GPU Capacitor Bulging & Leaking: Electrolytic Aging, ESR, Ripple & Replacement Diagnosis

2026-10-19 ｜ GPU Repair ｜ 30 min ｜ HK GPU Fix Editorial

Key Takeaways

A bulging capacitor has built up internal gas pressure and is losing its ability to filter power cleanly. Visual signs help, but measurement and rail-level testing confirm how far the fault goes. Replacement is parameter matching and board-level work, not a same-value swap.

What a Capacitor Does in the GPU Power Path

A graphics card's voltage regulator module, usually called the VRM, is a switching buck converter. It is built from a PWM controller, MOSFET switches, inductors (chokes), and capacitors. It chops the incoming roughly 12 V supply into pulses and smooths them into the low, stable voltage the GPU needs. Capacitors in this circuit store charge and filter out the remaining voltage ripple before power reaches the chip. Input capacitors stabilise the input and filter high-frequency noise. Output capacitors smooth the DC output and supply instantaneous current during transient load spikes.

Multi-phase VRMs split the load across several identical phases. That reduces heat per component and smooths the output. It also matters for understanding stress: capacitors sitting beside the hardest-worked phases, and beside MOSFETs, chokes, and heatsinks, live in the worst thermal environment on the card. This article does not rank which capacitors fail first on any particular card model. No model-specific failure ranking was found in the sources used here. The point is a general engineering one: heat and ripple exposure drive ageing, and the power path is where both are concentrated.

Two Capacitor Families on the Board

Graphics cards use more than one kind of capacitor. The distinction matters because the two main families age in different ways.

Wet aluminium electrolytic capacitors

A standard aluminium electrolytic capacitor, often called a wet capacitor, has two aluminium foil layers with a liquid electrolyte between them. The liquid is part of how the capacitor works. Over time, especially at elevated temperature, the liquid can dry out. As it does, capacitance falls and ESR rises sharply.

Polymer / solid capacitors

A polymer capacitor, often called a solid capacitor, uses a solid conductive polymer impregnated into the separator paper instead of a liquid electrolyte. This gives dramatically lower ESR and virtually eliminates the drying problem. At identical operating temperatures, lifespan is significantly longer, according to the technical explainer used for this article.

The word "solid" needs care. In marketing, "solid capacitor" usually means a solid or polymer aluminium capacitor. Tantalum capacitors are a different solid family with different failure behaviour. This article says "solid/polymer aluminium" when it means that family, and treats tantalum separately where relevant.

The trade-off between the families is qualitative, not a priced comparison. Electrolytics pack more capacitance into a given area at low cost. Polymer/solid types have much lower ESR and better heat transfer, but are bulkier per unit capacitance and cost more. Tantalum chip types are small with fairly low ESR but cost a lot. All types can fail.

A polymer capacitor is often described as having a flat top, which can help distinguish it at a glance from a scored-vent electrolytic. That is a type cue, not a health certificate. Solid capacitors can still fail, as explained in the misconceptions section below.

ESR and Ripple: The Two Ideas That Explain Most Failures

ESR

ESR stands for Equivalent Series Resistance. It is the parasitic resistance in series with the ideal capacitance of a real capacitor. High ESR turns energy into heat and degrades filtering.

For CPU and GPU power stages, one technical explainer gives 5–20 mΩ as illustrative "low" ESR values, and says a failing electrolytic can reach 500 mΩ or more, enough to cause instability. Those figures are presented here as that source's illustrative ranges, not as a universal specification for every graphics card. The same source offers a rule of thumb: measured ESR more than 3–5 times the datasheet value indicates a degraded capacitor that should be replaced. Other sources describe faulty capacitors as typically showing 2–4 times their rated ESR. These are overlapping rules of thumb from different sources, not a single hard threshold, and there is no verified ESR table for a specific GPU model in the sources used here.

Ripple

Ripple current is the AC component flowing through the capacitor while it filters a DC rail. In a buck converter, the ESR contribution to output ripple can be expressed as ΔV_ESR = ΔI × ESR. That relationship is the key to the whole failure story: as ESR rises, the same ripple current produces more ripple voltage on the rail, and more heat inside the capacitor.

Why Wet Electrolytics Bulge and Leak

Bulging and leaking are not a single-cause event. They are the visible end of a chain. The order matters.

