Pourquoi les pads thermiques tombent en panne : causes courantes et solutions

Les coussinets thermiques tombent généralement en panne parce que l'interface perd un contact complet et stable. Le coussinet lui-même peut être endommagé, mais ce n'est qu'une possibilité parmi d'autres. Une mauvaise épaisseur, une compression insuffisante, une dureté inadaptée, un serrage irrégulier, un mauvais placement, une contamination et le vieillissement à long terme peuvent tous augmenter la résistance thermique.

Une valeur en W/mK plus élevée ne peut pas sauver un entrebâillement d'air. Elle ne peut pas non plus réparer un dissipateur thermique qui a été soulevé d'une puce voisine par un pad trop épais.

Ainsi, la question utile n'est pas simplement : “ Est-ce un mauvais pad thermique ? ” C'est :

Qu'est-ce qui a changé dans le trajet complet, de la source de chaleur à la structure de refroidissement, en passant par le pad ?

Ce chemin comprend le composant, le pad, le dissipateur thermique, le boîtier, les fixations, les tolérances, l'environnement de fonctionnement et le processus d'assemblage. Le dépannage fonctionne mieux lorsque tous ces éléments sont vérifiés ensemble.

Que signifie réellement une panne de coussin thermique ?

Un coussinet thermique est un matériau d'interface thermique préformé. Il comble l'espace entre un composant générateur de chaleur et une surface plus froide, telle qu'un dissipateur thermique, une plaque froide, un boîtier métallique, un blindage ou un châssis.

Le coussin remplace l'air emprisonné par un matériau qui transfère la chaleur plus efficacement. Il peut également assurer une isolation électrique, un amortissement, un contrôle des vibrations ou un placement facile en production. Le plus large de HAKTAK Guide des matériaux d'interface thermique explique comment les tampons se positionnent par rapport aux graisses, mastics, gels, matériaux à changement de phase, au graphite et aux adhésifs conducteurs.

Une défaillance signifie que l'interface ne remplit plus une ou plusieurs fonctions requises. Cela ne signifie pas nécessairement que la pastille se soit fissurée ou ait fondu.

Défaillance thermique

Une défaillance thermique peut se manifester sous les formes suivantes :

• Température de jonction, de boîtier, de mémoire ou de point chaud supérieure
• Limitation thermique
• Puissance de sortie réduite
• Des ventilateurs qui tournent plus vite qu'avant
• Arrêts imprévus
• Une grande différence de température entre des appareils similaires
• Un point chaud local alors que la température moyenne semble acceptable

Défaillance mécanique

Une panne mécanique peut inclure :

• Extrusion du pad depuis l'interface
• Déchirure, perforation ou dommage sur les bords
• Déformation permanente
• Perte de récupération après une compression prolongée
• Mouvement sous l'effet des vibrations
• Pliage de circuits imprimés
• Contrainte sur un composant ou un joint de soudure
• Déformation du dissipateur thermique ou du boîtier

Pannes électriques et de production

De nombreuses pastilles remplissent d'autres fonctions que la simple dissipation thermique. Une pastille peut assurer la dissipation thermique tout en présentant un défaut au niveau de l'assemblage sur un autre plan.

Voici quelques exemples possibles :

• Réduction de la marge diélectrique suite à une surcompression
• Crevaison sur le bord tranchant d'un composant
• Graphite conducteur exposé à proximité d'un circuit
• Contamination des contacts, des composants optiques ou du revêtement de protection
• Transfert de l'adhésif depuis la pellicule de protection
• Un tampon s'est déplacé hors de sa zone cible
• Une doublure de protection laissée en place
• Variation de positionnement à l'origine de rebuts de production

C'est pourquoi dire que “ la température semble correcte ” ne constitue pas un test d'acceptation complet.

Symptômes d'un mauvais pad thermique et ce qu'ils signifient généralement

Le tableau ci-dessous constitue un point de départ et non un diagnostic définitif. Plusieurs causes de panne peuvent entraîner le même symptôme.

