📌 Key Takeaways
- •Under the same workload and the same room temperature, load temperatures are higher than they used to be. One third-party summary uses "5–10°C above the stable period" as a reference — that is that site's figure, not an official standard, and this article mentions it only as a reference, not a diagnostic threshold.
- •The fan spins up sooner and more often under light load — for example, it now spins up with a dozen browser tabs open when it previously did not.
- •One core runs clearly hotter than the others at the same clock speed, which may mean paste has thinned or shifted locally (pump-out) and contact is uneven.
- •Under sustained heavy load there is throttling (clocks pulled down) or repeated temperature-limit warnings (HWiNFO showing PROCHOT or a GPU thermal limit).
- •Under sustained load the machine becomes unstable, shuts down suddenly, or blue-screens — but other causes must be ruled out first; do not blame paste on sight.
First, the premise: this is the "still loud after cleaning" article, not another cleaning guide
The earlier article in this series (batch02, "Laptop Fan Loud or Stuck at Full Speed? Causes, How to Tell Them Apart, Safe Cleaning, and When to Send It In") already covered the basics: how to tell common causes apart, how to clean safely, software load, placement, and basic send-in timing. This article starts from a different premise: a proper cleaning has already been done, and the noise is still there. If that premise does not hold — if the machine has never actually been cleaned — read the basic article first instead of jumping into the advanced checks here.
Once cleaning is out of the way, the problem moves up a level. It is no longer "dust blocking the exhaust" but the fan mechanics themselves, the thermal interface, or the speed-control chain. This article works through six parts: telling sound types apart, bearing wear, dried thermal paste, other mechanical causes, fan-speed control failure, and single-fan failure in dual-fan gaming laptops — and finishes with the fan-versus-full-heatsink decision, DIY limits, and a send-in checklist.
Source levels are used consistently throughout, and this matters: official means a manufacturer's official support document or official article; official forum means a reply by a manufacturer's staff or moderator on its official forum, which is not the same as a formal support document; vendor FAQ means an official FAQ from a parts maker (for example a thermal-paste or fan manufacturer); third-party summary means a repairer or media guide, not official; forum experience means a user or technician sharing a case on a forum. Anything unconfirmed or disputed is labelled as such in the text, never written as a conclusion, and never dressed up as official guidance.
Chapter 1: Airflow noise vs mechanical noise — hear the difference before guessing the cause
Airflow noise and mechanical noise sound completely different. A UK repairer's practical rule, consistent with HP's and Lenovo's official classifications, puts it bluntly: if the sound is a smooth rushing noise that rises and falls with load, think cooling or airflow first; if there is grinding, clicking, scraping, or a new high-pitched whine, treat it as a mechanical fault and avoid heavy load before inspection, because a failing fan can slow down or stop, and overheating follows.
Sound comparison table
| Sound | Character | Likely direction | Source level |
|---|---|---|---|
| Smooth "whoosh" | Continuous, smooth, rises and falls with load | Normal physics of air through blades and exhaust; think cooling, airflow, or speed-control problems first | Physical phenomenon + third-party summary |
| Grinding | Rough, like something being ground down | Usually bearing failure, or a blade touching an obstruction | HP official article; Dell official forum says the same |
| Rattling | Something loose, vibrating, knocking | Loose screws or damaged blades | HP official article; Lenovo official glossary says rattling can mean a loose or damaged fan |
| Buzzing | Sustained low hum with vibration | Possibly a foreign object jammed in, loose fan mounting, or bearing wear | HP official article |
| Whine, irregular judder, scraping | High-pitched, intermittent, does not follow load | Points to a mechanical fault, not plain airflow noise | Third-party summary |
| Intermittent, varies on its own, present even at low load | Sound that "does not follow load" | Points to a bearing or mechanical imbalance rather than a cooling-load problem | Forum experience: a fan with a bad bearing "is not spinning faster, just louder" |
Three auxiliary tests (for telling sounds apart only — not repair steps)
1. Speed comparison test. Check fan speed (RPM) with HWMonitor, HWiNFO, Core Temp, or the manufacturer's own tool. If there is an obvious mechanical sound but speed has not risen to match, suspect mechanical imbalance or bearing wear rather than a cooling-load problem. One third-party summary notes that the RPM reading itself may be unstable or delayed in this state, and continued use carries a risk of the fan seizing suddenly. The more reliable step is the maker's own diagnostic: Dell, press F12 at boot for the diagnostics menu and run the fan test; HP, press Esc or F2 at boot for HP Hardware Diagnostics, go to Component Tests, and run the Fan Test. If anything is abnormal, note the error code for the repair visit.