1. Heat: ambient heat plus self-heating

Heat is described in the sources as the biggest driver of bulging. Two heat sources combine. The first is ambient heat from the card and the case. The second is self-heating from ripple current flowing through the ESR. Mounting a capacitor near heatsinks, power MOSFETs, or transformers can significantly shorten its life even if datasheet ratings look acceptable.

A widely repeated rule of thumb connects temperature and lifespan: roughly, every 10 °C rise above rating halves expected lifespan. This rule appears in more than one of the sources used here. It is a rule of thumb, not a precise prediction for any individual card.

2. Ripple-current overload

Exceeding the rated ripple current causes internal self-heating, accelerates electrolyte evaporation, and shortens lifespan. In switching supplies and DC-DC converters, ripple-current overload sets up a destructive feedback loop: internal heating degrades the electrolyte faster, ESR rises, and the higher ESR produces still more heating.

3. Ageing and electrolyte loss

Over time the electrolyte dries out, especially at high temperature. Capacitance falls, ESR rises, losses rise, the capacitor heats more and degrades faster, and at end of life it goes open-circuit as the dielectric dries up. Bulging and venting appear in this sequence as pressure builds.

4. Gas generation, pressure, and the vent

Electrolyte decomposition produces gas. In laboratory analysis of failed capacitors from the historical capacitor plague episode, hydrogen gas was identified in failed caps. Pressure builds until the scored vent on the can top — usually a cross- or K-shaped score — opens. The vent is designed to split at its seams and relieve pressure rather than allowing pressure to rupture the can.

5. Overvoltage and reverse polarity as accelerants

Overvoltage and reverse polarity are fault or installation causes, not normal ageing, and this article keeps them separate.

Overvoltage damages the oxide dielectric, which leads to leakage current, heating, gas generation, bulging, and in extreme cases vent rupture. Reverse polarity causes rapid dielectric breakdown, gas generation, and venting or violent failure. One exam-preparation source states that even roughly 1–1.5 V of reverse voltage can destroy the dielectric. That figure is cited as that source's statement. The core point is consistent across the sources: electrolytic capacitors are polarised, and reversed installation is a hard error, not a minor mistake.

A note on the historical "capacitor plague"

From 1999 to 2007, there was a higher-than-expected failure rate of non-solid aluminium electrolytic capacitors, caused by a faulty electrolyte composition that led to corrosion and gas generation. The episode was especially visible in motherboards, video cards, and power supplies. Laboratory analysis found hydrogen gas and an absence of the stabilising and depolarising compounds that normally bind excess hydrogen.

The well-known story about a stolen electrolyte formula is reported in the sources as attributed and claimed, not proven. Another source notes that the real picture probably also involved low-ESR designs, hotter and more power-hungry PCs, and counterfeits. This article therefore treats the plague as a historical manufacturing-quality episode. A modern bulged capacitor is far more likely to be ordinary heat, ripple, and ageing. Appearance alone does not diagnose "plague" capacitors.

Visual Signs vs Lookalikes

Visual inspection is the first step in diagnosis. It is necessary but not sufficient. The sections below separate genuine failure signs from the manufacturing materials that commonly confuse them.

Signs of a genuinely failed or failing wet electrolytic

A healthy cylindrical electrolytic top is flat, clean, and undeformed. The top often carries a cross-shaped or similar score that acts as a safety valve. The failure signs are:

• Domed or bulged top. A bulged, inflated, or dome-shaped top means gas has built up inside and deformed the metal cover. The component no longer meets its specifications.
• Vent split or cracking. The vent is a stamped impression forming seams, designed to split and relieve pressure rather than allowing pressure to rupture the can. Bulging or cracking of the vent is a failure sign.
• Casing sitting crooked. The bottom rubber plug can be pushed out by internal pressure, tilting the casing on the board.
• Electrolyte residue. Leakage may appear as wet, sticky residue, or over time as a dry, crusty residue around the base or oozing from the top. Colour is usually brown, yellowish, or reddish. Dried electrolyte residue is described in the sources as reddish-brown and as crusty rust-like brown deposits. The material is corrosive and can damage the board, traces, and nearby components. Leaked electrolyte is also conductive and can create unintended leakage paths between traces.
• Casing damage. Cracks, fissures, or breaks in the casing. In extreme cases there can be partial ejection of the casing if the vent fails to open.
• Surrounding board clues. Discoloration, burn marks, or darkened PCB around the capacitor indicate prolonged overheating or short circuits. Rust or corrosion on leads or the metal base suggests moisture or leaked electrolyte.

