Used CNC Machine Electrical Inspection Checklist — 30 Checks Before You Pay the Balance-CNCShare

The most expensive sentence in a used equipment deal is: “It powers on and it runs, so it’s fine.”

Powering on only proves it isn’t dead. It does not prove it will work. Especially on the electrical side — servo drives, spindle, tool changer, lubrication, safety circuits. Any one of them carrying a latent fault turns into a repair invoice plus months of downtime once it reaches your floor.

This guide breaks a field inspection into 5 stages and 30 checks. Print it out and tick them off one by one.

Note: this article covers inspection method and decision logic only, not machine-specific parameters. Wherever a specific value is needed, refer to the manual that came with the machine. Items marked verify against the manual should not be settled from experience alone.

1. Check the history before you check the machine (5 checks)

A machine can be cosmetically restored. Its history cannot be faked.

1. Nameplate, system model and drive nameplate must agree

Compare the nameplate model and serial number with what the control reports. Check three places in particular: the nameplate, the machine model setting inside the parameters, and the drive nameplate. If they disagree, the control or the drives have been swapped.

2. Read running time and part count

Most controls can report cumulative powered hours, cumulative parts cut, and spindle hours. “Old machine, very few running hours” is either genuinely idle time or a cleared counter.

3. Go through the alarm history

Look at the alarm log for historical alarms. What you want: repeated occurrences of the same alarm, servo overload, spindle alarms, battery alarms.

Alarm logs can be cleared. This item is evidence, never a conclusion.

4. Look for evidence the parameters were touched

Check parameter backup dates and modification traces. If the timestamp is obviously later than the machine’s last known operating period, someone has been in there recently — which may be maintenance, or may be concealment.

5. Ask why it is being sold

Stopped because of a fault, product line change, or equipment renewal — those are normal reasons. A vague “the owner got out of the business” deserves extra caution.

2. Checks with the power OFF (8 checks)

Do not switch it on yet. What you can see with the power off becomes invisible once it is running.

6. Open the electrical cabinet — smell it first

A burnt or acrid smell is a hard signal. Then look at: dust thickness, oil contamination, scorch marks, water traces, and whether the cabinet seal has hardened.

7. Cables and harnesses

Inspect each one: hardened or cracked insulation, damaged sheath, rodent damage, loose cable ties, and chafed wires inside the drag chains. The drag-chain sections are the worst offenders — wires break at bends.

8. Contactors and relays

Look for pitted, eroded, or welded contacts. Pull the armature by hand: the action should be crisp, never sluggish.

9. Drive cooling fans and filters

A dead fan is the most common latent fault. Fan stops → drive overheats → intermittent alarms, which are almost impossible to reproduce on site but will follow you home.

10. Absolute position encoder battery

The battery holder sits in the cabinet or on the motor side, depending on the machine. Measure the voltage; treat anything below its rated value as “replace” (verify the rated value against the manual).

The classic consequence of a low battery: absolute position is lost on every power-down, requiring re-homing or reference-point re-establishment on the next start-up. This is not a minor fault.

11. Grounding and insulation

Check that the ground wire is actually landed on the cabinet and has not been removed. With a megger, measure insulation resistance — use the acceptance value from the manual, not “looks about right”.

12. Look for swapped components or modifications

Empty slots, jumpered wires, unrecognised retrofit boards. A jumper is the classic “make the alarm go away” shortcut.

13. Transformers and power modules

Look for discoloration from heat and smell. Measure module output voltages after power-up.

3. Power on, but do not run it (6 checks)

Energise, but don’t move anything yet. This stage tells you what the machine thinks about its own health.

14. Power-up self-test and current alarms

Do not hit reset first. Record every alarm number and message, then decide which are “sat for a long time” and which are real.

15. Battery alarms

If absolute-position or battery alarms appear, the battery or the stored position is suspect. Do not clear them on site — ask whether the reference point has already been lost.

16. Servo and spindle enable

Release the E-stop and check whether the servos energise, whether there is abnormal noise, and whether the drive panel shows alarms. Noise at the moment of energising deserves particular attention.

17. E-stop circuit and guard interlocks

Test each one: panel E-stop, pendant E-stop (if fitted), guard door switch, light curtain (if fitted). A bypassed E-stop circuit is the most dangerous and the most common corner cut.