Le moment d'apparition du symptôme est un élément utile. Un problème qui survient immédiatement après l'assemblage est généralement lié à la géométrie, à la compression, au positionnement ou à la fixation. Une dérive progressive est plus susceptible d'être due au vieillissement, à la déformation rémanente après compression, aux vibrations, à une exposition à des produits chimiques ou à un déplacement au sein de l'empilement.

Pourquoi une épaisseur inadaptée du pad thermique entraîne une surchauffe

Une épaisseur incorrecte est l'une des causes les plus courantes de défaillance des pastilles thermiques. Elle est également à l'origine de certains des symptômes les plus difficiles à interpréter.

L'écart nominal en CAO n'est pas suffisant. Les produits réels présentent des variations liées à la hauteur des composants, à l'épaisseur de la soudure, au gauchissement du circuit imprimé, à la planéité du boîtier, à l'usinage du dissipateur thermique, à la compression du joint et au couple de serrage des vis.

Une plaquette doit fonctionner aux écartements minimal, nominal et maximal.

Que se passe-t-il lorsqu’un coussin thermique est trop fin ?

Une plaquette trop fine risque de ne pas être en contact avec les deux surfaces. Même si elle est légèrement en contact dans des conditions nominales, elle peut perdre le contact lorsque l'écart est maximal.

Le résultat peut inclure :

• Poches d'air
• Traces de contact peu marquées
• Résistance d'interface élevée
• Lieux incontournables de la région
• Écart de température entre les appareils
• Contact intermittent en cas de vibrations

Ce défaut peut être difficile à détecter depuis le côté d'un ensemble. Un plot peut sembler être placé au bon endroit tout en n'exerçant pratiquement aucune pression.

Que se passe-t-il lorsqu'un tampon thermique est trop épais ?

Un coussinet plus épais n'est pas automatiquement plus sûr. Il crée un trajet thermique plus long et nécessite souvent plus de force pour atteindre l'épaisseur installée.

Un coussin trop épais peut :

• Augmenter la résistance thermique volumique
• Pliage d'un circuit imprimé
• Soumettre les soudures ou les boîtiers à des contraintes
• Déformer un boîtier fin
• Extruder autour des bords
• Réduction de la marge d'isolation électrique après une forte compression
• Prevent another component from touching the same cold plate

This last point is easy to miss.

Suppose a GPU cooler contacts the GPU die, memory chips, and power components. New memory pads are thicker or harder than the original parts. They make strong contact with the memory, but they hold the cold plate slightly above the GPU die. Memory temperature may improve while GPU core and hot-spot temperatures become much worse.

The memory pad looks like it is working. The complete cooling assembly is not.

For a practical selection method based on the full gap range, see HAKTAK’s guide to sélection de l'épaisseur du pad thermique pour l'électronique.

A Simple Thickness Example

Assume a 1.5 mm pad is installed across a gap that varies from 1.0 mm to 1.35 mm.

The same pad can be close to over-compressed in one unit and under-compressed in another. That is why one nominal percentage is not enough.

Under-Compression and Over-Compression Failure

Thermal pads need pressure to conform to surface texture and replace trapped air. More pressure is not always better, though. There is a useful operating window.

Under-Compression Leaves Hidden Air Gaps

Under-compression can happen when:

• The pad is too thin.
• The pad is too hard for the available load.
• Screw torque is too low.
• The housing flexes instead of loading the pad.
• The surfaces are tilted.
• The pad area is large relative to the total clamp force.
• Component heights vary more than expected.

Large pad area deserves attention. Pressure equals force divided by area. A pad that becomes twice as large needs twice the force to reach the same average pressure, assuming other conditions remain equal.

The interface may look neatly covered but still have poor surface wetting.

Over-Compression Creates Mechanical and Electrical Risk

Over-compression can temporarily improve contact. The thermal gain may become very small after full contact is reached, while mechanical risk keeps rising.

Possible results include:

• Extrusion de tampon
• Permanent compression set
• PCB bow
• Package cracking
• Solder-joint strain
• Fastener overload
• Housing distortion
• Puncture over sharp edges
• Reduced final dielectric thickness

A soft pad may tolerate high strain with modest force. A firm pad at the same strain may load the assembly much more heavily. Compression percentage cannot be reviewed without the force-deflection curve.

