A winding, cable termination, or molded assembly can pass a basic dielectric withstand test and still contain insulation defects that will shorten its service life. Understanding partial discharge causes is therefore not an academic exercise. It is a practical requirement for finding localized weaknesses before they become tracking, flashover, insulation breakdown, or an in-service failure.
Partial discharge (PD) is a localized electrical discharge that bridges only part of the insulation system. It does not create a complete conductive path between energized conductors. Yet each discharge can erode insulating material, generate chemically active byproducts, and increase the likelihood that a small defect becomes a growing failure mechanism.
What Partial Discharge Reveals About Insulation
PD occurs when the electric field in a localized region exceeds the breakdown strength of the material in that region. The surrounding insulation may remain intact and the overall assembly may continue to operate. That is why a conventional hipot test, while essential for validating dielectric strength, does not always provide enough information about long-term insulation integrity.
The measured behavior matters as much as the initial presence of discharge. Partial discharge inception voltage (PDIV) indicates the voltage at which detectable activity begins as voltage rises. Partial discharge extinction voltage (PDEV) identifies where that activity stops as voltage falls. Discharge magnitude, pulse repetition rate, phase-resolved patterns, and changes over time can help distinguish a benign test artifact from a meaningful insulation defect.
A PD result must be interpreted in the context of the product, insulation technology, applied waveform, and applicable standard. A discharge level that is unacceptable in a high-reliability aerospace component may not have the same implication in another product category. The engineering objective is not simply to obtain the lowest possible numerical reading. It is to determine whether the measured activity is internal to the device under test, repeatable, relevant to the intended operating conditions, and within the product’s acceptance criteria.
Primary Partial Discharge Causes
The most common causes are localized field enhancement, dielectric discontinuities, contamination, and mechanical damage. In many real assemblies, more than one mechanism is present. A void may originate during manufacturing, then become more active because moisture or thermal cycling changes the local electric-field distribution.
Voids and Gas-Filled Cavities
Internal voids are among the most recognized partial discharge causes. They can exist within cast resin, potting compounds, laminated insulation, cable insulation, adhesive bonds, or winding impregnation. Because the dielectric constant of a gas-filled cavity differs from the surrounding solid dielectric, voltage does not distribute uniformly. The electric field inside the void can become high enough to ionize the gas even when the average field across the insulation is within design limits.
Voids may result from inadequate impregnation, trapped air, poor mixing of encapsulants, incomplete curing, shrinkage, or inconsistent material flow during molding. Their effect depends on size, geometry, location, gas pressure, and orientation relative to the electric field. A small void near a sharp conductor edge can be more consequential than a larger void located in a lower-field region.
Sharp Edges, Protrusions, and Poor Field Control
Electrical stress concentrates at sharp conductor edges, burrs, strand ends, damaged semiconductive layers, and abrupt changes in conductor geometry. This is particularly relevant in high-voltage connectors, cable accessories, transformers, motor windings, stators, and compact power-electronic assemblies.
A design can have adequate bulk insulation thickness while still experiencing PD at a single poorly controlled interface. Corona rings, stress-relief materials, smooth conductor preparation, and carefully designed termination geometry reduce local field enhancement. Manufacturing control is equally significant. A nicked foil edge or a stray wire strand can defeat an otherwise sound electrical design.
Delamination and Interfacial Gaps
Interfaces between dissimilar materials are frequent defect locations. Bond lines between resin and conductor, insulation tapes and windings, or coatings and substrates can separate due to inadequate adhesion, cure variation, differential thermal expansion, or mechanical loading. The resulting gap behaves much like an internal void, especially where the interface is exposed to high electrical stress.
Interfacial defects deserve attention because they can grow in service. Thermal cycling may repeatedly open and close a gap. Vibration can worsen poor bonding. Once PD begins at an interface, discharge byproducts and localized heating may further degrade adhesion, creating a self-reinforcing failure path.
Contamination, Moisture, and Surface Degradation
Contamination changes surface conductivity and can create discharge activity along insulation surfaces. Conductive dust, ionic residue from manufacturing, process fluids, flux residues, and salt deposits can distort the electric field or support tracking. Moisture can lower surface resistance, alter dielectric properties, and accelerate chemical degradation.
Surface discharge is especially relevant where insulation interfaces with air. Humidity, altitude, temperature, and surface cleanliness all influence the onset and character of activity. A component tested in a clean, climate-controlled laboratory may behave differently after exposure to industrial contaminants or repeated field temperature cycles.
