Ballbar and Laser Checks: Verifying Used CNC Accuracy

Two cheap tests reveal whether a used machine still holds tolerance, or just looks clean under the shop lights.

Published 2026-09-07 · Data as of 2026-09-07 · Market & data intelligence · Educational, not advice.

A used machine sells on how it looks and how many hours the counter shows. Neither tells you whether it still holds tolerance. Paint hides nothing about a worn ballscrew, and low hours on a machine that was crashed and never realigned mean less than a clean plot from a machine that ran three shifts for a decade.

Two tests cut through that. A ballbar check scores how well the machine moves in a circle, and a laser interferometer maps how accurately it positions along a straight line. Together they turn a guess into a number. Here is what each one measures and how to read what comes back.

What the ballbar actually tells you

A ballbar is a precision telescoping bar with a ball on each end that sits in magnetic cups. The machine runs a programmed circle while the bar records tiny changes in radius. Perfect motion draws a perfect circle. Real machines draw a distorted one, and the shape of the distortion names the fault.

This matters because most parts are made from arcs and blends, not single-axis moves. A machine can position fine in a straight line and still cut an out-of-round bore because two axes are not cooperating.

The plot reads like a fingerprint. Backlash shows up as steps where an axis reverses. Servo mismatch tilts the circle into an oval on the diagonal. Squareness error tilts it the other way. Stick-slip from worn ways or dry rails shows as spikes. Scale errors show as a circle that is simply the wrong size.

The single number that gets quoted is circularity, often reported in microns. Treat that number as a headline, not the story. Two machines can post the same circularity value for completely different reasons, and one fault is a cheap adjustment while the other is a rebuild. Always look at the shape.

What the laser check adds

The ballbar tells you how axes work together. A laser interferometer tells you how each axis behaves alone. It measures positioning accuracy — does the machine actually land where the program says — and repeatability — does it land in the same spot every time you send it there.

Repeatability is the number that should worry you most on a used machine. Positioning error can often be compensated in the control. Poor repeatability usually cannot, because it means something is loose, worn, or inconsistent, and no software offset fixes a bearing that wanders.

The laser walks the axis in steps, both directions, several passes, and produces a curve. A gentle slope across the travel is often a scale or thermal issue that compensation handles. A sharp local dip points to damage in one section of the ballscrew or rail — frequently the part of the travel where the machine spent its life making one family of parts.

Reading the report, not the pass line

Ask for the raw plots, not just a pass or fail summary. A report that says a machine meets a standard tells you little without knowing which standard, which tolerance band, and how the machine was warmed up first. A cold machine and a hot machine can post very different numbers, and a seller who tested cold is either careless or hoping you are.

Note the ambient conditions too. These tests are temperature sensitive, and a plot taken in a controlled room means more than one taken next to an open dock door in summer.

How this shapes what you pay

Accuracy data lets you price the machine on evidence instead of vibe. A clean ballbar and tight laser numbers justify the top of the range for that model, tooling, and controls. A backlash signature or a worn spot in the travel is not automatically a walk-away — it is a negotiating line item, priced against the cost of new ballscrews, bearings, or way work.

The stakes scale with the work. A general-purpose machine cutting loose-tolerance brackets can carry faults a jobshop would never notice. A 5-axis machine or a precision turning center living on tight-tolerance work has no room for a lazy servo, and the same distortion that is cosmetic on one machine is disqualifying on another.

The same logic runs across categories. Circular interpolation faults matter most on contouring work in a vertical machining center, while on a lathe you weigh spindle condition and axis wear differently again. Match the test to how the machine will earn its keep.

Do not let a seller test the machine for you and hand over a summary. Bring your own metrologist, or make the test a condition of sale with the results verified independently. The cost of both checks is trivial against the cost of a machine that cannot hold the tolerance you bought it for.

Next in The Gauge: what a spindle vibration and runout check reveals that a ballbar never will, and why it is the one test most buyers skip.