18. Lubrication, hydraulics, pneumatics

Check oil level, pressure and air pressure, and confirm the lubrication pump actually delivers on schedule (observe it in manual mode). Poor lubrication is the slow killer of ways and ballscrews — invisible in a short visit.

19. Coolant and chip conveyor

Run the coolant pump and the chip conveyor under power. These auxiliaries do not affect “it runs”, but they absolutely affect whether you can start production.

4. What only shows up when it moves (7 checks)

20. Jog each axis through its full safe travel

Listen for noise, watch for smooth running. Look specifically for: abnormal noise, stiction, juddering, and backlash.

21. Backlash

Do a reversal test using a limit switch or a dial indicator. Excess backlash is not necessarily mechanical wear — it may be a parameter temporarily adjusted (see section 6).

22. Spindle ramp-up and orientation

Step up from low to medium speed, listening for noise and feeling for vibration. Test spindle orientation (required before tool change). The location of the noise matters — bearing noise, belt noise and motor noise must be distinguished.

23. Tool changer full cycle

Do not just watch one tool change. Run 10 or more consecutive changes and look for jammed tools, dropped tools, and positioning error. The tool magazine is one of the most common problem areas on a used machine.

24. Coolant, chip conveyor and pneumatic interlocks

Run a short program in automatic mode and confirm the auxiliaries follow correctly.

25. Accuracy spot-check

A ballbar or laser interferometer is ideal; without them, at minimum cut a test part (circle, square, step) and check roundness, squareness and size consistency.

26. Run it long

Idle for at least 30 minutes. Many overheat and intermittent-alarm faults only appear under continuous running. Do not skip this stage.

5. Software and files (4 checks) — the most overlooked

27. Is the parameter backup complete?

Confirm you can export parameters, PMC ladder, pitch error compensation, macro programs and tool data. Miss one and every future problem gets harder.

28. Can the backup actually be restored? (critical)

Having a backup file is not enough — you must confirm this backup is restorable. A corrupt file, a version mismatch, or a missing PMC means there is no rescue when you need one.

Run this as a rehearsal. Half an hour before purchase beats three days of downtime later.

29. System media, CF card, and locks

Confirm the system program media is intact, whether it is encrypted, and whether a factory password lock is present. A locked machine with an unknown password is a half-usable machine.

30. Documentation

The random manual, electrical drawings, parameter tables and PLC program listing. Electrical drawings matter most — without them, every future line fault takes twice as long to trace.

6. Common concealment tactics (the most valuable section)

Tactic	What it looks like	How to defeat it
Clearing the alarm log	An implausibly clean history for the machine's age	Cross-check against running hours and test-cut results
Temporarily adjusting parameters to hide backlash	Backlash “just barely passes”	Check parameter modification timestamps; measure at several positions
Removing or bypassing alarm devices	A sensor “isn't there” or its wire is jumpered	Count components against the electrical drawings (checks 12 and 17)
Board swapping or borrowing power	Drive or power module looks “new”	Match drive nameplate against the factory serial number
Removing filters or fans	An oddly clean cabinet that still runs hot	Touch-test every fan by hand (check 9)
Running empty but refusing a test cut	Deflection with “no material available”	Insist on a test cut, even a small part

7. Turning findings into price

Finding problems does not mean you cannot buy the machine. It means turning them into price or remediation terms.

List the findings first: which are “must fix before production”, which are “can run with”.
Cost them: convert the must-fix items into money, including lost production time, and use that as your negotiating basis.
Put it in writing: remediation scope, delivery condition, balance payment terms, warranty coverage.
Hold back a balance: leave part of the payment until the machine has been powered, test-cut and accepted at your site.

Frequently Asked Questions

Q: What if there are no electrical drawings?

A: You can still buy it, but price the risk in. Without drawings, any future circuit fault can take twice as long to trace. At minimum require a parameter backup and the PLC program.

Q: Can I just clear a battery alarm?

A: Not recommended. When an absolute position encoder battery is low, clearing the alarm only hides the problem — the position is still lost on the next power-down. The correct fix is to replace the battery and re-establish the reference point.

Q: Can I sign off if it only ran empty, with no test cut?

A: Very risky. Empty running cannot expose load, accuracy or thermal deformation. At minimum require a small test part and keep the sample.

Q: How important is a parameter backup on a used machine?

A: Extremely. Parameters, PMC, pitch compensation and macro programs are all required, and you must verify that the backup is restorable. That is your only way back when something goes wrong.