Lookalikes: glue, flux, and why colour alone does not decide

A board that looks "dirty" is not automatically a board with leaking capacitors. Two manufacturing materials are routinely mistaken for electrolyte.

Fixing glue and silicone sealant. White rubbery material around a radial capacitor is commonly silicone sealant or glue deliberately used to hold heavier components against vibration. Other adhesives used in electronics include epoxy, silicone, thread-locker, and hot-melt. Their purpose is to keep large parts in place during assembly and handling and to damp vibration.

Aged glue. Glue darkens with age, absorbs moisture, and can turn mildly conductive and corrode resistor leads and copper traces. Brown "bond glue" has been blamed in vintage equipment for eating component legs and traces. Aged glue is therefore not "always harmless" — but it is still not electrolyte, and its presence alone does not prove the capacitor has leaked.

Flux and rosin residue. Solder-side and component-side rosin and resin residue is a normal manufacturing leftover. It traps moisture and dirt, creates leakage paths, and hides cracked solder joints. Residues split into non-polar types (rosin, resin, glue), polar or ionic types (flux activator halogens, salts), and weakly polar types (organic acids and bases). Flux residue can leave an ionic film that becomes a leakage path.

How to tell them apart is a matter of inspection judgement, not a laboratory test. The following clues help, but none of them settles an ambiguous case on its own:

The decisive step for an ambiguous case is electrical testing, described below, or professional inspection. The article states this plainly: if it looks like glue, that does not prove the capacitor is fine.

Dust is not treated here as a lookalike with a cited diagnostic mechanism. No source in the research established that dust mimics electrolyte residue. Dust can be blown or brushed away before inspection as general housekeeping. It should not be presented as a diagnostic confuser with a mechanism.

Symptoms: Generic, Multi-Causal, and Why That Matters

A degraded capacitor filters worse. High ESR means more ripple voltage on the rail, voltage drops, reduced efficiency, and heat. In a VRM output, that means the GPU gets a noisier, less stable supply.

The symptoms that can accompany capacitor degradation include random crashes or freezes, especially under load, blue screens, restarts, graphics artefacts, flickering, or crashes during gaming or video playback, and difficulty booting or POST errors. A severely degraded capacitor bank can leave the card unable to power up cleanly. In rarer, catastrophic cases, dielectric breakdown can create a direct short, which usually trips fuses or protection.

Every item on that list is generic. Many other faults — GPU chip faults, memory faults, driver faults, power-supply faults, overheating from other causes — produce the same list. This article states that explicitly: symptom overlap is the reason diagnosis is a process, not a glance. No symptom on this list proves a capacitor fault. A bulged capacitor is a fault in its own right, but it is not automatically proof that it is the card's only fault.

Diagnosis Logic: What Each Check Can and Cannot Prove

Diagnosis should follow a sequence: power off, visual inspection, symptom context, electrical screening, then professional confirmation. Always disconnect power before inspecting. Discharge before connecting any meter. A charged capacitor can damage a meter, so discharge must be verified with a voltmeter first.

Step 1: Visual inspection

Visual inspection can identify bulging, venting, and leakage. It is necessary but not sufficient. Capacitors that look visually perfect may still have high ESR, detectable only by measurement.

Step 2: Ordinary multimeter and capacitance meter

A standard DC multimeter cannot measure ESR, because steady DC cannot pass through a capacitor. An ordinary ohmmeter test and a capacitance-meter test will miss a high percentage of bad capacitors, according to workshop experience reported in the sources. That is presented here as workshop experience, not as a measured percentage.

A capacitance meter needs the capacitor disconnected from the circuit. Capacitance can still read near-normal while ESR is already badly degraded. ESR is often the first parameter to change as a capacitor ages.

Step 3: ESR meter — the right screening tool, with real limits

A purpose-built ESR meter measures ESR, usually without disconnecting the capacitor. Practical meters use a test voltage too low to switch on semiconductor junctions in the circuit, which is what makes in-circuit reading possible. The test signal is typically around 50–200 mV peak-to-peak at 100 kHz, according to one technical source.