2. Tilt test. Gently tilt the machine. If the sound changes clearly — louder, quieter, appearing, or disappearing — suspect mechanical imbalance, a bearing, or something loose. A UK repairer lists exactly this: "sound changes when tilting the laptop or pressing near the fan points to bearing wear or fan imbalance." An Acer Swift 3 user case (forum experience) matches: the fan suddenly got loud when the machine was tilted or lightly bumped, and replies pointed to looseness, displacement, or a foreign object.
3. Press test. Press lightly on the chassis near the fan — lightly is the operative word. If the noise weakens or disappears, suspect loose mounting or bearing play. A desktop case (forum experience) used a thumb pressed lightly on the fan hub, with the noise disappearing, to judge a bearing problem. Three warnings must travel with this test: light pressure only; if you are not comfortable, do not try it; and never reach inside a running machine.
A necessary warning about oiling. A folk remedy circulates online — peel back the fan sticker and put a drop of sewing-machine oil on the bearing. Dell community posts and iFixit replies mention it, but a Dell community reply also states plainly that even if it works, it is only short-term. This article does not recommend oiling: bearing wear is mechanical wear, oil does not restore it, and opening the machine to poke at the bearing or pull cables risks damaging other parts. If a mechanical fault is suspected, reduce heavy-load use and get the machine inspected promptly.
Chapter 2: Bearing wear — why there is still a grinding sound after cleaning
Sleeve bearings vs ball bearings
Laptop fans commonly use two bearing types: sleeve (oil) bearings and ball bearings. SANYO DENKI's official blog (vendor FAQ) and the GamersNexus third-party guide broadly agree: sleeve bearings start quieter and wear out sooner, with noise growing over time as lubrication dries; ball bearings last longer and are relatively stable at high speed, though cheaper ball bearings can carry a slight clicking sound. GamersNexus adds that sleeve bearings tend to "die suddenly," with no clear warning beforehand — there may occasionally be an early whine or a slowing speed, but it is hard to tell apart from other noises — and that high temperatures and sharp temperature swings accelerate lubrication failure.
| Item | Sleeve bearing | Ball bearing | Source level |
|---|---|---|---|
| Early noise | Quieter at first | Relatively stable at high speed; cheap units may carry a slight click | Vendor FAQ + third-party summary |
| Manufacturer-rated life | Around 20,000–30,000 hours (TITAN rates 25,000 hours at 25°C) | Around 50,000–100,000+ hours (TITAN rates 50,000 hours) | Vendor FAQ + third-party industry summary; figures disagree across sources |
| Ageing behaviour | Noise grows as lubrication dries; may die suddenly with little warning | Wears more slowly, more often gives advance signs | Third-party summary |
| Laptop caveat | Most life figures come from desktop-fan contexts or manufacturer ratings; real life depends on temperature, mounting orientation, and usage intensity | Same | Research note |
Intermittency is a strong bearing clue
An Acer Predator Triton 300 user case (forum experience) is worth quoting for its logic: a buzzing-plus-grinding sound that survived two cleanings, varying in loudness. The reply judged it a bearing problem, for a reason that helps readers triage: "if a blade were touching something, it would be constant, not intermittent; only a bearing problem varies with speed and temperature like that." That logic is a useful first filter — a constant, unchanging noise points first to a blade touching something or a foreign object; an intermittent noise that varies with speed and temperature points first to a bearing. It is a first filter only, not a diagnosis; diagnosis needs inspection.
