Key Takeaways
- Spindle runout >0.005 mm or bearing noise = reject or negotiate a significant price reduction (spindle rebuild: $2,000–$8,000).
- Ballscrew backlash >0.02 mm can be compensated in control, but backlash over 0.05 mm usually indicates worn nuts requiring replacement ($500–$3,000 per axis).
- Always run a test cut — a circle interpolation test reveals combined positioning error better than any individual measurement.
- Verify the control has current software and that the manufacturer still supports it — an unsupported control (obsolete Fanuc 0, some 90s-era Mazaks) can be expensive or impossible to repair.
- Ask for the machine's service history. A machine with oil-change records and documented PM is worth more than one with "ran in production."
Before You Visit
Get the machine serial number and model. Look up the manufacturer's published specifications for that model year: rated positioning accuracy, spindle speed range, axis travel, table load capacity. You need these to verify claims and assess whether the machine meets your requirements.
Ask the seller: How many hours? What was it machining (aluminium, steel, castings)? Any spindle or axis crashes? When was the last coolant flush? Is there a service history?
Spindle Checks (Most Critical)
- Runout at spindle nose: Chuck a precision test bar (or use a ground shank in the spindle taper). Place a dial indicator at the spindle face OD. Rotate by hand. Acceptable: <0.005 mm. Reject if >0.010 mm.
- Runout at 150 mm extension: Same test with indicator at the tip of a 150 mm test bar. This reveals bearing tilt errors not visible at the nose. Acceptable: <0.010 mm.
- Bearing noise: Run spindle at low, medium, and maximum speed. Listen for grinding, chattering, or humming that changes with speed. Any bearing noise is a red flag — rebuild required before reliable production.
- Thermal growth: Run spindle at 6,000 RPM for 20 minutes, then measure test bar runout again. Runout should remain within 0.005 mm of cold measurement. Large thermal growth indicates worn or loose bearings.
- Drawbar pull force: Ask if this has been checked recently. Should be within ±10% of manufacturer spec (typically 6–12 kN for CAT40). Low pull force means toolholders can pull out under load.
Axis Motion Checks
- Backlash (X axis): Position to mid-travel. Move +5 mm, then reverse. Record difference between commanded and actual position on dial indicator. Acceptable: <0.015 mm. Note: control can apply backlash compensation — ask current compensation values and compare to measurement.
- Backlash (Y axis): Same procedure on Y.
- Backlash (Z axis): Same on Z. Z backlash typically shows up as Z position error after tool changes — critical for depth accuracy.
- Repeatability: Command 10 returns to the same position from the same approach direction. Measure with indicator. Acceptable: <0.005 mm variation.
- Way smoothness: Traverse full axis travel at low rapid (25% of maximum). Feel and listen for stick-slip (jerky motion that indicates dry or scored ways). Run full travel at maximum rapid — vibration is normal; sudden noise or clunking is not.
- Ballscrew pitch error: The control's pitch error compensation table can compensate for wear-related lead errors. High compensation values indicate worn screws.
Table and Structure
- Table flatness: Place a precision straightedge across the table. Use feeler gauges to check gap along the full length. Acceptable: <0.010 mm per 300 mm. Heavy table sag or dips indicate crashed loads or worn linear guides.
- T-slot condition: Inspect all T-slots for damage (nicks, burrs, pulled threads). Minor damage is cosmetic; structural damage to slot flanges affects clamping reliability.
- Column squareness to table: Use a precision square on the table and check the column (or Z-axis travel) squareness. Acceptable: <0.01 mm per 300 mm.
ATC, Coolant, and Auxiliary Systems
- ATC cycle: Run 20 consecutive tool changes at full automatic. Count any faults, slow changes, or hesitations. One fault in 20 cycles is a yellow flag; two or more is a red flag.
- Tool retention at all carousel positions: Check for missing retention knobs, damaged forks, or wear in the carousel that prevents secure retention.
- Coolant pump and tank: Start coolant. Verify flow is adequate and consistent. Inspect tank for sludge, swarf buildup, and coolant contamination (rancid smell = microbial growth). Budget $200–$500 to flush and refill.
- Through-spindle coolant (if equipped): Verify pressure is within spec (typically 40–70 bar high-pressure TSC). Check for leaks at rotary union.
- Chip conveyor: Run with water. Look for jams, broken slats, and proper discharge at the side.
Control and Electrical
- Control boot time and alarm history: Boot the control and review the alarm history (usually Maintenance → Alarm Log on Fanuc). A machine with frequent axis alarms, estop faults, or servo errors has undisclosed issues.
- Software version: Record the CNC software version. Check with the manufacturer whether it is still supported and what the upgrade path is.
- Servo drive condition: Look inside the electrical cabinet (if permitted). Check for heat marks, burned contactors, or blown fuse indicators.
- Reference position: Home all axes. Note whether they home consistently to within one encoder count. Inconsistent homing on one axis indicates a reference switch, encoder, or ballscrew problem.
Test Cut (Non-Negotiable)
- Circle interpolation test: Program a 100 mm diameter circle at 1,500 mm/min feedrate in aluminium or wax. Measure the resulting circle with a CMM or precision bore gauge. Roundness should be within 0.020 mm for a machine in good condition.
- Hole positioning test: Drill or spot-drill a 5-hole bolt pattern with 50 mm radius. Measure hole centre positions against nominal with a CMM. Total position error should be <0.025 mm per hole.
- Surface finish test: Face mill a 100×100 mm area in aluminium. Measure Ra with a profilometer. A machine with good way condition and spindle bearings should achieve Ra ≤1.0 µm with a standard 50 mm face mill.
- Thermal stability over a 4-hour production run (if possible): Machine repeat parts over a 4-hour period and measure feature-to-feature variation. Thermal growth should stabilise within the first hour; ongoing drift above 0.020 mm/hour suggests spindle bearing or way lubrication issues.
Negotiation tip: If the machine fails 1–2 checks with known repair costs, use those costs to negotiate the purchase price. A machine that needs a $2,500 spindle rebuild is worth $2,500 less — the seller usually knows this and will negotiate.