What an ESR meter can show is straightforward: a high ESR reading is strong evidence the capacitor is degraded. One explainer puts it plainly: if the ESR reads high, the capacitor is almost certainly bad.

What it cannot prove is equally important, and all of the following limits must be understood before trusting a reading:

1. It does not measure capacitance. ESR and capacitance are different parameters. A good ESR reading says nothing about capacitance.
2. A short-circuited capacitor reads as falsely low ESR. A short looks like an "ideally low" ESR, which is a false "good". Pair the ESR check with an ohmmeter or multimeter check for shorts.
3. ESR depends on temperature and applied voltage. A capacitor with excessive ESR at operating temperature and voltage may test as good when cold and unpowered. Freeze spray is one workshop technique for intermittent capacitors, mentioned here only as a technician technique, not as do-it-yourself advice.
4. Parallel capacitors mask each other. The meter reads their combined ESR, which is lower than any individual capacitor's ESR. Nearby low-value resistors can also skew readings. A normal in-circuit reading does not fully exclude a masked fault. Desoldering one leg and testing out-of-circuit gives a definitive answer.
5. In-circuit accuracy is screening-grade, not lab-grade. For aluminium electrolytics in power-supply circuits, in-circuit readings are usually within 10–30% of the true out-of-circuit value, according to one vendor technical article. That is adequate for pass or fail decisions because faulty ESR is typically 2–4 times rated. Precise measurement needs out-of-circuit testing under specified conditions.
6. A charged capacitor can damage the meter. Verify discharge with a voltmeter first.
7. Readings are frequency-dependent. Dedicated meters typically run at 100 kHz, while aluminium electrolytic datasheet ESR is usually specified at 120 Hz. A meter reading is therefore a screening comparison — against a known-good unit or a datasheet limit — not a datasheet re-measurement.

What must not be presented as diagnosis

• "Bulged capacitor" does not mean "the card's only fault."
• "No bulge" does not mean "the capacitors are fine." Visually perfect capacitors can have high ESR.
• "Replaced the bulged capacitor and the card works" does not prove the diagnosis was complete. A single repair outcome says nothing about other capacitors on the same rail.

Noise: Coil Whine vs Capacitor-Related Sound

Noise alone does not diagnose a capacitor. The distinction matters because the most common high-pitched sound from a graphics card has a different cause.

Ordinary coil whine is normally inductor vibration through magnetostriction and, less commonly, piezoelectric vibration of ceramic MLCC capacitors. It is usually an acoustic nuisance, not by itself proof of a defect. Power-supply quality can worsen it. Coil whine is rarely indicative of a functional defect and generally does not affect performance or lifespan, according to the explainers used here.

Capacitor-related noise is different. One explainer notes that faulty capacitors on any board can emit a high-frequency squeal when failing, which is less common but potentially serious because capacitor failure can lead to instability or permanent damage. That source is low-authority and is hedged accordingly here. A bulging or venting capacitor may also hiss or pop.

The article-safe framing is this: the distinguishing question is what else is present. A card that only whines under high frame rates, with no other symptom, is a different story from a card that whines and crashes and shows a domed capacitor. Do not write "coil whine means capacitor failure," and do not write "capacitor failure sounds like coil whine."

Replacement Principles

Parameter matching: ESR first, ripple second, capacitance last

Substitution order matters. One distributor technical article sets the order as follows: ESR first, set by the ripple-voltage budget through ΔV_ESR = ΔI × ESR; ripple-current rating second; capacitance last. Substitute capacitance within ±20% of the original, according to that source's guidance. The same source warns that a substitute with 20% higher capacitance but 30% higher ESR is not an equivalent substitute.

The widely repeated workshop rules of thumb line up with that logic:

• ESR: same or lower.
• Ripple-current rating: same or higher.
• Voltage rating: same or higher, without going absurd. Higher-voltage capacitors are usually bigger and more expensive.
• Capacitance (µF): do not change unless necessary.
• Temperature rating: match or exceed the original. Replacing with a 105 °C rated part when in doubt is widely advised. The 105 °C label is a rating, not a guarantee. One tutorial's statement that a 105 °C rating usually indicates a premium grade with lower ESR is a heuristic, not a guarantee.
• Physical fit: the replacement must fit the diameter and height available. Check size before ordering.
• VRM and switching-regulator positions need ultra-low-ESR types. General-purpose capacitors may be inadequate there.
• ESR is not usually printed on the capacitor body. It must be looked up in the datasheet for the exact series.

