Must a Failed Fan Mean a Whole New Unit?

Three parts, three failure signatures

A stopped fan does not automatically mean a new fan unit. In a modular fan package the fan, the motor and the variable speed drive (VSD) are three separate parts, and usually only one of them has failed. The expensive mistake is ordering a whole unit before knowing which part that is — or, just as bad, replacing a healthy motor while a faulty drive keeps tripping the new one.

Start from the symptom, not from the spare-parts shelf. A fan that will not turn, a motor that hums but never starts, a drive that trips the moment it ramps up, and a fan that runs but moves too little air are four different failures, and each one points at a different part. The table below turns those symptoms into a first verdict — but a verdict is only a hypothesis until a measurement confirms it, which is the subject of the next section.

Two habits cause most misdiagnosis. The first is trusting the alarm text at face value: a drive reporting an overload trip may be telling the truth about the current, while the real fault is a seized fan bearing that drives the current up. The second is testing only at standstill: a motor that spins freely by hand can still fail under load, and a drive that looks healthy unloaded can trip at full current. Symptom plus measurement, never symptom alone, is what earns a parts order.

Write the symptom down before anyone touches the unit: what the operator saw, what the drive displayed, and whether anything changed recently — a filter change, a setpoint change, a power dip. That note costs two minutes and often settles the argument later about whether the replacement actually fixed the fault.

Isolate before you order

The isolation test is a fixed sequence: make safe, separate the three parts electrically and mechanically, then measure each one on its own. Skip a step and the test lies — a motor still wired to a faulty drive will condemn the motor, and a fan still coupled to a seized motor will condemn the fan.

Make safe first, and prove it. Isolate the supply, lock it off, tag it, and test for dead with a meter you have just proved on a known live source. On the drive side, wait out the maker's DC-bus discharge time — typically several minutes — before touching terminals, because a drive can hold a dangerous charge long after its display goes dark.

Then separate the parts. Disconnect the motor from the drive output, and uncouple the fan from the motor shaft where the construction allows it. Now each part can be judged alone. Spin the impeller by hand: it should turn smoothly through a full revolution with no graunch, no tight spot, and no blade contact with the casing. Check the blades for cracks, bent edges, and fouling, and check the tip clearance against the maker's figure — a blade that rubs the casing will fail again no matter what else you replace.

Measure the motor with a meter, not with your ears. Winding resistance should be balanced across the three phases: treat an imbalance above 2% of the mean as a failed winding until the maker says otherwise. Insulation resistance, measured at 500 V DC, should read at least 1 MΩ as a site acceptance figure — below that, the winding insulation is suspect and the motor does not go back into service on hope. Turn the shaft and feel the bearings: roughness, play, or a dry rumble means the bearings are part of the fault even if the electrics pass.

Measure the drive last, with the motor disconnected. Check that no output phase is shorted to earth, that the DC bus charges and holds as the maker describes, and that the fault history matches what the operator saw. Record every reading on the job sheet: the isolation test is only useful if the next engineer can see what was measured, not just what was concluded.

Repair or replace: decide on downtime, not habit

Once the failed part is known, the decision is a trade between three numbers: the cost of the part, the cost of the downtime, and the lead time to get the part on site. A cheap part with a six-week lead time can be more expensive than a dear part on the shelf next door, because every hour of lost cooling is an hour of derated production or a data hall on reduced redundancy. Put the downtime cost in the same calculation as the parts cost, or the calculation is fiction.

For the motor, rewinding is worth pricing only when the frame, shaft, and bearings are sound and the rewind quote lands well below a new motor — as a planning rule, below about 60% of the new-motor price — with a lead time the site can live with. Above that, or when the frame is cracked or the shaft is scored, a new motor is the cheaper decision even before downtime is counted. For the drive, board-level repair makes sense only from a repairer who can test the unit under load and warrant it; a drive that returns untested is a second outage waiting for a date.

For the fan itself, repair is a narrow case. Clean fouling, yes. Replace a damaged blade set, yes, if the maker supplies it as a balanced assembly. Weld or straighten a cracked impeller, no — a cracked impeller is a balance and safety failure, and it gets replaced as an assembly. The whole-unit swap still has one legitimate use: when two of the three parts have failed together, or the casing and mounting are damaged beyond the failed part, replacing the assembly can be faster than rebuilding it piece by piece. That is a measured exception, not a habit.