Chapter 3: Dried thermal paste — temperature gives it away first, noise is only the result
Dried paste does not itself make a sound. It makes the fan run at full speed for longer because heat no longer transfers to the heatsink properly, so the fan has to work at high speed to hold temperatures down — which is why the machine is "still loud after cleaning." So suspecting paste means looking at temperatures, not just listening.
Temperature signs that point to dried paste
The core logic, consistent across several third-party summaries, is: compare against your own machine's history, not against one absolute number. Different CPUs have very different idle and load temperatures, so "50°C at idle" can be normal on one machine and a warning sign on another.
- Under the same workload and the same room temperature, load temperatures are higher than they used to be. One third-party summary uses "5–10°C above the stable period" as a reference — that is that site's figure, not an official standard, and this article mentions it only as a reference, not a diagnostic threshold.
- The fan spins up sooner and more often under light load — for example, it now spins up with a dozen browser tabs open when it previously did not.
- One core runs clearly hotter than the others at the same clock speed, which may mean paste has thinned or shifted locally (pump-out) and contact is uneven.
- Under sustained heavy load there is throttling (clocks pulled down) or repeated temperature-limit warnings (HWiNFO showing PROCHOT or a GPU thermal limit).
- Under sustained load the machine becomes unstable, shuts down suddenly, or blue-screens — but other causes must be ruled out first; do not blame paste on sight.
What dried paste looks like once opened is for inspection staff only — this article does not teach users to remove the heatsink module themselves: paste gone dry, cracked, crumbly, squeezed to the edges, or leaving a bare patch in the middle.
Paste ageing: disputed, so no fixed number of years
Thermal-paste maker Noctua's official FAQ states that NT-H1 and NT-H2 have a rated service life of up to 5 years on a CPU, and after 5 years users should check temperatures from time to time and need not replace unless the delta temperature (CPU temperature minus room temperature, compared at 100% load) has clearly risen; some customers report good results after more than 10 years. Unopened paste lasts 2 years after opening. That must be qualified: Noctua's 5-year figure applies to its own products only and must not be applied to every laptop's factory paste. The third-party rule of thumb "replace every 2–3 years" is just that — a rule of thumb. A laptop-specific factor from third-party summaries is that thin laptop heatsinks and sharp thermal cycling make pump-out or separation more likely, and long gaming sessions, frequent hot-cold cycles, and sustained high-temperature operation age paste faster — a third-party claim, not official guidance. The conclusion is one sentence: judge by temperature monitoring, not by the calendar.
Reasonable expectations after repasting
This article writes no guaranteed temperature-drop figure. One third-party repairer claims 10–20°C from repasting; that is a promotional claim from that site, with no model and no test conditions, and must not be written as a guarantee. Real improvement depends on how dried the old paste was, CPU power draw, heatsink design, room temperature, and test method — badly cracked old paste can improve a lot, while paste that was not dry may barely change at all.
The correct comparison is Noctua's delta method: compare CPU temperature minus room temperature (delta) at 100% load, which cancels out summer-versus-winter room differences. SunbeamTech's third-party guide adds: use the same heavy workload each time (for example a 10-minute stress test), record the delta, and re-measure the same way after repasting before comparing. The fair verdict sentence is: "if the delta clearly falls after repasting and the fan no longer runs at full speed constantly, paste was one of the factors; if temperatures barely change, check the fan itself or heatsink contact next." An HP official-forum Omen case reinforces this: where grinding noise and overheating coexisted, the reply listed four possibilities together — fan bearing or blade wear, blocked or damaged fins, dried paste, and incorrect reassembly — warning users not to treat a repaste as the whole answer.
One binding rule many people miss: whenever the heatsink has been separated from the CPU or GPU, the old paste must be cleaned off and replaced; never layer new paste over old paste (third-party summary and iFixit repair guide agree). Paste fills microscopic gaps; thicker is not better, and it needs even heatsink pressure to spread. This is one of the key advanced plots behind "still loud after cleaning": many DIY cleanings remove the heatsink but never repaste, or repaste badly, leaving poor conduction and a fan stuck at full speed.
