Capacitor Plague

The capacitor plague refers to an era, roughly 1999 to 2007, when enormous numbers of aluminum electrolytic capacitors failed prematurely because of a defective electrolyte formulation. The flawed electrolyte lacked the additive chemistry needed to keep its composition stable over time, so it corroded the capacitor internals and generated hydrogen gas as it degraded. Affected capacitors bulge at the scored vent on top, split open, or leak brown electrolyte down their sides, and the equipment they sit in, motherboards, monitors, power supplies by the million, became unstable or died years before its expected service life. The widely reported origin story involves an incompletely copied electrolyte formula moving between competing manufacturers, but whatever the exact industrial history, the engineering lesson stuck: the electrolyte is the life-limiting component of the entire assembly.

The role of ESR

Equivalent series resistance (ESR) is the unwanted internal resistance every real capacitor has in series with its ideal capacitance. In an electrolytic, ESR is dominated by the electrolyte, so it is the aging signal: as electrolyte dries out or vaporizes, capacitance sags below rating and ESR climbs sharply. High ESR makes the capacitor dissipate real power as heat under ripple current, and that self-heating accelerates further electrolyte loss, a runaway loop that ends in venting. The vast majority of electrolytic field failures show up as rising ESR long before a dead short, which is why an ESR meter, not just a capacitance meter, is the core diagnostic tool: a capacitor can measure "correct" capacitance while its ESR has already tripled and the rail it filters has become noisy garbage.

What to look for in repair

Visual inspection catches the obvious cases: a domed or split top, crusty residue at the base, a lifted sleeve, or a popped vent. But capacitors that look perfect can be marginal, so measure ESR in circuit where the topology permits, compare against a low-ESR reference table for the given capacitance and voltage, and replace suspect units with quality low-ESR, high-temperature (105 °C) parts from reputable manufacturers, matching or exceeding the original ripple-current rating. Replace in groups: if one unit of a batch has failed, its siblings share the same chemistry, the same hours, and the same heat.

Why miners still care

The plague era is over, but its failure mode is permanent physics. Electrolytics age out of every power chain eventually, and mining hardware ages them fast: a PSU delivering kilowatts of continuous load in a warm, dusty environment is close to an accelerated life test. Aging electrolytics on power-supply and control boards are a frequent root cause of brownouts, random restarts, boards that drop out under load but pass idle tests, and machines that only crash on hot days, symptoms easily misblamed on firmware or hashboards. Heat is the multiplier, and thermal fatigue compounds the problem by cracking the solder joints under the very components that are drying out. When a machine develops power-quality ghosts, recapping or a proper PSU-level diagnosis, the kind of component-level work offered through D-Central's repair service, is routinely cheaper than replacing the unit, and keeping working hardware alive is its own quiet form of decentralization.

Component evolution has changed the picture without ending it. Modern hashboards lean on ceramic and solid-polymer capacitors, which contain no liquid electrolyte and therefore cannot dry out or vent, one reason the boards themselves tolerate years of heat. But the power supplies feeding them still depend on large aluminum electrolytics for bulk energy storage, because nothing else delivers that much capacitance per dollar. The plague-era lesson therefore lives on at the PSU: the electrolytic capacitors are the wear item, their lifetime halves for roughly every 10 °C of additional heat, and the fan that cools them is the cheapest life-extension device in the whole machine.

Cross-check PSU specs in the ASIC PSU reference.