As a planning allowance only — not a quotation — a single-part swap on an accessible AHU or fan wall typically takes one to three hours of site work, against half a day or more for a full unit change with duct and mounting rework. That gap is where the modular saving lives: fewer parts bought, less rework, and a shorter outage. Where the failed part is not on site, the decision flips to whatever restores cooling first — which is why the spare-parts list below exists.

Size the replacement fan

A replacement fan is sized on the duty point, not on the old part number. The duty point is two numbers together: the airflow the system needs, in m³/h, and the static pressure the fan must overcome at that airflow, in Pa, measured at the installed point with the filter in its loaded state. A fan chosen on airflow alone, or on the old catalogue page, will miss the duty point the moment the system resistance differs from the original installation — and it almost always differs, because filters load, ducts change, and grilles get replaced.

Hold margin explicitly. Size the fan so the duty point sits on the fan's P-Q curve with 10–15% static-pressure margin above the loaded-filter total, and check the duty at the design temperature, not at a mild test-bench temperature: hot air is less dense, so the same volumetric flow carries less cooling. Check the absorbed power at the duty point too, and confirm the motor you pair it with covers that power with its own margin — a fan and a motor sized in isolation can still fail as a pair.

Then match the interface, because a fan that fits the curve but not the casing is not a replacement. Record the footprint and mounting pattern, the inlet and outlet sizes and flange drilling, the rotation direction viewed from a stated side, the weight and lifting points, and Photograph the old mounting before removal: a bolt pattern remembered wrong is a second site visit. If any interface dimension differs, check in writing that the casing, duct, and guard still fit — otherwise the 'equivalent' fan becomes a fabrication job.

Size the replacement motor

The replacement motor starts from the load, not from the old rating plate. If the duty point has changed — a different fan, a changed system resistance, a higher design temperature — the shaft power at the new duty is the number that sizes the motor, and the old motor's kilowatt rating may now be wrong in either direction: too small and it overloads, too large and it runs inefficiently and costs more than it should. Where the duty is unchanged, matching the old rating, frame, and speed is the safe route, but still verify it against the measured duty rather than assuming it.

Specify the efficiency class in writing: IE3 or higher where regulation and supply allow, and IE4 or IE5 where the duty is continuous and the energy case closes — the motor runs for years, so a one-class improvement pays back many times over its premium. Match the mounting form (foot, flange, or face, in the maker's IM code), the frame size and shaft height, the pole count and therefore the speed, and the terminal-box position, so the new motor drops onto the old plinth without rework.

Then the environment and the drive. Match the ingress protection (IP) rating to the site — dust, washdown, and outdoor exposure do not forgive an undersized IP rating — and specify insulation class F with a class B temperature rise for continuous duty (S1), which buys thermal headroom on hot days. If the motor will run on a VSD, say so on the order: VSD-duty insulation, and shaft earthing or insulated bearings where the maker requires them, are what keep drive-induced bearing currents from becoming the next failure. A motor ordered without that note is a standard motor being asked to survive a drive duty.

Size the replacement VSD

The replacement drive is sized on current, not on kilowatts alone. Its continuous rated output current must cover the motor's full-load current at the duty point, with the maker's overload rating — commonly 110% for 60 seconds — available above that for starting and transient peaks. A drive chosen on kilowatt rating while the motor draws a higher current at low speed or high temperature will trip exactly when the site needs it most, so current is the sizing number and kilowatts are the cross-check.

Match the control interface to the existing system before ordering, because a drive that cannot talk to the site controller is a rewiring project. Record whether the site uses a 0–10 V or 4–20 mA speed reference, a digital run/stop, or a fieldbus such as Modbus RTU or BACnet, and confirm the replacement drive provides that interface with spare inputs and outputs for alarms and status. Keep the old drive's parameter set: photograph every screen or export the file, because re-entering ramp times and current limits from memory is how a new drive inherits an old misconfiguration.