Need a laptop repair expert?
Free diagnosis · Transparent pricing · No fix, no fee · 180-day warranty
Intel's official troubleshooting points can back this section: check that the cooling solution matches the CPU's socket and power; make sure the heatsink is installed correctly with even pressure at all mounting points; make sure paste quantity is correct; confirm fan blades are unobstructed and spin freely. The research note flags that Intel support article 000005791's search-result URL path does not match its topic, so this article paraphrases the points and does not link that URL; the link should be verified before citation.
Chapter 4: Other mechanical causes that survive a cleaning
Deformed or imbalanced blades
Damaged blades show up as rattling plus a feeling of vibration. HP's official article states that rattling comes from loose screws or damaged blades, fixed by tightening screws, checking blades, and replacing the fan if necessary. An HP official-forum staff reply adds: during inspection look for "cracked blades, loose mounting screws," and "if you see wobble or uneven rotation, replacing the fan is the safest option." The imbalance clues echo Chapter 1: noise present without a matching speed rise, unstable RPM readings, and abnormal vibration felt in the chassis. Unconfirmed note: no public source confirms that "the cleaning process itself bent the blades" is a common cause, so this article does not write that; "damaged or deformed blades" is listed only as a check item, with causes including impact, ageing, or foreign-object strikes.
Foreign objects a cleaning can miss
Beyond hair and fibres, there are object types a cleaning easily misses — especially when the cleaning "never removed the fan itself and only blew air through the exhaust":
- Internal cables or ribbons touching the blades. A Dell G3 case reply (forum experience) notes blades can strike internal wires or ribbons; an older Dell official-community post had a case of "a displaced wire touching the fan," fixed by moving it back; an iFixit Acer case also lists "an internal cable touching the fan when the laptop is moved."
- A screw dropped inside the machine. An ASUS ROG official-forum reply warns that a screw dropped during assembly rattles when the machine is shaken; if you are not sure whether a screw fell in, do not power on, because it can short-circuit and damage the board.
- Debris caught between the fan and the heatsink. An HP official-forum technician reply mentions tape or mesh sometimes sits where the fan meets the heat pipe, and to check whether anything is caught between the fan and the fin stack.
Loose mounting and incorrect reassembly
Loose mounting shows up as buzzing, rattling, and vibration sounds, and the fan suddenly gets loud when the chassis is tilted or lightly bumped. HP's official article lists "loose fan mounting" as one cause of buzzing; HP official-forum replies say loose screws or mounting problems let the fan vibrate loudly. "Incorrect reassembly after cleaning" is this article's key plot point: an HP Omen case reply (official forum) explicitly lists "incorrect reassembly — a wire or part may be touching the fan after reassembly"; Intel's official guidance stresses that the cooling solution needs "even pressure at all mounting points," since a loose heatsink module hurts cooling and creates vibration. Damaged fins deserve a mention too: the HP Omen reply notes "dust-clogged or bent fins block airflow and trap heat," and bent fins can also change the airflow path and add noise.
Confusion items — do not get these wrong
- Hard-drive mechanical sound. A Dell community reply (forum experience) warns that on older machines still using spinning hard drives, the drive motor can be mistaken for fan noise; removing the drive distinguishes them. This applies only to old HDD machines, not SSD machines — the article states that.
- Full speed at boot is normal. An ASUS manual (third-party mirror page) notes that fans running at full speed during boot is normal behaviour and they slow down once the system loads. Do not write "loud at boot" up as a fault.
- ASUS's official line. An ASUS ZenTalk staff reply says abnormal fan noise should first be checked against the official FAQ "Notebook/Desktop/AIO Troubleshooting - Computer Overheating and Fan issues," and if it persists, to contact local customer service. This research did not obtain a direct link to that FAQ, so this article only writes "ASUS officially advises following its troubleshooting FAQ first" and does not invent a URL.