Calculate the Compression Window

The basic compression equation is:

Compression (%) = (épaisseur à la livraison – épaisseur après pose) / épaisseur à la livraison × 100

Run this calculation at:

1. Écart minimum
2. Écart nominal
3. Écart maximal

Then compare every condition with supplier compression data and the assembly’s force limit. HAKTAK’s detailed guide to taux de compression du pad thermique covers this process and the related dielectric concerns.

Pad Hardness and Conformability Can Break a Good Design

Thermal conductivity often gets the large print on a datasheet. Hardness and compression force may get a small graph near the back. In real products, that small graph can decide whether the thermal path works.

Why a High-Performance Pad May Be Too Hard

Highly filled materials can be firmer. That is not automatically bad. A firm pad can offer dimensional stability and easier handling.

Trouble starts when the product cannot provide enough pressure to compress it.

The pad may:

• Touch only surface peaks
• Fail to follow enclosure curvature
• Overload fragile components
• Hold a heat sink above another device
• Produce different results as screw torque varies

Shore hardness is useful for screening, but it does not fully describe assembly behavior. Two pads with similar Shore 00 values can have different stress-versus-strain curves.

Ultra-Soft Pads Have Their Own Problems

Very soft pads conform well at low pressure. They are useful over delicate components and wide height variation.

Still, soft material can be awkward. It may:

• Stretch during liner removal
• Tear around narrow features
• Move during assembly
• Extrude under high compression
• Stick to tools or gloves
• Lose dimensional accuracy in unsupported shapes

The best choice is not “soft” or “hard” by itself. It is the material that reaches full contact inside the allowed force window and remains stable through production and service.

Poor Contact, Surface Flatness, and Assembly Tolerance

A thermal pad can only conform so far. It cannot repair unlimited housing warp, a tilted heat sink, or a badly located fastener.

Rough, Warped, and Tilted Surfaces

Common causes of uneven contact include:

• Heat sink machining tolerance
• Cast housing distortion
• PCB warpage
• Component height differences
• Solder thickness variation
• Uneven gasket compression
• Thermal expansion mismatch
• A screw boss that acts as an early mechanical stop

If one side of a pad compresses by 30% and the other side barely touches, temperature will not be uniform.

Contact paper, pressure-sensitive film, temporary witness material, 3D measurement, cross-sectioning, or a controlled pad-imprint inspection can help reveal the pattern. Each method has limits, so it should be used carefully.

Uneven Screw Torque Can Tilt the Interface

Fastening one corner completely before the others can tilt the heat sink. It is a bit like tightening one leg of a wobbly table and hoping the other three follow.

A controlled sequence may include:

• Bringing all screws into light contact
• Tightening in a cross pattern
• Applying torque in two or more stages
• Using calibrated tools
• Recording final torque
• Adding mechanical stops where required

Torque is only an indirect measure of clamp force. Friction, thread condition, washer design, and housing stiffness also matter.

Pad Footprint and Misalignment

An undersized pad can leave part of the heat source uncovered. An oversized pad may add force over areas that do not transfer useful heat.

Poor geometry can also create:

• Interference with screw bosses
• Folded edges
• Material over connectors or test points
• Narrow sections that tear during liner removal
• Electrical-clearance problems
• Placement ambiguity for operators

Good die-cut design should follow the useful thermal overlap while allowing realistic placement tolerance.

And yes, check the liners. A clear PET liner can be surprisingly hard to notice. Leaving it in place adds a poorly controlled insulating layer and blocks the pad from conforming.

Why High W/mK Thermal Pads Can Still Fail

Thermal conductivity describes a bulk material property. It does not directly report the temperature of a GPU, MOSFET, battery module, LED, or controller.

The complete interface includes:

• Bulk resistance through the pad
• Contact resistance at both surfaces
• Épaisseur finale comprimée
• Effective contact area
• Heat spreading in the component and housing
• Airflow, liquid cooling, or natural convection after the heat leaves the interface

Thickness Can Outweigh a Conductivity Advantage

Ignoring contact resistance for a simple comparison: R = t / (k x A)

For the same contact area:

• A 10 W/mK pad at 2.0 mm gives a simplified t/k value of 0.20.
• A 5 W/mK pad at 0.8 mm gives a simplified t/k value of 0.16.