Then the electrical environment. Confirm whether the site needs a built-in or external EMC filter for its emission class, and whether harmonic mitigation — a line reactor or DC choke — is required to keep total harmonic current distortion within the site or utility limit. Check the maximum motor cable length the drive allows, and add an output choke or dv/dt filter when the run is long, or the motor insulation pays for the cable length. Finally, confirm the cooling and mounting: a drive in a hot, dusty enclosure needs its ventilation and derating checked, not just its catalogue current.

Why separability beats a sealed unit

This is where the modular route earns its keep against the sealed EC fan. An EC fan is built as a sealed, moulded unit: motor, impeller, and electronics in one package. When one element fails in the field, repair is impractical and the whole unit has to be swapped out — ABB's own account of the difference. A modular package keeps the three parts separate, so each can be replaced independently, and the healthy parts keep working.

The industry evidence is a recent ABB and NOVENCO package pairing NOVENCO's ZerAx axial fan with an ABB IE5 permanent-magnet motor and variable speed drive (reference: ABB and NOVENCO fan package for data centre cooling, also reported by Cooling Post). ABB reports that replacing EC fans with the package in a data centre cooling application of more than 1,000 fans cuts annual energy consumption by 15–40%, depending on the age of the existing installation, with a return on investment in under two years for many retrofit projects at current energy costs (reference). The same account puts system efficiency at up to 85% in a typical AHU installation and the package design life at 15–20 years (reference). Those are the makers' figures for their package, not a promise for any other product — but the direction is the point: separability and system efficiency together, not either alone.

The retrofit market is moving the same way from the other side. ebm-papst's NEXAIRA.Fit programme, which replaces outdated ventilation fans with high-efficiency EC fans plus digital optimisation, now carries an energy-efficiency guarantee backed by operating data; the maker reports that replacing fan technology alone can cut energy consumption by up to 50%, and that additional digital optimisation can raise savings to as much as 70%, depending on application and operating conditions (reference: ebm-papst adds energy efficiency guarantee to NEXAIRA.Fit retrofits). Note the distinction kept throughout: EC here is an industry concept attributed to its makers, not a MEGA Tech product claim. And when you compare a sealed unit against a modular package, compare the cost of the next failure too, not just the price of this purchase.

The spare-parts list

A spare-parts list is a document, not a shelf of hopeful boxes. For every fan type on site, it records the part, the exact data needed to reorder it without returning to site, the quantity held, and the lead time if it is not held. The table below is the minimum: a list that records only part names, without ratings and dimensions, will fail at the first night call-out.

Set quantities by criticality and lead time, not by round numbers. Hold on site what stops the process: one spare motor and one spare drive per critical fan type, and one spare impeller assembly per fan type where the impeller is a wear or damage item. For non-critical duty, holding the data and a confirmed supplier with a quoted lead time can be enough — but that lead time must be real, rechecked yearly, and written on the list.

Store and count like it matters, because a spare that cannot be found is not a spare. Label every part with the fan type, the duty, and the date received; store motors and drives dry, upright, and protected from dust and condensation; rotate bearings by hand at each count so they do not brinell in storage; and count the stock twice a year against the list, reconciling every difference the same week. The count is also the moment to check that the recorded data still matches the installed units — fan types drift over years of piecemeal replacement.

After the swap: the commissioning checklist

A swap is not finished when the new part is bolted in; it is finished when the rebuilt unit passes the same tests a new installation would. Run the checklist in order, and record every reading — the sheet is the evidence the repair worked, and the baseline the next fault will be judged against.

First, the electrical checks with the unit isolated, then energised. Confirm the rotation direction against the arrow on the casing before the fan runs under load — a fan running backwards still moves some air, which is why backwards fans survive for weeks while quietly overheating the space. Then measure the running current on each phase at the duty point: the current must sit at or below the motor's full-load current, and the imbalance between phases must stay within 5% of the mean. Above 10% imbalance, stop and investigate — supply, connections, or winding — rather than signing off.

Then the mechanical and airflow checks. Vibration on the bearing housings, measured as RMS velocity, should meet the acceptance value agreed before the swap; as a commonly written site acceptance figure, 2.8 mm/s RMS is the new-installation acceptance level to agree in writing, with higher readings triggering investigation rather than acceptance. Measure the airflow at the duty point and confirm it within ±10% of the design duty, and check the noise against the site limit at the agreed distance — a unit that meets airflow but breaches noise has still failed its acceptance.