Polarity: the hard safety rule

Electrolytic capacitors are polarised. The negative side is marked with a stripe on the body for radial parts, and with a stripe, dot, or notch for surface-mount parts. The PCB silkscreen shows a positive mark or shaded outline. Reversed installation leads to dielectric breakdown, rapid gas generation, and venting or violent failure. Polarity checking is a non-optional step. This article does not present capacitor replacement as a casual first do-it-yourself job.

Why this is board-level work

Capacitors are soldered to thick copper power planes that act as powerful heatsinks. A low-power iron will struggle. One workshop account says a temperature-controlled iron may be needed for motherboard power planes. Graphics card boards are multi-layer boards with the same power-plane problem. The specific wattage for GPU boards was not verified in the research, so this article phrases it as a general point — multi-layer boards with heavy power planes need serious desoldering heat — and cites the motherboard account as an analogy rather than a GPU specification.

Other reasons this is board-level work:

• A wrong substitute — too-high ESR, too-low ripple rating, wrong polarity — can make filtering worse or fail violently.
• Leaked electrolyte is corrosive and conductive. Board cleaning and trace inspection are part of the job, not just swapping the can.
• A shorted or open companion component, such as a MOSFET, choke, or controller, will not be fixed by a capacitor swap.

Replace one, or the group?

Workshop practice quoted in the sources says: "Replace all of the same brand as the bad one. If it is on its last leg the others are not far behind." That is forum advice, not a measured rule. The article-safe phrasing is this: a technician will often assess the whole rail or bank, because capacitors on the same rail have shared the same heat and ripple stress — but whether to replace one or several is a judgement call for the repairer after testing, not an automatic rule. Do not write "always replace all capacitors," and do not write "replacing one always fixes it."

What replacement cannot do

If the GPU chip itself, memory, MOSFETs, or PCB traces are damaged, capacitor replacement alone will not restore the card. If electrolyte has corroded traces or caused shorts, cleaning and trace repair may be needed beyond the swap. If the root cause is external — poor case airflow, or a failing power supply delivering dirty power — swapping capacitors without addressing the cause invites repeat failure. That is presented here as reasoning from the heat and ripple chain and the power-supply stress point, hedged as reasoning rather than as a measured rate.

No claim is made here about typical repair cost, turnaround time, or outcomes for capacitor replacement on graphics cards. No such figures were verified in the research. For cost, this article states only that professional inspection is charged under the site's published three-tier range HK$400–2,800, with the actual fee set by the post-inspection quote.

Prevention: Risk Reduction, Not Guarantees

None of the measures below carries a measured prevention-effectiveness figure. All of them are risk reduction, not guarantees.

• Heat. Keep the card and case cool. Heat is the biggest bulging driver, and capacitors near heatsinks, MOSFETs, and transformers age faster. Practical corollaries, stated as general practice rather than measured claims: keep good case airflow, do not block intake or exhaust, and keep the PC in a ventilated spot.
• Dust and cooling hygiene. Dust build-up insulates heatsinks and fans and raises temperatures. This article does not claim that dust directly causes capacitor bulging, because no source in the research measured dust as a capacitor-specific mechanism. Dust cleaning is presented only as general cooling hygiene: dust worsens the heat that accelerates ageing.
• Power-supply quality. A poorly regulated or low-quality power supply can introduce voltage fluctuations that worsen power-delivery stress. A stable, adequate power supply reduces electrical stress on the card's power path. That is a qualitative engineering point, not a brand recommendation. This article names no "good" or "bad" power-supply brands.
• Do not keep running a card with known bulged capacitors and hope. A bulging capacitor has already started to fail internally even if the device still powers on. Left in place it keeps losing capacitance. The safe advice is to stop using the card for heavy load and get it inspected. Continued use is not presented here as safe.
• Load. Sustained heavy load means sustained heat and ripple. Avoiding unnecessary sustained full load on an ageing card is prudence, not a proven life-extension formula.

Common Misconceptions

FAQ

Can a bulging capacitor cause crashes or artefacts?

