GreatLight CNC Machining Factory logo
CNC Machining
Rapid Prototyping
Materials
Industries
News
About GL

Get Instant Quote

Repair sourcing guide

CNC Machine Card Repairing Service: How to Choose One Without Losing a Week

A dead axis drive or control card stops the whole cell. This guide is written for plant engineers and maintenance buyers who need to pick a cnc machine card repairing service, not a general electronics shop. Read it and you will know which evidence to collect before shipping a board, which questions separate a real repair house from a parts swapper, and when replacement beats repair.

Board-level fault diagnosisSpares sourcingWarranty in writingRepair vs replace call
cnc machine card repairing service
Quick answer

Key takeaways

Send the fault history, not just the boardAlarm codes, axis number and the moment it fails cut diagnostic time in half.
Ask for the repair report format firstA shop that shows a sample report will tell you what was replaced and why.
Warranty length is a proxy for confidenceShort or verbal coverage usually means a board swap, not component-level rework.
Repair wins below roughly 40 percent of replacement costAbove that line, a new or refurbished card with a real spares chain is safer.
Decision table

What a cnc machine card repairing service should be judged on

Score each supplier against these rows before you release a purchase order.

CheckWhat good looks likeRed flag
Fault evidenceYou supply alarm codes, axis number, load readingsShop asks for the board alone, no context
Component-level reworkCapacitors, drivers, opto-isolators replaced and listedOnly whole-board substitution described
Test methodLoad or simulator test under power, printout suppliedContinuity check only, no powered run
Spares chainGenuine or graded equivalents named in the quoteUnnamed 'compatible' parts, no source
Warranty termsWritten period plus rework and refund pathVerbal promise, no document
TurnaroundDiagnosis window stated up front in daysVague 'as soon as possible'
Data handlingNDA available, uploads kept confidentialCustomer programs shared without consent
Where the money goes

Why a cnc machine card repairing service matters to a running plant

A control card, drive card or servo card sits between the CNC controller and the mechanics. It reads motion commands and pushes current into the motor. When one of those boards dies, the machine does not run slowly. It stops, often mid-cycle, sometimes with a tool still in the cut.

The direct cost is the lost spindle hours. The indirect cost is harder to see. A hot job slips, an operator is parked, and the next cell downstream is starved. On a five-axis cell running unattended overnight, a single card failure can wipe out a full shift of output before anyone notices.

That is why repair sourcing is a maintenance decision, not a purchasing formality. Dropping the board at the nearest electronics shop usually costs more in the end. Those shops fix televisions and industrial drives with the same bench, and they rarely have the test rig to run a card under load.

GreatLight runs 127 high-precision CNC machines across three wholly-owned plants, so board failures are a familiar problem here. We repair cards for our own cells and for outside plants, and we quote from the same evidence list: alarm history, axis, and symptoms under load.

Evidence first

What to collect before the board leaves your plant

Most bad repairs start with a thin work order. The board arrives with a sticky note that says 'axis not moving.' The technician has no alarm number, no axis ID, and no idea whether the fault appears at power-up or under load. Diagnosis then becomes guesswork billed by the hour.

Write down the alarm code exactly as the control displays it, including the axis letter or spindle number. Note when the fault appears: during homing, during rapid moves, or only when the spindle reaches a certain speed. If the card overheats, record how long the machine runs before it trips.

Photograph the board on both sides before packing, with the connector labels visible. Note any burn marks, swollen capacitors, or discolored traces. If a fuse blew, say so. If the same card failed twice in a month, say that too, because repeat failures point at a wiring or grounding issue the repair shop cannot see.

Pack the board in anti-static foam, not bubble wrap alone. Static damage during shipping creates a second fault that looks like a bad repair. Label the box with the work order number so the receiving bench can match the board to your notes.

  • 1
    Alarm code and axisCopy the exact text from the control screen.
  • 2
    Failure conditionsPower-up, homing, rapid traverse or load.
  • 3
    Board photosBoth sides, before packing, labels visible.
  • 4
    Failure historyHow many times, over what period.
Rework vs swap

Component-level rework and what it actually means

A real repair shop works at component level. It desolders and replaces the failed electrolytic capacitor, the gate driver, the opto-isolator, or the power module. It reflows cracked solder joints on connectors, which is a common cause of intermittent faults on older drives. Then it tests the board under power.

A parts swapper does the opposite. It pulls a used board from stock, fits it, and calls that a repair. That can work once. It also means you receive a board with unknown service hours and no fix for the original fault. If the root cause was a failing cooling fan or a loose ground, the swapped board dies the same way.

Ask what was replaced and where the parts came from. Genuine manufacturer parts cost more but carry known ratings. Graded equivalents can be acceptable for discontinued cards, provided the shop names the source and the electrical rating. Unnamed compatible parts are a red flag.

Insist on a powered test, not a continuity check. A card can show correct resistance on every pin and still fail once it sees switching current. A load bank or a machine simulator catches that. Ask for the test printout with the return shipment.

