Partial Discharge Testing: The Key to Long-Term IGBT Module Reliability

Partial Discharge Testing for IGBT Modules: The Key to Assessing Long-Term High-Voltage Insulation Reliability

As power systems push towards higher voltages and greater power densities—driven by applications like electric vehicles, renewable energy inverters, and industrial motor drives—the long-term reliability of insulation systems within power modules has become a paramount concern. While standard tests like Hi-Pot (high potential) testing can confirm the basic dielectric strength of an IGBT module, they fail to detect subtle, progressive degradation mechanisms that can lead to catastrophic failure in the field. This is where Partial Discharge (PD) testing becomes an indispensable tool for engineers. It serves not just as a quality check, but as a predictive method to evaluate the health and longevity of a module’s high-voltage insulation system.

Unlike a complete breakdown, partial discharge is a localized dielectric breakdown of a small portion of a solid or fluid electrical insulation system under high voltage stress. These are essentially tiny, repetitive sparks or arcs that don’t immediately bridge the electrodes. However, their cumulative effect is devastating, slowly eroding insulation materials and paving the way for a complete and often explosive failure. Understanding and implementing PD testing is crucial for ensuring the robust, long-term performance of high-voltage IGBT modules.

Understanding the Physics of Partial Discharge

To grasp the importance of PD testing, one must first understand the phenomenon itself. A perfect insulation system is homogeneous, with uniform dielectric properties. In reality, manufacturing imperfections are unavoidable, leading to microscopic voids, contaminants, or interfaces within the insulation material, such as the crucial silicone gel that encapsulates the internal components of an IGBT module.

The Mechanism of Discharge

When a high voltage is applied across the module, the electric field is not uniform. It intensifies at sharp points, material interfaces, or within these imperfections. Consider a tiny gas-filled void trapped within the silicone gel. The gas (often air) has a much lower dielectric strength than the surrounding gel. As the voltage rises, the electric field stress across this void can exceed the gas’s breakdown voltage, causing ionization and a small, rapid electrical discharge. This discharge is “partial” because it is contained within the void and does not bridge the main terminals of the module.

This process repeats with every cycle of the AC voltage, creating a continuous stream of micro-discharges. Each discharge event releases energy in several forms:

• Electromagnetic Emissions: High-frequency current pulses.
• Acoustic Energy: A faint crackling or hissing sound.
• Chemical Reactions: Formation of ozone and other corrosive byproducts.
• Light and Heat: Localized temperature increase and light emission.

It’s the cumulative energy and chemical byproducts of these thousands of discharges that degrade the insulation, carbonizing the surfaces of the void and creating conductive pathways (treeing) that eventually lead to complete insulation failure.

Key PD Measurement Parameters

PD testing is designed to detect the high-frequency current pulses generated by these discharges. The key metrics engineers look for are:

• Partial Discharge Inception Voltage (PDIV): The minimum voltage at which partial discharge activity begins to occur as the voltage is increased. A high PDIV is desirable, indicating a robust insulation design.
• Partial Discharge Extinction Voltage (PDEV): The voltage at which partial discharge activity ceases as the voltage is decreased from a higher level. Ideally, PDEV should be close to PDIV, but it’s often slightly lower. A large difference can indicate more severe insulation issues.
• Discharge Magnitude (pC): The amount of charge transferred in a single PD pulse, measured in picocoulombs (pC). International standards often set maximum acceptable pC levels (e.g., <10 pC) for a given test voltage.

Why Standard Hi-Pot Testing Isn’t Enough

A common question is why PD testing is necessary when a module already passes a Hi-Pot test. The two tests serve fundamentally different purposes and reveal different information about the insulation’s health.

The Hi-Pot test is a brute-force check. It applies a high DC or AC voltage for a short duration (typically one minute) to see if the insulation can withstand it without a complete breakdown. It’s a simple pass/fail test that answers one question: “Can the insulation survive this high voltage right now?”

PD testing, conversely, is a more sensitive diagnostic tool. It answers a different, more critical question: “Even if the insulation can withstand the voltage now, are there underlying defects that are actively degrading it and will cause it to fail prematurely in the future?”

The following table illustrates the critical differences and why relying solely on Hi-Pot testing can create a false sense of security.

Aspect	Hi-Pot (Dielectric Withstand) Test	Partial Discharge (PD) Test
Purpose	Verifies the ultimate dielectric strength against complete breakdown.	Detects localized, incipient faults and ongoing degradation processes.
Nature of Test	Destructive or potentially damaging; stresses the entire insulation system.	Non-destructive diagnostic; identifies flaws before they become critical.
Information Provided	Simple Pass/Fail (leakage current below a threshold).	Quantitative data (PDIV, PDEV, discharge magnitude in pC) and qualitative patterns.
Failure Indication	Indicates gross manufacturing defects or severe damage.	Indicates microscopic voids, contamination, delamination, or poor design.
Long-Term Reliability Prediction	Poor. A module can pass Hi-Pot but fail months later due to PD activity.	Excellent. The absence of PD at operating voltage is a strong indicator of long-term reliability.