In this example, the lower-conductivity pad has lower bulk resistance because it is much thinner. Real interfaces also contain contact resistance, so this is not a product ranking. It is a reminder to check the whole geometry.

Compare Thermal Impedance at Realistic Pressure

Thermal impedance data can be more useful than W/mK when the test thickness, pressure, temperature, and surface conditions are known. HAKTAK’s article on conductivité thermique en fonction de l'impédance thermique explains why those conditions must travel with the reported value.

A pad tested at high laboratory pressure may look excellent. The same material can perform poorly in a lightweight plastic enclosure with low clamp force.

This is also why high W/mK does not always mean better cooling performance. The label matters, but the installed interface matters more.

Thermal Pad Aging, Compression Set, and Loss of Contact

Some failures are present on the first day. Others develop slowly.

A pad can pass initial thermal testing and still lose contact after heat aging, thermal cycling, vibration, chemical exposure, or years under compression.

Compression Set

Compression set is the permanent deformation left after a material has been compressed for a defined time and then released.

In an assembly, high compression set can contribute to:

• Lower recovery
• Reduced contact pressure
• Greater sensitivity to tolerance or housing movement
• Higher thermal resistance
• Poor rework behavior
• Local hot spots after aging

The problem is not always visible. The pad may remain in place but no longer push firmly against both surfaces.

Fastener relaxation and gasket aging can make this worse. Several parts in the stack may lose load at the same time.

Hardening, Softening, and Chemical Change

Long exposure to heat can change polymer behavior. The direction and amount depend on the formulation.

A pad may:

• Become harder
• Become softer
• Lose elasticity
• Change surface tack
• Show increased oil bleed
• Swell after fluid exposure
• Lose adhesion to a carrier or reinforcement

Materials near oil, fuel, coolant, cleaning agents, plasticizers, or process chemicals need compatibility testing with the actual fluid. A generic label such as “chemical resistant” is too broad for reliable design.

Thermal Cycling and Power Cycling

Semiconductor packages, copper, aluminum, PCB laminates, solder, and polymers expand at different rates. Every heating and cooling cycle moves the stack a little.

Repeated movement can:

• Change pressure distribution
• Rub or shear the pad
• Increase edge damage
• Relax fasteners
• Open a small contact area
• Shift a poorly retained pad
• Stress adhesive or reinforcement layers

Solid pads are generally less prone to classic grease pump-out. They can still lose effective contact through mechanical movement and compression set.

Vibration and Shock

Automotive electronics, industrial equipment, rail systems, outdoor telecom hardware, and mobile devices experience vibration or impact.

Vibration can reveal:

• Inadequate pad retention
• Abrasion against sharp edges
• Fastener relaxation
• Housing movement
• Weak adhesive
• A design with almost no compression margin

The temperature may remain stable during a short bench test and drift only after environmental exposure.

Oil Bleed, Contamination, Delamination, and Adhesive Failure

Not every thermal pad failure is a simple thickness problem.

Is Oil Bleed Always a Failure?

Some silicone-based pads can show a small amount of surface oil or low-molecular-weight material. The acceptable amount depends on the formulation and application.

Minor bleed does not automatically prove poor heat transfer. However, migration can become a serious concern near:

• Electrical contacts
• Relays
• Ensembles optiques
• Cameras and sensors
• Conformal coatings
• Paint or bonding surfaces

Acceptance should follow supplier limits and application testing, not a quick visual opinion.

Dust, Fingerprints, and Cleaning Residue

Soft, tacky surfaces attract contamination. Dust and fibers can create local contact defects. Oil from handling can interfere with adhesion. Aggressive cleaning residue may change the surface or attack nearby materials.

Good handling controls may include:

• Keeping liners in place until assembly
• Holding pads by liner tabs or edges
• Using clean gloves or tools
• Protecting cut parts in sealed packaging
• Controlling cleaning chemistry
• Separating silicone-free materials from silicone-contaminated tools where necessary

Pressure-Sensitive Adhesive Can Add Another Failure Layer

Adhesive backing helps placement, but it also adds thickness and another interface.