It can be associated with instability, crashes under load, artefacts, and boot difficulty, because high ESR means noisier and less stable power delivery. All of those symptoms are generic and multi-causal. A bulged capacitor is a fault that needs attention, but symptoms alone do not prove it is the only fault or even the main fault on a given card.

If the card still works, can I keep using it with a bulged capacitor?

That is not advisable. A bulging capacitor has already started to fail internally even if the device still powers on, and left in place it keeps losing capacitance. The safe course is to stop using the card for heavy load and get it inspected. Continued heavy use risks further damage. This article makes no claim that a bulged capacitor is certain to rupture violently — venting is the designed behaviour, and violent rupture is the failure-of-the-vent case.

Can I test a capacitor with an ordinary multimeter?

A standard multimeter cannot measure ESR, which is usually the first parameter to degrade. An ordinary resistance or capacitance test can miss many bad capacitors. An ESR meter is the screening tool, but it has limits: it does not measure capacitance, a shorted capacitor reads falsely low, parallel capacitors mask each other, and readings vary with temperature and voltage. A normal in-circuit reading does not fully exclude a masked fault.

Is it enough to replace only the visibly bulged capacitor?

Sometimes one capacitor is the only failed part. Sometimes it is one of several stressed parts on the same rail, or a symptom of wider heat, ripple, or power-supply stress. Whether to replace one or several is a judgement call for the repairer after testing the rail, not an automatic rule. Replacing one capacitor and seeing the card work does not prove the diagnosis was complete.

Do solid or polymer capacitors ever bulge or fail?

Polymer and solid aluminium capacitors avoid the dry-out mechanism, but they can still fail through gradual parameter drift, leakage current, rising ESR, and mechanical stress. Solid tantalum capacitors fail differently, typically as a short or low insulation resistance, and manganese-dioxide-cathode types can ignite at breakdown. A flat top is a type cue, not a health certificate.

What should a replacement capacitor match?

ESR first, ripple-current rating second, capacitance last. As rules of thumb: ESR same or lower, ripple rating same or higher, voltage same or higher without going absurd, capacitance within ±20% of the original per one distributor source, temperature rating matched or exceeded, physical fit checked, and polarity verified. VRM and switching-regulator positions need ultra-low-ESR types; a general-purpose capacitor may be inadequate there.

Sources

All sources accessed 2026-10-19.