The Partial Discharge Testing Procedure: A Practical Guide

Performing a PD test requires a specialized setup and a controlled electromagnetic environment to avoid false readings from external noise. The test is typically performed according to standards like IEC 60270 (for general HV test techniques) and IEC 61287-1 (for power semiconductor modules).

Essential Test Equipment

1. AC High-Voltage Source: A low-noise, variable AC power supply capable of generating the required test voltage.
2. Coupling Capacitor (Ck): A PD-free capacitor with a much larger capacitance than the device under test (DUT). It provides a path for the high-frequency PD pulses.
3. Measuring Impedance (Zm): A circuit (usually RLC) placed in series with the DUT to detect the PD current pulses and convert them into voltage signals.
4. PD Detector/Analyzer: An instrument that processes the signals from the measuring impedance, filters out the 50/60 Hz power frequency, and displays the PD pulses in terms of magnitude (pC) against the phase angle of the test voltage.
5. Device Under Test (DUT): The IGBT module, with its terminals properly connected to the test setup.

Step-by-Step Test Execution

The core procedure involves carefully applying voltage and monitoring for the onset of PD activity:

1. Calibration: The system is first calibrated by injecting a known charge (e.g., 10 pC) to ensure the detector’s measurements are accurate.
2. Voltage Ramp-Up: The AC voltage is slowly and smoothly increased from zero. The PD detector continuously monitors for any discharge activity.
3. PDIV Identification: The exact voltage at which repetitive PD pulses first appear and exceed a predefined threshold (e.g., 5-10 pC) is recorded as the Partial Discharge Inception Voltage (PDIV).
4. Test Voltage Hold: The voltage is often raised further to a specified test level (e.g., 1.5 times the rated operating voltage) and held for a short period to check for stability.
5. Voltage Ramp-Down: The voltage is then slowly decreased, and the point at which the PD activity ceases is recorded as the Partial Discharge Extinction Voltage (PDEV).

A successful test requires the PDIV to be significantly higher than the module’s maximum continuous operating voltage, and the measured PD magnitude at a specified test voltage must be below the accepted limit (e.g., <10 pC). This ensures a sufficient safety margin and confirms the integrity of the insulation system.

Case Study: Pre-empting Field Failures in a Wind Turbine Application

Problem: A renewable energy company experienced a series of unexpected IGBT module failures in their wind turbine converters after approximately 18-24 months of operation. The failures were catastrophic, causing significant downtime and costly repairs. All failed modules had passed the standard manufacturer’s Hi-Pot test during production.

Investigation & Solution: A root cause analysis was initiated. Decapsulation of the failed modules revealed evidence of electrical treeing and carbonization tracks within the silicone gel near the high-voltage terminals—a classic sign of long-term partial discharge activity. To validate this hypothesis, the engineering team implemented a PD testing protocol for all incoming power modules. They tested a new batch of modules from the same supplier and found that while all passed the Hi-Pot test, nearly 15% exhibited a PDIV that was only marginally above their peak operating voltage. This low margin was a critical red flag.

Result: By rejecting the modules with low PDIV, the company prevented potentially defective components from being installed in their turbines. They worked with the IGBT module manufacturer, who traced the issue to an inconsistent vacuum-potting process for the silicone gel, which was trapping microscopic air bubbles. The manufacturer refined its process and adopted PD testing as a 100% outgoing quality screen. The implementation of routine PD testing as a qualification and incoming inspection tool drastically reduced in-field failure rates, validating it as a critical step for ensuring the long-term reliability of systems like EV inverters and wind turbines.

Conclusion: PD Testing as a Cornerstone of High-Voltage System Reliability

For engineers designing, qualifying, or deploying high-voltage power electronics, partial discharge testing should not be viewed as an optional or overly complex procedure. It is a fundamental requirement for ensuring long-term operational safety and reliability. It provides insights into the health of an IGBT module’s insulation system that no other test can offer, moving beyond a simple check of dielectric strength to a predictive assessment of manufacturing quality and future performance.

Key takeaways for every engineer:

• PD is a leading cause of long-term insulation failure in high-voltage IGBT modules.
• PD testing is non-destructive and can detect hidden defects (voids, contaminants) that a Hi-Pot test will miss.
• A high PDIV is a critical indicator of a well-designed and properly manufactured insulation system.
• Adopting PD testing in qualification and IQC processes can significantly reduce field failures, warranty costs, and protect brand reputation.

As the industry continues to adopt wide-bandgap semiconductors like SiC and GaN, which enable even higher voltages and switching frequencies, the electrical stress on insulation systems will only intensify. This makes a deep understanding and rigorous application of partial discharge testing more critical than ever before. It is an investment in knowledge that pays dividends in reliability.