A wrong-size CNC part can originate in the drawing, units, tool diameter, compensation, runout, machine scale, backlash, flex, stock movement, or measurement method. The pattern across directions and feature sizes determines which branch to test first.
Measure the Error Pattern First
A wrong-size CNC part can originate in design scale, import units, tool diameter, toolpath side, machine calibration, backlash, deflection, or measurement.
Measure more than one feature in X, Y, and where relevant Z. Compare programmed and actual dimensions, then note whether error is proportional, a constant offset, direction-dependent, localized, or different between internal and external features. One measured square cannot distinguish every cause.
Units and Scaling Errors
Start by comparing the CAD dimension, CAM geometry, posted units, commanded test distance, and finished feature in that order.
Check document units, import units, CAM units, post output, controller interpretation, and any scaling field. A 25.4 relationship strongly suggests inch–millimeter confusion, while an unexplained mirror or rotation points elsewhere. Correct the earliest source stage instead of compensating with arbitrary machine scale.
Tool Diameter and Compensation
An outside profile, inside pocket, and hole respond differently to cutter compensation and tool diameter errors.
A constant difference may come from tool diameter, cut side, stock allowance, or an incorrect zero. An error that grows with size can indicate scale or calibration. Compare at least two programmed lengths before editing steps-per-distance; tool deflection and measurement method can mimic calibration error on one feature.
Constant Offset Versus Proportional Error
Use a simple square, circle, and measured axis move to separate software scale from direction-change or cutting-load effects.
Backlash, loose motion parts, runout, cutter deflection, stock movement, and thermal changes can affect finished size. Direction reversal tests and repeated light passes help separate play from cutting load. Inspect and service only according to the machine's documented adjustment procedure.
Mechanical Play and Cutting Deflection
Do not calibrate steps from one loaded cut until tool deflection, stock movement, runout, and measurement uncertainty have been considered.
Internal pockets and external profiles respond oppositely to a wrong tool diameter or offset direction. Verify whether CAM is cutting inside, outside, or on the line and whether finishing allowance remains. Measure the actual cutter when appropriate instead of trusting an unlabeled library name.
Verify With a Calibration Artifact
Record before-and-after values and keep calibration model-specific; correction without a baseline can move one feature closer while harming another.
Use a calibration artifact with two lengths, a circle, an internal pocket, and an external profile, then measure with a suitable tool and consistent method. Save programmed values, measured values, tool, material, direction, and machine state so the correction can be attributed to a mechanism.
Dimensional Accuracy Measurement Checklist
| Check | Method | Result | Pass/Fail |
|---|---|---|---|
| Design dimension | Record the nominal value and tolerance | ________________ | ________________ |
| Machine-axis calibration | Compare commanded and measured travel | ________________ | ________________ |
| Tool diameter | Measure or verify actual cutting diameter | ________________ | ________________ |
| Runout | Measure at the tool with a suitable indicator | ________________ | ________________ |
| Workholding distortion | Measure before clamping and after release | ________________ | ________________ |
| CAM allowance | Review roughing and finishing stock | ________________ | ________________ |
| Measurement method | Use the same instrument, datum, and temperature state | ________________ | ________________ |
The broader topic belongs in Build a CNC Feeds-and-Speeds Experiment Log You Can Reuse. For a closely related but separate task, use How Bit Diameter and Flute Count Trade Detail for Strength and Measure CNC Dust Collection at the Cutter, Not the Hose.
Questions About Why Your CNC Cuts the Wrong Size—and What to Measure First
Why does a CNC part measure correctly in one direction but not the other?
An axis-specific scale, play, guidance, or measurement-datum problem may be involved. Compare commanded and measured travel separately on X and Y before changing toolpath compensation.
Can clamping make a CNC part the wrong size?
Yes. Clamps can bow thin stock or compress soft material, so measure the part while held and again after release using the same datum and instrument.
How do I separate tool-diameter error from machine calibration error?
Cut a controlled feature with a verified tool, measure machine travel independently, and compare internal and external dimensions. The error pattern helps distinguish scale from cutter or runout effects.
What measurements should go in a dimensional accuracy checklist?
Record the nominal dimension and tolerance, actual measurement, instrument, datum, stock condition, clamp state, tool diameter, runout, axis, and corrective test.