Chapter 5: Fan-speed control failure — BIOS, EC, fan curves, and power modes
This is the chapter most easily mistaken for a hardware fault: the fan itself is fine, but the machine is "telling it to run at full speed." An Intel technical document (official, low-level, written for system designers) supplies the concept: fan speed should be controlled from the processor's digital temperature sensor (DTS); in plain language, fan speed should normally follow temperature. If the fan runs at full speed while temperatures are low, something in the control chain — sensor to EC to fan — is wrong.
Maker thermal-mode comparison table
| Brand / tool | Modes | Fan behaviour | Notes |
|---|---|---|---|
| Dell Power Manager | Optimized / Cool / Quiet / Ultra performance | Cool raises fan speed to cool the chassis, with more noise; Quiet lowers fan plus processor speed for less noise, at some performance cost; Ultra performance raises processor plus fan speed — fastest but loudest, switchable with Fn+T | Official note: if there is a fan fault, Thermal Management is disabled — usable as a diagnostic clue. Source: Dell user guide (official) |
| Lenovo Vantage / Legion Zone | Performance / Balance / Quiet, switchable with the Fn+Q shortcut | In Balance the machine shifts dynamically between Performance and Quiet with the workload; Quiet trades performance for silence | A non-compliant power adapter may prevent switching to Performance. Source: Lenovo user guides (official) |
| HP | Fans are managed automatically by the system; no manual speed control | Some business models have a BIOS "Fan Always On" option (affects idle behaviour only); gaming models control heat indirectly through OMEN Gaming Hub or HP Command Center Quiet/Performance profiles; OMEN 16 offers no custom fan curve | Source level: HP staff statements on the support forum, not manual text. Do not teach forcing speeds with third-party SpeedFan |
| ASUS ROG Armoury Crate | Silent / Performance / Turbo / Windows / Manual | Manual mode allows a custom fan curve (dragging temperature-speed points) and power-target adjustment; Turbo also raises the power limit, so temperatures can be higher than in Performance; Turbo needs the original adapter on wall power | Official wording warns that manual GPU overclocking can destabilise the system and beginners should read the overclocking guide first; most users only need the fan curve. Source: ROG official guides |
| MSI | Cooler Boost; MSI Center User Scenario | With Cooler Boost on, fans run at maximum speed and loud air noise is normal; in MSI Center, Extreme Performance's gear icon switches fan mode (Auto / Advanced custom slider) | For an MSI machine still loud after cleaning, check Cooler Boost first — the Fn+F7 physical key can enable it by accident. MSI's official troubleshooting order: turn off Cooler Boost, clear the exhaust, reset the EC, update the BIOS. Source: MSI support documents (official) |
| Acer NitroSense | Auto (default) / Max / Custom; CoolBoost | Custom adjusts CPU and GPU fans separately; CoolBoost raises maximum fan speed — Acer's site figures: +10% fan speed and +9% CPU/GPU cooling versus auto mode | Notebookcheck's third-party report says newer Acer laptops lock fan control at the BIOS/EC level, so third-party tools cannot read or control the sensors; Acer users should stick to official NitroSense and third-party speed tools are not recommended. Source: Acer user manual plus site (official) + third-party summary |
Common signs of control failure
These are composite patterns assembled from several sources — labelled as a composite judgement with mixed source levels (official-forum user reports, forum experience, official FAQ), not written as conclusions:
1. At idle or low temperature (40–50°C), the fan surges to full speed intermittently and falls back. A Dell G3 community case (official-forum user report): the BIOS fan curve was badly written, and a later official BIOS update improved it.
2. The fan runs at 100% constantly while temperatures are not high — possibly a fan speed-sensor fault. An iFixit Q&A (forum experience) notes that if the tachometer or thermal wire of a 4-wire fan breaks, the board gets no feedback. Unconfirmed: wire-colour definitions differ by maker; this article does not hard-code them.