• CapsMod — Cap Basics – Electrolytic & Polymer Capacitors — https://www.capsmod.net/caps-basics.html
• Wikipedia — ESR meter — https://en.wikipedia.org/wiki/ESR_meter
• Specap — Advanced Capacitor Testing: ESR Meters, LCR Analyzers & Impedance Analysis — https://specap.com/resources/blog/advanced-capacitor-testing-beyond-multimeter
• ScienceInsights — How to Measure ESR of a Capacitor: 3 Methods — https://scienceinsights.org/how-to-measure-esr-of-a-capacitor-3-methods/
• AKKN — Reading Bulge, Leakage, and Open-Circuit Evidence to Pick a Safer Electrolytic Replacement — https://www.akkn.com/news/8435.html
• Witonics — Why Do Capacitors Bulge? Causes and How to Prevent It — https://witonics.com/blogs/buying-guides/why-do-capacitors-bulge-causes-and-how-to-prevent-it
• MarcTech2 — Why Electrolytic Capacitors Fail (And How to Prevent It) — https://www.marctech2.com/news/why-electrolytic-capacitors-fail-and-how-to-prevent-it
• ePanorama — Electrolytic capacitor failures — https://www.epanorama.net/blog/2010/03/23/electrolytic-capacitor-failures/
• Wikipedia — Capacitor plague — https://en.wikipedia.org/wiki/Capacitor_plague
• Hackaday — Capacitor Plague tag page — https://hackaday.com/tag/capacitor-plague/
• Spiceworks Community — Does these capacitors seem blown? — https://community.spiceworks.com/t/does-these-capacitors-seem-blown/300421
• PC Hardware Pro — Identify swollen or leaking capacitors on the motherboard — https://www.pchardwarepro.com/en/como-identificar-condensadores-con-fugas-o-hinchados-en-la-placa-base/
• Tom's Hardware Forum — Bad Caps — https://forums.tomshardware.com/threads/bad-caps.476483/
• All About Circuits Forum — White rubbery material on circuit board — https://forum.allaboutcircuits.com/threads/white-rubbery-material-on-circuit-board.71245/
• EEVblog Forum — Why so many types of glues used in securing electronic components? — https://www.eevblog.com/forum/projects/so-many-types-of-glues-used-securing-electronic-components-is-it-necessary/
• Audio Science Review — Bond glue issues and pictures — https://www.audiosciencereview.com/forum/index.php?threads%2Fbond-glue-issues-and-pictures.10329
• Tube Maze — Kenwood L-07M restoration — https://tubemaze.biz/restoring-the-kenwood-l-07m-a-pound-of-glue-and-a-transistor-that-found-itself/
• OPCBA — PCBA Residue Reliability: Flux, Ionic & Cleaning — https://opcba.com/does-residue-on-pcba-have-any-effect-on-reliability/
• AKKN — Applying Ripple-Current and ESR Limits When Sourcing Capacitors for Switching Regulators — https://www.akkn.com/news/8428.html
• All About Circuits Forum — Capacitor replacement — https://forum.allaboutcircuits.com/threads/capacitor-replacement.135489/
• AnandTech Forums — Question on replacement caps — https://forums.anandtech.com/threads/question-on-replacement-caps.1686914/
• Practical Machinist — Electronics: How important are matching voltages on electrolytic capacitors — https://www.practicalmachinist.com/forum/threads/electronics-how-important-are-matching-voltages-on-electrolytic-capacitors.271505/page-3
• Lighthouse LEDs — Power Supply Filter Capacitor: Smoothing Ripple Voltage — https://lighthouseleds.com/blog/power-supply-filter-capacitor-ripple.html
• Accio — Capacitor Polarity Identification Guide — https://www.accio.com/plp/capacitor-polarity-identification
• Testbook — Electrolytic reverse-polarity Q&A — https://testbook.com/question-answer/what-happens-when-an-electrolytic-type-capacitor-i--630c81334346cedbe0f37ecc
• IC Online — Are capacitors polarized? — https://ic-online.com/news/post/are-capacitors-polarized-a-simple-guide-to-avoid-circuit-damage
• Fly-Wing — Electrolytic Capacitor — https://www.flywing-tech.com/blog/electrolytic-capacitor/
• PCB Store — What is a VRM? — https://pcbstore.com.bd/glossary/vrms
• Ruigu — Understanding Voltage Regulator Modules — https://ruigus.com/2025/11/28/understanding-voltage-regulator-modules-a-complete-guide/
• Corsair — What are VRMs on a Motherboard? — https://www.corsair.com/lm/tw/explorer/glossary/what-are-vrms-on-a-motherboard/
• GeeksforGeeks — Voltage Regulator Module — https://www.geeksforgeeks.org/electronics-engineering/voltage-regulator-module/
• YouTube / XDedu — 顯示卡為什麼會吱吱叫？電路板上會『唱歌』的電感和電容 — https://www.youtube.com/watch?v=Vg2kmGxv26E
• Tom's Hardware Forum — High pitch whistling… Is it coil whine? — https://forums.tomshardware.com/threads/high-pitch-whistling-coming-through-sound-systems-is-it-coil-whine.2378457/
• AEANET — What Is GPU Coil Whine? — https://www.aeanet.org/what-is-gpu-coil-whine/
• IOPscience — Reliability and Failure Mode in Solid Tantalum Capacitors — https://iopscience.iop.org/article/10.1149/2162-8777/abf728
• Weishi Electronics — Polymer Capacitor Failure — https://weishielectronics.com/polymer-capacitor-failure/
• Mouser / Panasonic comparison PDF — Which is better, Aluminum Electrolytic or Conductive Polymer Aluminum Solid Capacitor? — https://www.mouser.pe/pdfDocs/whichisbetter,aluminumelectrolyticcapacitororconductivepolymeraluminumsolidcapacitor.pdf
• TME / Panasonic — Panasonic Electrolytic Capacitors — https://www.tme.eu/en/news/about-product/page/62186/panasonic-electrolytic-capacitors/
• TrueGeometry — Reduced system performance due to capacitor issues — https://www.truegeometry.com/api/exploreHTML?query=Reduced%20system%20performance%20due%20to%20capacitor%20issues
• TrueGeometry — Capacitor Failure Modes — https://blog.truegeometry.com/api/exploreHTML/8ae61fe43705424c6da4d520062e765c.exploreHTML