Mechanical Damage and Aging
Insulation does not need to be visibly cracked to develop PD susceptibility. Repeated vibration, assembly stress, handling damage, abrasion, and excessive bend radius can create microcracks or loosen interfaces. In rotating machinery, electromagnetic forces and thermal expansion can progressively compromise turn insulation and winding support.
Aging mechanisms also change the material itself. Heat can embrittle polymers, ultraviolet exposure can degrade exposed surfaces, and chemical attack can reduce dielectric strength. PD may be the initiating defect, but it can also be a later symptom of insulation already weakened by environmental or mechanical stress.
Why Applied Voltage and Environment Change the Result
PD testing is highly dependent on test conditions. An observed discharge level cannot be separated from the voltage waveform, frequency, ramp rate, coupling arrangement, and environmental state of the device under test.
Alternating-current, direct-current, and impulse applications do not stress insulation identically. AC testing commonly supports phase-resolved analysis because repetitive discharge pulses can be related to the voltage phase angle. Under DC stress, charge accumulation and field redistribution may delay or alter discharge behavior. Fast-switching power electronics add another consideration: steep voltage edges can create high local electric stress that a low-frequency test may not fully represent.
Air density is another variable. At reduced atmospheric pressure, external corona can begin at lower voltages. This matters for aerospace equipment, high-altitude applications, and test laboratories where environmental conditions are not controlled. Temperature and humidity can similarly affect surface discharge, material conductivity, and moisture absorption.
These dependencies do not make PD testing less useful. They make test definition more critical. The applied voltage, conditioning period, environmental conditions, and pass-fail criteria should reflect the product’s insulation system and intended service profile.
Separating a Real Defect From Test-System Noise
A partial discharge measurement system is sensitive by design. Without appropriate shielding, grounding, calibration, and fixture design, external interference can resemble a device defect. Switching supplies, digital electronics, fluorescent lighting, nearby high-voltage equipment, poor ground connections, and loose fixture hardware can all introduce unwanted signals.
A credible test setup begins with a controlled measurement circuit. The coupling device, measuring impedance, calibrator, test object, high-voltage source, and acquisition system must be appropriate for the voltage and capacitance of the application. Calibration establishes the relationship between the measured signal and apparent charge. It does not prove that every pulse originates within the test object.
Correlation is the key diagnostic discipline. Engineers should compare measurements with and without the device under test, observe whether pulse patterns track the applied voltage phase, inspect the fixture for external corona, and repeat the test after controlled changes. When the signal remains stable across these checks and changes predictably with applied voltage, the case for genuine PD is much stronger.
Building a Useful Partial Discharge Test Strategy
The best response to partial discharge causes is not a single end-of-line test. It is a coordinated approach that connects insulation design, process control, and diagnostic measurement.
At the design stage, evaluate electric-field concentration at edges, interfaces, and terminations. Material selection should account for dielectric properties, thermal expansion, moisture behavior, expected voltage waveform, and manufacturing capability. A material with excellent dielectric strength on a datasheet may still be a poor choice if the production process consistently leaves voids or weak interfaces.
During manufacturing, control the factors that create defects: surface preparation, cleanliness, resin mixing, vacuum impregnation, cure schedules, winding tension, conductor trimming, and handling. PD screening is most valuable when it feeds corrective action upstream. Repeated activity in the same phase pattern or voltage range may point to a specific tooling, material, or assembly issue.
For qualification and production testing, define a repeatable procedure with documented ramp rates, dwell times, voltage levels, noise limits, calibration methods, and acceptance criteria. Relevant requirements may be drawn from IEC 60270 or product-specific standards, but the test plan must be tailored to the actual insulation system. Standard alignment provides a measurement framework; application knowledge determines whether the framework answers the failure-risk question.
High-quality instrumentation matters because the measurement chain must detect small, repeatable signals in the presence of high voltage and electrical noise. Systems with controlled high-voltage output, stable acquisition, appropriate coupling components, and traceable calibration support help test teams reduce uncertainty and make defensible decisions.
The practical goal is to make partial discharge a design and process indicator rather than a late-stage surprise. When a measurement is tied to known field stresses, controlled environmental conditions, and disciplined fixture validation, it can reveal where insulation reliability is being lost – while there is still time to correct it.
