Buying a Used CNC: 7 Essential Checks Before You Commit

• 1
  What matters most
• 2
  Used machine vs. new machine vs. outsourced machining
• 3
  Why engineers still consider buying a used CNC processing center
• 4
  What to inspect before buying a used CNC machine
• 5
  The real cost of restoring old machines
• 6
  When a used machine is the wrong answer
• 7
  How to evaluate a used CNC processing center in 7 steps

The economics are simple. A used CNC processing center that sold for a six-figure sum a decade ago can change hands for a small fraction of that. If your parts sit in a wide tolerance band and the machine only has to run a handful of setups, the payback can come in months rather than years.

The second reason is capability you could not otherwise afford. A large used machine may give you 4,000 mm of travel or a Ø400 mm rotary table for the price of a small new mill. That changes which jobs you can quote, not just how cheaply you run the ones you already have.

The third reason is schedule control. When a machine sits in your own shop, you are not queuing behind another company's production runs. For maintenance spares and legacy parts, that matters more than the hourly rate.

None of this is an argument for buying blind. The savings are real, but they are conditional on the mechanical condition of the frame and on how long you can keep the control alive.

Start with the spindle. Ask for a taper contact check with bluing, a drawbar force reading, and a runout measurement at the gauge line. Taper contact below roughly 80 percent or runout beyond 0.01 mm means bearings and regrind are in your near future. A spindle rebuild is the single largest line item on most restoration budgets.

Then check the geometry. A granite square and a dial indicator on the table will show squareness between X and Y, and a test bar in the spindle will show spindle-to-table perpendicularity. If the machine has linear guides, look for scoring and check backlash on each axis. Box ways tolerate wear better but need more scraping work.

Pull the maintenance history. Look for spindle hours, ball screw replacement, and any crash record. A machine that has been crashed and repaired can still be good, but you want to know where the repair happened so you can inspect that area harder.

Finally, check the electrical cabinet. Blown drives, scorched contactors, and evidence of rodent damage all point to a machine that sat unused in a damp warehouse. Corrosion on the control boards is hard to reverse.

• 1
  Spindle taper contactBluing check; below about 80 percent means regrind or replacement.
• 2
  Axis squarenessGranite square plus indicator; look for deviation across the full travel.
• 3
  BacklashMeasure on each axis; worn ball screws cannot be compensated away indefinitely.
• 4
  Control healthParameter backup, battery condition, and drive fault history.

The purchase price is only the first number. Add rigging and transport, which for a large machine can run into five figures once you include cranes and permits. Add a foundation or isolation pad if the machine needs one. Add power drops, air, and coolant handling.

Then add the rebuild. New spindle bearings, reground taper, ball screw replacement, way scraping, and a control retrofit are the usual items. A full control retrofit changes the machine's capability but also its cost profile, and it takes weeks of integration and testing before you cut a first good part.

Tooling is easy to forget. A used machine often arrives with a partial magazine. You may need new holders, a pull stud set, and workholding that matches your parts. For a 40-taper machine, that alone can add a meaningful amount.

The right way to judge the deal is total cost of ownership per good part. Take the purchase price, add transport, rebuild, tooling, and the downtime you will absorb during commissioning, then divide by the number of parts you expect to run before the machine needs another major service.

If your prints call for ±0.005 mm on tight-tolerance features and you need that from the first article, a used machine is a gamble. You can reach that band after a proper rebuild, but the rebuild has to be verified with a ballbar test and a test cut before you trust it on production.

If your volume is low and irregular, capital tied up in a machine that idles is worse than paying a higher hourly rate to an outside shop. Prototypes and short runs rarely justify the floor space and the maintenance overhead.

If the control is orphaned and no third party supports it, walk away unless you are already planning a retrofit. A machine you cannot get serviced is a machine you cannot schedule.

If the parts need certification traceability, remember that the machine is only one link. You still need the inspection equipment, the material certificates, and the process documentation that your customer's quality system requires.