Possible problems include:

• Adhesive transfer to the liner
• Uneven tack
• Air trapped during placement
• Edge lifting
• Poor adhesion after heat or humidity
• Residue during rework
• Higher thermal resistance

Full-area adhesive should not be added by habit. Natural tack, selective adhesive, split liners, or mechanical retention may work better in some designs.

Reinforcement and Laminate Delamination

Pads may contain fiberglass, PET, PEN, graphite, or other carrier layers. Reinforcement improves handling and puncture resistance, but a layered construction introduces more interfaces.

After heat, humidity, chemical exposure, or repeated flexing, inspect for:

• Separation between layers
• Bubbles
• Cracked edges
• Wrinkles
• Carrier exposure
• Loss of electrical isolation

Thermal Pad Failure Modes Change by Material

The phrase “thermal pad” covers several material families. Their risks are not identical.

For conventional compressible sheet interfaces, HAKTAK’s silicone thermal pad range provides options around thickness, hardness, tack, dielectric performance, and custom die-cut formats.

When component heights vary too much for one sheet thickness, thermal putty versus thermal pad selection becomes relevant. A dispensable thermal conductive gap filler may also reduce assembly stress in complex or large-area interfaces.

The alternative still needs validation. Changing material families replaces one set of risks with another; it does not remove engineering work.

Thermal Pad Failures by Application and Industry

The same basic physics applies across industries, but the most important failure mode changes with the product.

GPUs, AI Servers, and High-Power Accelerators

These assemblies may use pads on memory, power stages, controllers, and heat spreaders. Several components often share one cold plate.

Key risks include:

• Wrong replacement thickness
• Hard pads lifting the cold plate from a GPU or accelerator package
• Incomplete VRAM coverage
• Uneven screw torque
• Service technicians mixing several thicknesses
• High local heat flux hiding behind an acceptable average temperature

Monitor core, hot-spot, memory, inlet, and coolant temperatures where available. One sensor cannot tell the whole story.

MOSFETs, IGBTs, and Power Modules

Power electronics often need heat transfer and electrical insulation at the same interface.

Watch for:

• Puncture over package edges
• Reduced dielectric thickness after compression
• High clamp force
• Planéité du radiateur
• Power-cycling movement
• Local pressure around screws

Electrical testing should be repeated after mechanical and environmental stress, not only on fresh material.

EV Batteries and BMS Electronics

Battery systems may have large interfaces, multiple height levels, flexible trays, and strict stress limits.

Common failure paths include:

• Large total compression force
• Cell or PCB loading
• Coolant exposure
• Gap variation across a wide area
• Vibration
• Assembly movement over long service life

A material selected from one small test coupon may behave differently over a large battery module.

Automotive ECUs and ADAS Modules

Automotive housings face temperature cycling, vibration, humidity, oils, cleaners, and long qualification periods.

Traceability matters here. A production failure may come from a material lot, a housing change, screw-torque drift, or a new liner process. Root-cause analysis should preserve all four possibilities.

LED Modules

LED reliability is strongly linked to temperature. Pads or insulating sheets may sit under metal-core boards, drivers, or housings.

Uneven board contact can create local hot spots even when case temperature looks reasonable. Optical contamination and long heat exposure can also influence material choice.

Telecom, Industrial, and Outdoor Electronics

These products may run continuously in sealed or partly sealed enclosures.

Relevant concerns include:

• Humidité
• Dust
• Continuous high temperature
• Enclosure distortion
• Long service intervals
• Limited maintenance access

A low-cost pad replacement can become an expensive field visit. Long-term validation matters more than saving a few minutes in the lab.

How to Find the Root Cause of a Thermal Pad Failure

Replacing the pad immediately may hide the evidence. A controlled workflow is more useful.

Step 1: Confirm the Thermal Symptom

Compare the unit under consistent conditions:

• Ambient and inlet temperature
• Power or workload
• Fan or pump speed
• Firmware and control settings
• Heat sink cleanliness
• Sensor location
• Test duration

A dirty heat sink or changed fan curve can look like a pad failure.

