Rectifier overheating has three root causes—underrated current or duty, a failing thermal interface, or leakage growing with temperature—and a fixed measurement protocol followed by a fault tree finds the right one faster than a part swap. Measure the case temperature at steady state, work through the three branches in order, and the fix is usually mechanical or a derating change. Whether the rectifier comes from Good-Ark or another supplier, this protocol, the tree, and the retest criteria apply the same way.
Symptoms That Point to Overheating
The symptoms—hot case, efficiency loss, thermal shutdown, discolored board or solder, early field failures—are all downstream of a temperature that was never verified at the harshest working condition. Record ambient temperature, load current, and case temperature together before changing anything; the diagnosis is a comparison between measured loss and rated capability, and without the three numbers together it cannot start.
Measure Correctly—and Safely
Measurement quality decides the diagnosis, and safety comes first:
- De-energize before installing sensors. The tab or exposed pad of a rectifier can sit at a live circuit potential. Mount the thermocouple with the circuit off, then verify insulation before re-applying power.
- Attach the sensor to the case, not the plastic. Use a fine-gauge thermocouple bonded or clamped to the metal tab near the die. Soldering can damage the part and change the heat path; thermally conductive adhesive or a clamp is preferred. A sensor taped to the plastic body can read 20–40 °C cooler.
- Use a thermal camera for distribution. Set the emissivity for the surface—bare metal reflects, painted or anodized surfaces emit—or the image lies.
- Soak to steady state, not to a fixed time. Heat up until the case temperature stops changing at a defined rate—for example, less than 1 °C change over 10 minutes. Fifteen to thirty minutes is only a preliminary reference; large heatsinks and sealed enclosures can take much longer.
The Rectifier Overheating Fault Tree
| Observed symptom | Branch | First evidence to collect |
|---|---|---|
| Case temperature above the design target at full load | Branch 1: current / duty | RMS and average load vs IF(AV) at the actual case temperature |
| Case hot while the electrical loss looks within budget | Branch 2: mounting / interface | TIM, torque, copper area, solder voids |
| Case keeps climbing after load stabilizes, or hot with load removed | Branch 3: leakage at temperature | IR at operating reverse voltage and junction temperature |
Work the branches in order; most field complaints resolve in Branch 2 before the part is ever at fault.
Branch 1: Current and Duty Underestimation
IF(AV) is an average-current rating valid at a stated case temperature with a defined heat-flow route. Three mistakes reappear in field returns: continuous duty treated as average, peak current confused with average, and maximum ambient ignored. The junction responds to time-averaged power, but a high peak with poor duty still drives the case temperature up, and every degree of underestimated ambient is a degree of case temperature the datasheet did not give you.
A worked check on a generic 10 A-class rectifier: at 10 A with a maximum VF of 0.76 V at the operating temperature, forward loss is about 7.6 W; with 3 °C/W junction-to-case resistance that is roughly 23 °C of rise from case to junction. If the measured case sits at 95 °C, the junction is near 118 °C—inside a 150 °C ceiling, but with leakage and interface resistance still to add. If the case actually runs at 110 °C, the junction is near 133 °C and the margin is nearly gone. The case-temperature target, not the headline amps, decides the design.
Branch 2: Mounting and Thermal Interface
The second branch is mechanical, and it is the one that survives a part swap. Dry or missing thermal interface material, incorrect screw torque, an undersized insulation pad, small copper area on board-mount parts, and solder voids under an exposed pad all add resistance to the case-to-ambient path. Airflow and conformal coating belong on the same checklist: a heatsink in a dead air pocket, or a coat that insulates the tab, produces the same symptom with the same wrong conclusion. Re-torque, replace the TIM, and enlarge the copper before changing the semiconductor—it is faster, cheaper, and fixes a large fraction of “the part is too small” verdicts.
Branch 3: Leakage at Temperature
The third branch is the one that appears months after launch: leakage grows with junction temperature and reverse voltage, and the leakage power (VR × IR) adds heat to the very temperature rise that increases it. The signature is measurable without lab equipment: the case temperature keeps climbing after the load has stabilized, or the part runs hot with the load removed and only reverse voltage applied. For the diagnosis, include the leakage term to the loss budget at the maximum junction temperature and reverse voltage before blaming the part. The full mechanism and the stability check are covered in the dedicated thermal-runaway guide; Schottky rectifiers are the most exposed family because their low forward drop and high leakage share the same barrier physics. If the branch points to a family change, the Schottky rectifier diodes category on the Good-Ark site is the starting point for comparing parts by voltage class and leakage.
Fixes, Retest Protocol, and When to Change the Part
| Root cause | Corrective action | Verification |
|---|---|---|
| Current / duty underestimated | Derate, reduce duty, or move to a higher IF(AV) class | Re-run the loss math at maximum load and ambient |
| Mounting / interface | Re-torque, replace TIM, enlarge copper, add airflow | Re-measure case temperature at steady state |
| Leakage at temperature | Improve the heat-flow route, lower reverse stress, or switch family | Confirm the stability check passes at the hot junction |
Retest with a fixed protocol so before-and-after numbers are comparable: same enclosure or chamber, same load profile, same sensor location, and the same steady-state criterion. Record case temperature, calculated junction temperature, and margin to TJ(max) at each stage. If the margin is still thin after the mechanical fixes, the honest options are derating, a higher current class in the same package, or a family with better high-temperature leakage—each a design decision, not a lottery ticket.
Engineering note. The protocol above follows the rating structure used in rectifier datasheets—Good-Ark datasheets state IF(AV) at a stated case temperature, VF and IR at temperature, and RθJC for the junction calculation—and the steady-state criterion recommended for thermal measurement. No single soak time replaces a defined stability criterion, and no thermocouple reading is valid if it changes the mounting it is measuring. Thermal-runaway mechanics are treated in the dedicated guide; this article covers the diagnostic order.
Frequently Asked Questions
How do I measure the case temperature of a rectifier?
De-energize the circuit, attach a fine-gauge thermocouple to the metal tab near the die with adhesive or a clamp, then test at maximum ambient and load. Soak until the case temperature is stable—for example, less than 1 °C change over 10 minutes—and use a thermal camera with the correct emissivity for heat distribution.
How long should I soak before trusting the measurement?
Until the case temperature reaches a defined steady-state criterion, not a fixed number of minutes. Fifteen to thirty minutes is only a preliminary reference; sealed enclosures and large heatsinks can need much longer to stabilize.
Why does my rectifier overheat even though the current is below its rating?
The IF(AV) rating is valid at a stated case temperature and assumes a working heat-flow route. Continuous duty, high ambient, poor mounting, or small copper area can push the case beyond the rating point even when the average current looks fine.
Is the tab safe to touch or measure while the circuit is live?
No. The tab or exposed pad can sit at a live circuit potential. De-energize before mounting sensors, verify insulation, and treat the tab as live during operation.
How do I know if leakage current is the cause?
If the case temperature keeps climbing after the load stabilizes, or the part runs hot with the load removed, compare the leakage power (VR × IR at the working condition) with the forward loss. When leakage becomes a meaningful fraction of the budget at the hot junction, it is the likely driver.
Conclusion
Rectifier overheating is a diagnosis, not a verdict: measure at steady state with a safe, repeatable protocol, work the three-branch fault tree in order, and fix the mechanical path before changing the part. Verify with the junction calculation and the same retest protocol, and treat the case-temperature target as the specification. If the numbers still do not close, submit the rectifier part number, reverse voltage, load waveform, measured case temperature, and enclosure conditions to Good-Ark for an application review through the contact page.