FANUC Servo Amplifier: Repair or Replace? A Decision Table and Four Cost Drivers-CNCShare

When a servo amplifier alarms, the first question on site is: “Repair or replace?”

Most people compare two numbers — the repair quote against the new-part price. That is often the wrong call. Most of what you actually pay never appears on the quote.

An amplifier repaired for a few thousand, refitted, and then failing again two weeks later with ten days of downtime — versus a new unit for twice that, installed the same day. Which one costs more? The answer is obvious, yet at the moment of decision people only see the quote.

Note: this article covers decision logic and cost structure only, not prices, which vary enormously by machine model and sourcing channel. Where specific values matter, refer to the machine manual and actual quotations.

1. First: is this amplifier even repairable?

Before discussing cost, answer a more basic question — is it worth repairing at all?

1. The alarm type largely determines repairability

Auxiliary faults (dead fan, aged electrolytic capacitors, oxidised connectors): highest repairability, lowest cost, and the result is essentially like new.
Power module faults (overcurrent, short circuit, blown devices): depends on severity. Light damage is repairable; heavy damage has usually propagated to the drive board.
Control board / communication faults: repairable, but demands real capability from the repair house, and long-term stability depends entirely on their workmanship.
Encoder interface faults: sometimes the problem is not in the amplifier at all — it is on the motor side or in the cable. Troubleshoot before deciding. Do not strip the unit prematurely.

The most common mistake here: seeing an amplifier alarm and immediately concluding “the amplifier is bad”. A significant share of amplifier alarms actually originate in the motor, cabling or parameters.

2. Assess the physical damage

Open it up. Look for burnt smell, blown components, burnt-through PCB, bulged or leaking capacitors. A burnt-through PCB or large-scale carbonisation is structural damage — even repaired, long-term stability cannot be guaranteed.

3. Check model and spare-part availability

An older machine’s amplifier may already be discontinued. The question then is not “can we afford the repair” but “will there be a part next time”. If the model is discontinued and spares are scarce, repairing it now may only postpone a full retrofit.

4. Is this unit failing repeatedly?

Repeated failures of the same amplifier within a short period is a strong replacement signal. It usually means: a supply problem, a poor cooling environment, chronic overload, or an upstream fault on the motor side. Repairing the amplifier in that situation treats the symptom only.

2. Three tiers of repair — the risk differs enormously

“Repair” is not one thing. There are at least three tiers, with wildly different pricing and reliability.

Tier	Scope	Reliability	Best for
Tier 1: consumable replacement	Fans, electrolytic capacitors, connectors, fuses	High — effectively like new	Auxiliary faults, preventive maintenance
Tier 2: board-level repair	Power modules, driver ICs, PCB repair	Moderate — depends on the shop	Localised power or control-board faults
Tier 3: full refurbishment	Mass component replacement, PCB rework	Low — not advised for critical equipment	Emergency repairs on obsolete machines

Key question to ask: what exactly was replaced, whether power modules were changed, whether ageing tests were run, and how long the warranty is. “It’s fixed” with no detail carries the highest risk.

3. Three replacement options

1. New unit of the same model

Most reliable, zero compatibility risk, highest price. First choice for critical equipment where downtime is expensive.

2. Used or refurbished unit of the same model

Substantially cheaper, but watch: warranty length, whether ageing was performed, and the reliability of the source. Refurbished quality varies widely — this is the biggest uncertainty.

If you are evaluating a used machine, a related warning: ask specifically whether the amplifier on it is “newly replaced” — sometimes that is precisely how a fault is being concealed. See Used CNC Machine Electrical Inspection Checklist.

3. Cross-model substitution

Highest risk. Involves model compatibility, firmware version, parameter matching, physical dimensions and pin definitions. Do not do this on critical equipment unless the manufacturer provides a documented substitution and retrofit plan.

4. The real cost model: four factors

This is the core of the article. When deciding, list all four — not just the first.

1. Direct cost

Repair fee versus new-part price. The only number you can read straight off the quote.