Step 2: Map the Cooling Path

Write down the path:

Component -> thermal bloc-notes -> spreader or housing -> heat sink -> air or coolant

Mark every contact and every possible bottleneck. The pad may not be the dominant resistance.

Step 3: Inspect Contact Evidence

Before cleaning the assembly, photograph:

• Pad position
• Compression marks
• Core paste imprint
• Torn or folded edges
• Oil or residue
• Liner condition
• Screw locations
• Areas with no visible contact

Poor paste imprint on a neighboring die is strong evidence that pad thickness or stiffness has changed the stack.

Step 4: Measure the Real Stack-Up

Measure or calculate:

• Minimum, nominal, and maximum gap
• Supplied pad thickness
• Épaisseur comprimée
• Component height
• Housing and heat sink flatness
• PCB bow

Avoid measuring a soft used pad with uncontrolled finger pressure. That number can be nicely precise and still wrong.

Step 5: Review Force and Fastening

Vérifier :

• Screw torque
• Tightening sequence
• Clip or spring load
• Thread condition
• Mechanical stops
• Gasket interaction
• Total pad area

A torque change can affect both thermal contact and board stress.

Step 6: Separate Cause Categories

This separation prevents every problem from becoming a supplier complaint before the assembly has been checked.

Step 7: Change One Variable at a Time

Do not change the pad brand, thickness, paste, screw torque, fan curve, and heat sink at the same time. The new temperature may improve, but the reason will remain unknown.

Use controlled A/B samples where possible.

Step 8: Verify the Fix After Stress

A corrective action should survive:

• Heat aging
• Cyclage thermique
• Mise hors et sous tension
• Vibration
• Humidity or fluid exposure where relevant
• Reassembly or service handling

Initial temperature is necessary. It is not lifetime evidence.

Thermal Pad Testing Standards and Reliability Checks

No single test certifies the complete thermal pad interface.

ASTM D5470 for Thermal Impedance

ASTM D5470-17(2024) covers steady-state thermal impedance measurement and calculation of apparent thermal conductivity for thermally conductive electrical insulation materials.

It is useful, but the standard itself notes that idealized test conditions do not directly match most applications. Pressure, thickness, temperature, specimen area, and surface conditions should be reported with the result.

Hardness and Compression Set

ASTM D2240 is commonly used for durometer hardness. ASTM D395 addresses compression set for rubber materials.

These properties support material comparison, but they do not replace a full force-deflection curve or assembled validation. A small coupon in a laboratory does not know that a real PCB is thin, warped, and loaded by four screws.

Environmental Testing

IEC 60068 methods are often used to structure temperature change, vibration, shock, damp heat, and related environmental tests.

The exact profile should match the product. Five mild cycles do not prove suitability for an automotive module expected to survive years of repeated thermal and mechanical stress.

Flammability and Dielectric Requirements

UL 94 classifications describe material flammability behavior under defined tests. They do not prove low thermal impedance, compression stability, or electrical insulation in a compressed and aged assembly.

Dielectric testing should consider:

• Épaisseur finale comprimée
• Edge geometry
• Puncture risk
• Humidity and contamination
• Thermal aging
• Mechanical cycling

Test the Complete Assembly

Official AMD guidance for thermal interface materials emphasizes spreading, mounting pressure, and elimination of air gaps. Those factors explain why material data must be connected to the actual cooling assembly.

Reliability should also measure thermal performance after exposure, not merely confirm that the pad remains visible. A Parker Chomerics gap-pad reliability report illustrates this approach by tracking thermal impedance through thermal shock, thermal cycling, and vibration stages.

For a broader map of test methods, HAKTAK’s guide to normes communes de test TIM explains why the method and test conditions must be reported beside the value.

