Module 3 Electrical Fundamentals study materials
Module 3 Electrical Fundamentals study resources

Ageing Reservoir Capacitor: Out of Tolerance, Hot, and Short of a Spare

Module 3 Electrical Fundamentals · Put your knowledge to work in context.

All case studies

Ageing Reservoir Capacitor: Out of Tolerance, Hot, and Short of a Spare

A 28 V DC avionics supply uses a 4700 µF, 40 V, 105 °C electrolytic capacitor as the reservoir across the rectified output of a 400 Hz transformer-rectifier unit. After nine years in service the full-load ripple has risen from 40 mV to 180 mV peak-to-peak and the capacitor can is slightly domed. Removed and measured, the capacitor reads 3150 µF and 0.35 Ω equivalent series resistance; a serviceable sample of the same type reads 0.02 Ω. Leakage current at 28 V measures 3.5 mA against a limit of 1 mA. A 2.2 kΩ bleed resistor is permanently connected across the capacitor, which is mounted on a card handled during troubleshooting. Stores hold only a 3300 µF, 50 V, 85 °C capacitor in the same can size, and the aircraft has a midnight departure. The unit's ripple limit at full load is 100 mV peak-to-peak.

1. Assess the measured capacitance against the component tolerance, and show how the capacitance and ESR readings together account for the ripple growth.

Read the worked answer

Rated 4700 µF ±20% gives a lower acceptance limit of 3760 µF. The measured 3150 µF is 33% below nominal and therefore out of tolerance on its own, regardless of any other measurement.

Ripple across a reservoir is set by the capacitor impedance at the ripple frequency: Z ≈ ESR + 1/(2πfC).

  • As new: 1/(2π × 400 × 4700 × 10⁻⁶) = 0.085 Ω, plus 0.02 Ω ESR ≈ 0.105 Ω.
  • Degraded: 1/(2π × 400 × 3150 × 10⁻⁶) = 0.126 Ω, plus 0.35 Ω ESR ≈ 0.476 Ω. That is about 4.5 times the impedance, so for the same ripple current the ripple voltage rises roughly 4.5-fold - matching the observed 40 mV to 180 mV.

The ESR term dominates: the loss of capacitance alone would only raise ripple about 1.5-fold. High ESR means internal heating of I²R, which is why the can is domed - internal pressure from evaporated electrolyte. The 3.5 mA leakage, over three times the limit, confirms dielectric degradation. The part is unserviceable on three independent counts and is a venting risk; do not return it to the unit.

2. Calculate the bleed-resistor time constant, the discharge time to 5 V, and the stored energy, and state the handling precautions for the card.

Read the worked answer

τ = RC = 2200 Ω × 4700 × 10⁻⁶ F = 10.3 s. Time to fall from 28 V to 5 V: t = τ ln(28/5) = 10.3 × 1.72 ≈ 17.8 s. To 12 V: 10.3 × ln(2.33) ≈ 8.7 s. Energy: W = ½CV² = 0.5 × 4700 × 10⁻⁶ × 28² ≈ 1.84 J.

Consequences: the bleed resistor is a slow discharge path, not a safety device that makes the card instantly safe - a card pulled and handled within a few seconds of power-down still holds most of its charge. Wait at least 30 s (about 3τ, roughly 5% of initial voltage if the resistor is healthy), then measure the terminals with a meter before touching. Verify the bleed resistor itself: if it has drifted open, the charge persists indefinitely, so a resistance check across the capacitor (after discharge) is part of the inspection.

The 1.84 J is below the usual high-voltage threshold but still a genuine hazard: with a low-ESR capacitor the available discharge current is limited mainly by circuit resistance and can pit probe tips, weld a screwdriver blade, burn skin or ignite loose FOD. Discharge through a resistor, not a direct short across the terminals, and avoid shorting a charged capacitor with a metallic tool near the card.

3. Should the 3300 µF, 85 °C capacitor be fitted so the aircraft makes its departure? Justify the decision.

Read the worked answer

No - three independent grounds, any one of which is sufficient to reject it.

  1. Capacitance: 3300 µF is 30% below nominal and below even the −20% acceptance limit. Ripple impedance rises accordingly and the measured 100 mV ripple limit would be breached, giving the same symptom the unit was removed for.
  2. Temperature rating: 85 °C in a location designed around a 105 °C part. Electrolytic life roughly halves for every 10 °C of operation above rating, so service life and reliability fall sharply; the failure mode is exactly the dry-out and venting already seen.
  3. Ripple-current and ESR: a same-size can is no guarantee of equal ripple-current rating or ESR. An under-rated part overheats, domes and vents, and can short, taking the rectifier with it. Fitting a part not in the illustrated parts catalogue is also a configuration-control violation.

Action: order the correct 4700 µF, 40 V, 105 °C part and check its ripple-current rating against the circuit; if the aircraft must move, use the operator's approved deferral or a serviceable exchange unit, not a downgraded capacitor. When the unit is repaired, also check the bleed resistor value, the rectifier diodes and downstream loads for the elevated ripple that has been present.