Cost check

When repair stops making sense

Repair earns its place when the board is discontinued, when a replacement lead time runs into weeks, or when the repair quote lands well under the price of a new card. As a working rule, repair is the better call below roughly 40 percent of replacement cost, assuming the shop offers a written warranty.

Replacement wins when the card is still in production, when the failure damaged the backplane or adjacent drives, or when the same board has failed three times in a year. At that point the fault is systemic, and another repair just buys you a few more weeks.

Age matters too. A board with leaking capacitors across the whole power section will fail again at the next weak point. If the shop reports corrosion from coolant ingress or heat damage across multiple zones, replace rather than repair.

Consider the spares you already hold. If a spare card sits on the shelf, swap it in, get the machine running, and let the failed board go out for repair without pressure. That removes the rush fee and gives the bench time to do the job properly.

Supplier screen

How to screen a cnc machine card repairing service provider

Start with the questions that a bench cannot fake. Ask for the test equipment list. A shop that repairs servo drives should own a load bank, an oscilloscope, and a simulator for the control families it claims to support. If the answer is vague, the repair is probably a swap.

Ask how the shop handles confidentiality. Machine programs, part drawings, and controller backups often travel with the board. Uploads should be secure and confidential, and an NDA should be available on request. A shop that shrugs at this question is a risk to your customer programs.

Check the quality system behind the bench. ISO 9001:2015 gives you a documented process for incoming inspection, rework records, and final test. For automotive work, IATF 16949:2016 adds traceability expectations that matter when a failed card can stop a delivery line.

Finally, ask for the turnaround in days, not in adjectives. A shop that quotes diagnosis and repair windows up front is used to planning. Vague answers usually mean your board sits in a queue while another job runs.

Workflow

Step by step: how to run a repair job cleanly

  • 1
    1. Capture the fault in writingRecord the alarm code, axis, and the conditions that trigger it. Photograph the board and the machine nameplate before removal.
  • 2
    2. Pull the board with power locked outLock out the cabinet, wait for the DC bus to discharge, and use an ESD strap. Label every connector before you unplug it.
  • 3
    3. Pack and labelAnti-static foam inside a rigid box. Put the work order number and your contact on the outside, not just inside.
  • 4
    4. Ask for a diagnosis before authorizing reworkRequest the fault finding result and a firm quote. Confirm what gets replaced and which parts are used.
  • 5
    5. Agree the test standardPowered test under load or on a simulator, with a printout. A continuity check alone is not acceptance.
  • 6
    6. Fix the root cause on the machineClean or replace cabinet fans, check filter mats, and re-torque ground and bus connections before the card goes back in.
  • 7
    7. Record the outcomeLog the repair date, replaced parts, and warranty end. Track repeat failures on the same axis so you can justify replacement.
FAQs

Repair sourcing questions engineers ask

Can you repair a card if the manufacturer no longer supports it?

Often yes, provided the failed components are still obtainable or a graded equivalent exists. Discontinued drive and control cards are a common repair case because replacement stock has dried up.

The shop should name the substitute part and its rating in the quote. If no equivalent exists and the board is potted or the processor is locked, repair may not be possible.

How do I know the repaired card will hold up?

Look for a powered test under load and a written warranty period. A bench that runs the card on a simulator or load bank catches faults that a multimeter misses.

Also fix the machine-side cause. Overheating, coolant ingress, and loose grounds kill replacement cards just as fast as the originals.

Should I send one board or the whole drive rack?

Send the failed board first, with the fault history. If the fault moves after the repair, the problem may sit in the backplane, the power supply, or the motor cable.

Sending the whole rack costs more in shipping and risks disturbing connectors that were working fine.

What does a repair report usually contain?

A useful report lists the fault found, the components replaced with part numbers, the test performed, and the test result. It should reference your original alarm code so you can confirm the symptom matched.

Ask to see a sample report before you ship anything.

Is there a minimum order for card repair?

At GreatLight there is no minimum order quantity. A single card can go through the same intake, diagnosis, rework, and test path as a batch.

For plants with several identical machines, batching boards from the same model can save shipping and setup time.

How fast can a card be quoted and returned?

We return a quotation and free DFM analysis within 12 hours, and production can start within 24 hours of approval. Machined replacement parts ship in 3–5 days.

Repair turnaround depends on the fault and part availability, so the diagnosis window is stated before you authorize rework.

Send us the board and the fault history

Tell us the alarm code, the axis, and what the machine was doing when it stopped. We will come back with a diagnosis plan and a firm quote, not a guess.

12-hour quote100% inspectionNDA on request

Follow

More from the shop floor

We publish setup notes, tooling trials and inspection data from the factory floor.

FacebookTikTokYouTubeLinkedInInstagramThreadsPinterest

Trusted by engineers and manufacturers worldwide

Tesla Ford Motor Company BYD Auto Denso Magna International Boeing Airbus Medtronic KUKA FANUC