Common Thermal Pad Failure Causes and Solutions

When to Replace the Pad and When to Redesign the Interface

Replace the pad when it is:

• Torn
• Folded
• Punctured
• Contaminated
• Permanently deformed
• Missing identification
• The wrong grade or thickness
• Removed from an assembly without an approved reuse process

Redesign the interface when:

• No single pad thickness covers the full tolerance range.
• Required compression overloads the PCB or component.
• The heat sink remains tilted or warped.
• A pad is being used to compensate for a structural problem.
• Field failures return after repeated pad replacement.
• The interface needs both very low stress and large gap coverage.

Repeatedly changing pads in a bad mechanical stack is like putting new tires on a bent wheel. The new part may help for a while, but the underlying geometry is still waiting.

Conclusion

Thermal pads rarely fail because of one number on a datasheet. They fail as part of an interface system.

Wrong thickness can leave an air gap or lift a shared heat sink away from another chip. Poor compression can reduce contact. Excessive compression can damage the pad, PCB, component, or dielectric path. Aging, vibration, chemicals, contamination, adhesives, and assembly variation can then change a design that looked fine on day one.

The practical solution is to diagnose the entire heat path. Measure the real gap. Check minimum, nominal, and maximum compression. Review hardness and force. Inspect contact evidence. Control placement and torque. Then test thermal, mechanical, and electrical performance after realistic environmental stress.

The best thermal pad is not simply the one with the highest W/mK. It is the one that keeps reliable contact without overloading the product through its intended life.

Foire aux questions

What Are the Signs of a Bad Thermal Pad?

Common signs include rising component temperature, thermal throttling, high fan speed, inconsistent temperatures between units, poor contact marks, cracking, tearing, hardening, extrusion, oil migration, or loss of elasticity. These symptoms can also come from incorrect thickness, poor fastening, dust, airflow, or heat sink problems, so the complete cooling system should be checked.

Can Thermal Pads Cause Overheating?

Yes. A pad can cause overheating when it is too thin to make contact, too thick to fit the stack, too hard to compress, badly positioned, contaminated, or installed with a liner still attached. Thick or firm pads can also lift a shared cold plate away from a CPU or GPU die.

What Happens If a Thermal Pad Is Too Thick?

An overly thick pad can increase thermal resistance and assembly force. It may bend the PCB, stress components, squeeze out around the edges, reduce dielectric thickness, or prevent a neighboring component from touching the heat sink. The result can be worse temperatures even when the new pad has a higher conductivity rating.

What Happens If a Thermal Pad Is Too Thin?

A pad that is too thin may not contact both surfaces, especially at the maximum production gap. This leaves insulating air pockets and creates hot spots. Contact can also become intermittent during vibration or thermal cycling.

How Much Should a Thermal Pad Compress?

There is no universal compression percentage for every pad. The correct value depends on thickness, hardness, force-deflection behavior, surface flatness, pad area, gap tolerance, and component stress limits. Compression should be calculated at minimum, nominal, and maximum gap and compared with supplier data.

Do Thermal Pads Lose Effectiveness Over Time?

They can. Heat, compression set, thermal cycling, vibration, humidity, chemical exposure, and fastener relaxation may reduce contact or change material behavior. A pad can remain visually intact while its installed thermal impedance increases.

Is Oil Coming from a Thermal Pad a Sign of Failure?

Not always. A small amount of bleed can be formulation-dependent. Excessive migration may still contaminate contacts, optics, coatings, or bonding surfaces. Acceptance should follow supplier limits and thermal, electrical, and contamination testing for the application.

Can Two Thermal Pads Be Stacked?

Stacking creates an extra interface where air, misalignment, and contact resistance can develop. It also makes compression less predictable. A single correctly selected pad is normally preferred. Stacking should only be used when the supplier and product validation support it.

Can a Removed Thermal Pad Be Reused?

Reuse is generally risky. A removed pad may be torn, contaminated, stretched, or permanently compressed. Its final thickness and pressure may no longer be repeatable. Production, high-power, and reliability-critical equipment should normally receive a new specified pad.

Why Did GPU Temperatures Rise After Replacing Thermal Pads?

The replacement pads may have the wrong thickness, hardness, or position. They may be holding the cooler away from the GPU die, or they may not contact the memory and power components. Check the GPU paste imprint, pad compression marks, screw sequence, all liners, and the exact original thickness requirements before trying another material.