Finally, the trial run. Run the unit continuously at the duty point for at least 4 hours, logging current, vibration, bearing temperature, and airflow at the start, at one hour, and at four hours. Then keep it on a 7-day monitored run, with the drive fault log and site temperature log reviewed at the end of the week, before sign-off. A fault that appears on day three is still the repair's fault; a sign-off on day one would have missed it. File the sheet with the spare-parts list, so the next engineer starts from measurements, not from memory.

Where MEGA Tech fits — and where it does not

MEGA Tech is a custom cooling-fan manufacturer. What we do, in the terms our customers use: custom cooling-fan development and manufacturing, fan selection, air-path review, sample validation, and batch production. Our published range is DC axial fans, selected by operating condition — airflow and static pressure at the installed point — not by part number, with samples validated before batch orders.

Our position in this market is plain. We make replacement fans. We supply cost-effective cooling fans for businesses. We provide samples and testing for startups. And we grow with startups and small businesses. That is the whole of the claim: no invented model numbers, no lifetime figures for products we have not published, and no EC product of our own — EC in this article is an industry concept, described through its makers ABB, NOVENCO, and ebm-papst, not a MEGA Tech range.

We do not supply motors, drives, liquid loops, coils, or whole air-handling units as standalone products, and we do not rebuild a customer's drive or rewind a customer's motor. Where a site needs a modular package with a separately sourced motor and drive, we work to the interface: the fan duty, the mounting, and the airflow the package must deliver. Our manufacturing holds ISO 9001, CE, and RoHS. Every article on this site carries the same boundary: third-party figures are attributed to their makers, and MEGA Tech figures only ever describe MEGA Tech products that are actually published.

To ask MEGA Tech for a selection or a sample, send the fan size, voltage, airflow and static-pressure target, quantity, and constraints — plus the operating conditions the fan must survive. The selection starts from the duty point, and the sample proves it before any batch is discussed.

Engineer FAQ

Q: The fan stopped completely. Is the whole unit dead? A: Not necessarily. Work the isolation sequence: supply and fuses, drive power stage and fault log, motor winding resistance and insulation, then the fan shaft by hand. In a modular package, a dead fan usually means one dead part, and the measurement tells you which one.

Q: Can I replace the motor with a different kilowatt rating? A: Only if the new rating matches the shaft power at the actual duty point, with margin, and the frame, mounting, and speed match the installation. A bigger motor is not automatically safer — it changes the drive sizing, the starting current, and the efficiency at part load — so size from the duty, not from the shelf.

Q: The new drive trips where the old one did not. What changed? A: Usually the parameters, not the hardware. Compare the ramp times, current limits, and protection settings against the old drive's set, then check the control signal and the motor cable length. A drive that inherits a misconfiguration inherits the trip that goes with it.

Q: How do I know the vibration reading is acceptable? A: Agree the acceptance value in writing before the swap — 2.8 mm/s RMS is a commonly written new-installation acceptance figure — then measure at the bearing housings at the duty point and compare like with like. A single number without a location, a duty point, and an agreed limit proves nothing.

Q: What do you need from me to select a replacement fan? A: Airflow and static pressure at that airflow, the temperature basis, voltage tolerance, control interface, connector, envelope, noise target, and life assumption. With those eight inputs the selection starts from the duty point; without them it starts from a guess.

Questions for a Design Review

Bring these answers, with measured values attached, before a replacement part is released. To ask MEGA Tech for a selection or a sample, send the fan size, voltage, airflow and static-pressure target, quantity, and constraints — plus the operating conditions — to inquiry@cnmegatech.com or on WhatsApp, and the selection starts from the duty point.

• Which part failed — fan, motor, or drive — and which measurement proves it?
• What are the airflow and static pressure at the duty point, and where is the margin?
• Does the replacement motor match the duty power, frame, mounting, and IE class?
• Does the replacement drive cover the motor full-load current and the site control interface?
• What spares are held for this fan type, and what is the real lead time for the rest?
• Did the rebuilt unit pass rotation, current balance, vibration, airflow, and the 4-hour trial?