2. Downtime cost (usually the dominant term)

Downtime cost = output per hour × expected downtime

For a critical machine, the hourly loss can far exceed the difference between repair and replacement. Therefore:

Repair takes 7 days, new unit arrives today → replacement is actually cheaper
Repair takes 1 day, new unit has a 30-day lead time → repair wins

Lead time and repair turnaround must be asked about and written into the decision. This is the most commonly overlooked item.

3. Risk cost

How long will the repair hold? How long is the warranty? What is the probability of a second failure?

Rule of thumb: if the repair warranty is shorter than three months, or the shop will not commit to one, risk cost is high.

4. Hidden cost

Routinely ignored, but real:

Whether parameters need resetting
Whether re-commissioning is required (gain, zero point, limits)
Test cutting and accuracy re-verification
Additional labour and downtime

A new unit usually carries more hidden cost than a repair (parameters and firmware), but the repair’s hidden cost is the possibility of paying twice.

5. Decision table: when to repair, when to replace

Situation	Recommendation	Reason
Fan, capacitor, connector failure	Repair	Very low cost, high reliability
Alarm root cause is motor/cable, not the amplifier	Troubleshoot first — do not strip	Replacing the amplifier cannot fix the root cause
Localised board fault, shop offers warranty, fast turnaround	Repairable	Clear direct-cost advantage
Model discontinued, spares scarce	Lean replace — even consider retrofit	Avoid having no part next time
Same unit failing repeatedly in a short period	Replace, and find the root cause	Without fixing upstream it will fail again
PCB burnt through, large-scale carbonisation	Replace	Structural damage
Critical equipment, very high downtime loss	Replace (new unit first)	Downtime cost dominates
New-part lead time far exceeds repair turnaround	Repair (as a bridge)	Restore production now, order the new unit in parallel

6. Three things to do before touching anything

1. Back up before you strip anything

Parameters, PMC ladder, pitch error compensation, macro programs — export and verify restorable before removing anything. Amplifier replacement may require parameter resets; without a backup you have sealed off your own retreat. Follow the standard ALL I/O and SRAM backup procedure for your control, and treat “the backup file exists” as insufficient until you have confirmed it actually restores.

2. Record model and version

Photograph the amplifier nameplate: model, specification, firmware/version identifier, factory serial number. None of this can be missing if a cross-model substitution is ever considered.

3. The verification sequence after replacement

Power-up self-test with no alarms
Jog each axis, listening for noise and watching for smooth running
Step from low to medium speed gradually
Verify backlash and positioning accuracy have not shifted
Run continuously for 30 minutes or more to rule out intermittent alarms

7. Three common misconceptions

Misconception 1: the cheaper quote is always cheaper.

Skip downtime cost and you make “saved three thousand, lost thirty thousand” decisions.

Misconception 2: a new unit guarantees stability.

If the root cause is the motor, cabling or parameters, a new unit alarms just the same. An amplifier alarm does not mean the amplifier is bad — this cannot be overstated.

Misconception 3: keep repairing an old machine.

If the model is discontinued, one successful repair does not mean a part exists next time. Work backwards from “how many more years does this machine need to run”.

Frequently Asked Questions

Q: Does an amplifier alarm always mean the amplifier must be replaced?

A: No, and the proportion is significant. Encoders, power cables, motor windings and parameter faults can all present as amplifier alarms. Trace the signal chain first and confirm the fault location before deciding.

Q: What warranty length is reasonable on a repaired unit?

A: There is no universal standard, but warranty length is an important indicator of the repair house’s confidence. A very short warranty, or a verbal promise with nothing in writing, carries clearly higher risk.

Q: Why does a new unit need re-commissioning?

A: Because firmware versions, parameter defaults and gain settings may differ. Putting a new unit into production without commissioning risks accuracy drift at best, and vibration or alarms at worst.

Q: Is it still worth repairing an obsolete machine’s amplifier?

A: It depends how much longer the machine is planned to run. For short-term use, repair is fine; for long-term use on a discontinued model, evaluate a retrofit plan alongside, to avoid a cycle of repeated repairs.