3. Third-party monitoring software (e.g. HWiNFO) reading EC sensors frequently can trigger abnormal events; MSI's official FAQ tells users to turn off HWiNFO's EC Support and states plainly that turning it off does not affect fan control. Writing note: abnormal readings in monitoring software do not equal a broken fan.
4. Fan behaviour changes abruptly after a BIOS update. An HP forum user reported a machine going from quiet to loud after a BIOS update, attributing it to changed fan behaviour; disputed: another reply said to check the BIOS release notes first. This article mentions both sides.
Settings a user can safely check (without touching advanced BIOS values)
All of these are user-level operations supported by the makers: first, switch thermal modes in the maker's control centre (Dell Power Manager, Lenovo Vantage Fn+Q, OMEN Gaming Hub Quiet, Armoury Crate, MSI Center, NitroSense); second, in Windows Settings, System, Power & battery, set the power mode to "Balanced" or "Best power efficiency" (third-party summary, consistent with Lenovo's official "Best power efficiency is quieter and cooler" concept); third, in advanced power settings, Processor power management, Maximum processor state and System cooling policy (Active or Passive) — a third-party summary, not confirmed maker by maker, labelled as such; fourth, MSI's EC reset, official steps: shut down, unplug power, hold the power button for 30 seconds, plug power back in and boot (works for non-removable-battery models too; first boot after reset needs the adapter); fifth, update the BIOS plus chipset or graphics drivers (HP staff advice), and back up data before a BIOS update.
The risk statement must be included: do not tell users to change advanced BIOS thermal values or unlock fan curves at random. A Dell community moderator notes that makers deliberately do not open up much fan control, because wrong settings can cause overheating damage (forum experience); HP likewise locks manual curves for safety and warranty reasons.
Maker diagnostics: what you can run before sending the machine in
Dell's official troubleshooting document (official) gives two steps: first, run a full Scan Hardware in SupportAssist; second, boot with F12 into Diagnostics, choose Advanced Test, tick only the fan items, run the Thorough test, and note the error code or validation code for the repair visit. HP has a boot-time F2 diagnostic (mentioned in an HP staff reply). Lenovo Vantage also has a hardware scan, but no official document was found confirming a fan-specific test item in Vantage — marked unconfirmed, so this article only writes the general line "use the maker's diagnostic tools." Bringing a code to the repair visit makes inspection much faster.
Chapter 6: Dual-fan gaming laptops — when one side fails, how to tell
Architecture background
Gaming laptops commonly have one fan each for the CPU and GPU, with several heat pipes; some higher-end models use a vapour chamber or a shared-pipe design where CPU and GPU share heat pipes. That statement comes from a third-party summary and is not officially confirmed, so this article phrases it as a "common design," not a specification claim.
Identification method comparison table
| Method | How | Reading | Limits and source level |
|---|---|---|---|
| Exhaust temperature by feel, left vs right | Compare the two exhaust outlets | The side with fins should feel clearly warmer; if that side feels only lukewarm, about the same as the fin-less side, it may be poor thermal contact (aged or shifted paste) or heat-pipe failure — not necessarily a broken fan | Feel is an empirical rule of thumb, for initial reference only, not a diagnosis. Forum experience, unconfirmed |
| Sound location | Listen for which side the abnormal sound comes from | Sustained grinding, clicking, or scraping mostly points to bearing wear; if the noise clearly comes from one side, that side's fan can be isolated | Distinguish "wind noise" (high speed, normal) from "mechanical noise" (fault sign). Third-party summary |
| Monitoring software readings | Use HWMonitor, HWiNFO, MSI Afterburner, and similar tools for CPU/GPU temperatures and fan speeds | Read together with symptoms and the maker's specifications | Some laptop fans do not report RPM at all; safe temperatures vary by processor, model, and load, so no single temperature or speed figure condemns a fan. Third-party summary. Also see Chapter 5: MSI officially notes third-party tools reading EC sensors can trigger anomalies — if readings look abnormal, close such tools and observe again |
| Boot error messages | Watch for fan-related errors at boot | Dell BIOS reports "CPU fan failure" or "System fan failure" and demands action; Dell Power Manager disables thermal management when a fan fails (see Chapter 5) | Firmware-level errors are more reliable than third-party software readings. Source: Dell manual (official) |
Zero-RPM mode vs real failure: avoid misjudging
These are usually normal: fans stopped at desktop idle, fans stopped under light load, fans spinning briefly during the boot self-test and then stopping. These point towards real failure: the GPU or CPU is already hot but a fan still reads 0 RPM; manual speed adjustment has no effect; one side's fan never spins while the other side does; a fan twitches but cannot get going; or the failure comes with throttling or shutdowns. This comparison comes from a third-party summary (a retailer tech blog, originally about graphics cards — flagged here as a conceptual analogy applied to laptop dual fans).
