at a controlled temperature, since an aluminum block moves about 0.02 mm across a 50 mm span between 15 °C and 30 °C. Capability on these features should sit around a Cpk of 1.33 or better, and the shop has to prove it. The difference between commercial and precision tolerance rests exactly on this equipment and inspection stack.

The cost boundary between the two bands

Tolerance and cost do not rise together; they accelerate. Loosening a feature from ±0.1 to ±0.05 mm may add a finishing pass, while moving from ±0.05 to ±0.01 mm adds a CMM run, thermally stable fixturing, and a slower feed.

The rule of thumb holds that each time tolerance halves, the cost on that feature climbs a step, and scrap risk climbs with it. A shop quoting a part with ten ±0.01 mm features at the same price as one with two is either betting on luck or quietly planning to inspect lightly. The tolerance versus cost guide shows how the curve bends upward.

Where to draw the line on a drawing

The money-saving move splits the part. Mark only the features that truly need precision: the bearing bore, the mating datum, the seal face. Leave everything else at the standard default.

A bore locating a shaft and the face it seats against may justify ±0.01 mm; the outer profile, mounting holes, and cable cutouts almost never do. Group the tightest features so the machine cuts them in one setup, which protects the datum relationship between critical surfaces.

Where precision tolerance gets wasted

Over-tolerance proves just as expensive as under-tolerance. A cover held to ±0.01 mm when it only needs to look neat costs a CMM check on a feature that will never carry load. A bolt-hole pattern marked ±0.01 mm instead of the standard ±0.1 mm forces the shop to fixture and verify each hole, turning a fast operation into a slow one.

Look at every tight callout and ask what happens if it moves 0.1 mm. If the answer is nothing, the tolerance functions as decoration. Engineers tighten callouts out of caution, but the shop must meet every number on the drawing, so that caution becomes machine time and inspection time on the buyer’s invoice.

How the measuring gauge sets the band

A tolerance remains only as good as the gauge verifying it. A caliper resolves about 0.02 to 0.03 mm, so a ±0.01 mm callout checked with a caliper stays unverifiable. That is why precision features move to a CMM, which resolves 0.001 mm but costs more time per part.

When a gauge cannot read the called-out band, the inspection record cannot defend the dimension, and the part risks rejection even when the geometry itself sits within spec. Matching instrument resolution to tolerance band keeps inspection credible across both standard and precision features.

Tolerance stacking and datum strategy

Add tolerance stacking to the measurement picture: a chain of three loose features can drift the same distance as one tight one, so it sometimes proves cheaper to widen a single callout and control the stack with a datum strategy. When the tolerance budget runs tight across several mating features, ask the shop to review the stack-up rather than tightening every number.

That review often frees enough tolerance to drop the part back into the affordable band. Grouping critical features in one setup preserves datum relationships and prevents accumulated error from shifting mating faces out of alignment during assembly.

Matching tolerance to fit type

A useful rule reserves the precision band for fits that actually engage. A clearance hole bolting to a mating part can live at ±0.1 mm; a slip fit bore wants ±0.01 to ±0.02; a press fit or bearing seat wants ±0.005 to ±0.01. Flatness on a sealing face justifies a tight callout, while flatness on a decorative flange does not.

Read the full bands in the CNC tolerances guide before locking the drawing, and ask the shop which callouts they would tighten or relax during a DFM review. That conversation usually locates the tolerance that was costing money for no functional reason.

⚙

Ready to Start Your CNC Project?

Get a free quote within 24 hours. ISO 9001 certified, 20+ years experience, low-volume friendly from 1 piece.

Get a Quote