What a single-side failure does to cooling: the CPU or GPU that side serves loses most of its cooling, which can cause throttling, random restarts, or thermal shutdowns; Dell officially warns that a fan fault disables the whole Thermal Management. In a shared-pipe design, one side failing may drag both sides down — that concept is unconfirmed by official sources, so this article writes "may"; no official document was found stating "single-side failure effects" word for word — marked as unconfirmed official wording, and this article sticks to describing symptoms.
Chapter 7: Fan replacement vs full heatsink module — the decision, DIY limits, and send-in judgement
Replacement decision table
| Situation | Replacing just the fan may be enough | Replace the whole heatsink + heat-pipe assembly | Source level and notes |
|---|---|---|---|
| Fan's own mechanical fault: bearing noise (grinding or clicking), broken blade, not spinning at all with power present, boot fan-failure error with an undeformed heatsink | Yes | Fans and heatsinks are separate parts on many models, so the fan alone can be replaced while keeping the original heatsink. HP official-forum accepted solution: open it up and check — the fan with its housing is a separate component from the heatsink; another reply says buy the matching fan by the HP service manual part number. Tom's Hardware forum Dell Latitude E6430 case: if the heatsink is fine and no repaste is needed, replacing just the fan works (forum experience) | |
| Heat pipe bent or blocked | Yes | HP forum veteran reply: the heatsink body is a metal block that rarely breaks, but heat pipes can bend or clog, and then replacement is needed | |
| Heat pipe failed from a crack or pinhole leak: exhaust not warm plus temperatures that will not come down | Yes | Forum experience, unconfirmed: ordinary users cannot verify a heat-pipe leak themselves, so this article writes "suspected heat-pipe problem — send in for inspection" and does not teach self-diagnosis | |
| Laptop heatsinks are model-specific parts | Buy against the part number; do not mix and match | Same | HP forum volunteer reply |
DIY limits: steps ordinary users should not do themselves
These are the research team's recommended limits, not official bans, and this article labels them "recommended for repair service" rather than "officially forbidden," except where noted:
1. Removing the heatsink module and repasting: the whole bottom cover, battery cable, fan cable, and heatsink screws must come off in order; screws must be tightened diagonally in stages, and uneven pressure causes poor conduction; using the wrong conductive paste, or letting it overflow, can short-circuit (third-party repair guides). Users without experience should send the machine in.
2. Soldered or ribbon-style fan connectors: a 4-wire fan carries power plus tachometer plus PWM speed control (iFixit forum experience; wire colours differ by maker, no confirmed universal standard). If a connector is soldered to the board or the ribbon is extremely fine, forcing it can tear board solder joints — listed as a send-in job. This point has no official source; it is repair common sense, marked as unconfirmed official wording.
3. Suspected heat-pipe failure: users cannot verify it; it needs professional inspection.
4. Opening a machine still under warranty: it may affect warranty terms — check the warranty terms before opening. A third-party retailer blog mentions that removing a heatsink can affect warranty cover, but this is not confirmed maker by maker officially, so this article writes "may affect."
What users can safely do: blow out the exhaust from outside with compressed air (HP staff advice), switch thermal modes, reset the EC, run official diagnostics, update the BIOS or drivers, and check Task Manager for abnormal processes.
Send-in checklist
If any of the following applies, send the machine in for inspection (priced by written quotation after inspection, with no price assumed in advance):
- Boot reports a CPU or System fan failure, or Dell Thermal Management has been disabled
- One side's fan never spins, or twitches and cannot get going, while the other side is normal
- Mechanical noise (grinding, clicking, scraping), especially pinned to one side
- Under heavy load the CPU or GPU repeatedly exceeds 90°C, throttles, restarts randomly, or shuts down from heat
- An exhaust that should be warm is not (suspected thermal-contact or heat-pipe problem)
- Still at full speed after an EC reset plus BIOS or driver updates plus thermal-mode switching
- The heatsink was removed before but never repasted, or repasted badly
- Still within the warranty period (check warranty terms first, to avoid losing cover by opening the machine)
Pricing
All repairs are priced by written quotation after inspection. Confirmed standard charges:
| Item | Charge |
|---|---|
| Battery | from HK$580 |
| Screen | from HK$1,280 |
| Board chip-level repair | from HK$880 |
| Keyboard | from HK$480 |
| Fan, thermal paste, heatsink module, and other repairs | By written quotation after inspection |
FAQ
Q1: The machine is still loud after cleaning — does that mean the last cleaning was not done properly?
A1: Not necessarily. This article's premise is that a proper cleaning has already been done. If the cleaning "never removed the fan itself and only blew air through the exhaust," it is true that some things can be missed: an internal cable touching a blade, or debris caught between the fan and the fin stack. But bearing wear, dried thermal paste, speed-control failure, and single-side fan failure are all advanced causes that cleaning cannot fix. So "still loud after cleaning" does not mean the last cleaning was poor — it means the problem has moved to a different level and needs to be separated out using this article's methods.
Q2: The fan bearing makes a strange noise — can I just put a drop of oil in it?
A2: This article does not recommend it. The folk remedy of peeling back the fan sticker and adding sewing-machine oil is mentioned in Dell community posts, but a Dell community reply also states plainly that even if it works, it is only short-term. Bearing wear is mechanical wear: oil may quiet the grinding briefly, but the wear itself does not heal, and in the end the fan still needs replacing. Do not open the machine yourself to poke at the bearing or pull cables either, as that risks damaging other parts. The right move is to reduce heavy-load use, get the machine inspected promptly, and have the fan priced by written quotation after inspection.
Q3: How do I tell zero-RPM mode apart from a real fault?
A3: Fans stopped at desktop idle, stopped under light load, or spinning briefly during the boot self-test and then stopping are mostly normal zero-RPM or power-saving behaviour. But if the GPU or CPU is already hot and a fan still reads 0 RPM; or one side's fan never spins while the other side does; or a fan twitches but cannot get going; or the problem comes with throttling, random restarts, or shutdowns, it points towards a real fault. The hardest evidence is a boot error: Dell BIOS reports "CPU fan failure" or "System fan failure" — if you see that, stop guessing and send the machine in.
Q4: Should I replace the fan or the whole heatsink module — and when must I send it in?
A4: If the fault is confirmed to be the fan itself — bearing noise, a broken blade, not spinning at all with power present, or a boot fan-failure error with an undeformed heatsink — replacing just the fan may be enough, because on many models the fan and heatsink are separate parts. But if a heat pipe is bent or blocked, or a crack or pinhole leak is suspected (exhaust not warm plus temperatures that will not come down), the whole heatsink and heat-pipe assembly needs replacing; ordinary users cannot verify a heat-pipe leak, so suspicion alone means sending it in for inspection. Send it in if any checklist item applies — boot fan-failure error, one side never spinning, mechanical noise pinned to one side, repeated 90°C-plus temperatures under heavy load, or no improvement after an EC reset and updates. All prices are by written quotation after inspection.
Need Professional Laptop Repair?
Laptop King has 15+ years of repair experience in Shek Mun, Sha Tin. Free diagnosis, transparent pricing, no fix no fee. 180-day warranty for your peace of mind.