Thermal design for a rectifier is one arithmetic chain: heat generated at the junction flows through the junction-to-case resistance, the mounting interface, and the heatsink-to-ambient path, and the sum of those temperature drops sets the junction temperature. If TJ stays below the datasheet maximum with margin, the part lives; if it does not, no headline current rating saves it. This guide walks the full chain with a worked example and a sign-off checklist.
The Thermal Circuit: Junction to Ambient
Think of heat like current and temperature like voltage. The junction generates a heat flow equal to the total loss PD; it then passes through a series of thermal resistances—junction to case (RθJC), case to heatsink or board (RθCS), and heatsink or board to ambient (RθSA)—before reaching the surrounding air. Every resistance in the chain drops some temperature:
TJ = TC + (PD × RθJC), and TC = TA + (PD × (RθCS + RθSA))
The chain view explains the most common thermal failure mode: a design that quotes a low RθJC but ignores the interface and heatsink stages is missing most of the circuit. In a board-mounted part the “heatsink” is the PCB copper itself, and its area, thickness, and via count are part of RθSA.
For a board-mounted part the chain branches: heat spreads through the exposed pad into the top-layer copper, down through vias into inner planes, and out to the enclosure or airflow. Each branch adds a parallel path rather than one series path, which is why copper area and via count are first-order variables in a board-level thermal design.
Reading RthJC and RthJA Correctly
Two thermal symbols appear on every rectifier datasheet, and they answer different questions. RθJC (junction-to-case) is a property of the device and its package: it describes the temperature drop from the silicon die to the case or exposed pad, independent of how the part is mounted. RθJA (junction-to-ambient) includes the mounting: it assumes a defined board, copper area, and airflow, and it is only valid under those conditions.
The gap between the two numbers shows how much the mounting matters. The AMBRP5100, for example, lists 3 °C/W junction-to-case and 50 °C/W junction-to-ambient—the difference is the board-level heat path. A 5 A part on a tiny pad behaves much closer to RθJA than to RθJC, which is why “the current looks fine on paper” designs run hot in the field. Always ask what mounting the RθJA figure assumes before comparing parts.
The measurement context matters too: RθJC is defined by thermal-measurement standards such as JEDEC JESD51 with the case held at a defined temperature, so the number is repeatable and comparable across suppliers. RθJA is less standardized—every vendor states a different board—which is exactly why it cannot be used for cross-part comparison without reading the test conditions.
The Junction Temperature Calculation, Worked Example
Do the arithmetic at the real working condition, with the maximum loss and the complete chain. Worked check on a 5 A, 100 V board-mounted part: at 5 A with the maximum VF of 0.76 V, conduction loss is about 3.8 W. With 3 °C/W junction-to-case resistance, the junction-to-case rise is roughly 11 °C. If the case sits at 95 °C on the assembled board, the junction is near 106 °C—comfortably inside a 150 °C ceiling for the conduction term alone.
Then add what the first pass left out. At 106 °C junction, the leakage term is no longer the microamp-level 25 °C number; the datasheet’s 5 mA maximum at 100 °C reverse-leakage figure means a 100 V reverse bias can add roughly 0.5 W at the hot junction, which raises TJ further. The honest calculation iterates: recompute VF and IR at the new junction temperature, or confirm the result with a case-temperature measurement on the prototype. A single pass with 25 °C typical values is an estimate, not a thermal design.
The interface numbers belong in the same pass. A dry metal-to-metal contact can add roughly 1–2 °C/W or more at typical tab pressure, while a correctly applied TIM brings the interface below about 1 °C/W—both as engineering guidance, not datasheet values. At 3.8 W, the difference between the two is several degrees of junction temperature before the heatsink is even considered.
Mounting Quality Changes the Real Rth
The interface between the case and its heat path is where most thermal margin is lost. Bare metal-to-metal contact traps air in microscopic gaps; a thermal interface material (TIM) fills them. Screw torque controls the pressure that seats the TIM—too little leaves a gap, too much risks cracking the package. On board-mounted parts, solder voids under the exposed pad and undersized copper area add resistance that no external heatsink recovers.
The rule for the design phase: never let the datasheet RθJC stand in for the assembled RθCS + RθSA. Measure the case temperature on the first prototype, compute the junction temperature with the real chain, and confirm the margin before freezing the mechanical design. Installation-specific guidance for tab packages is covered in the TO-220 mounting guide; board-level rules for the exposed-pad packages are covered in the D2PAK and PDFN design guides.
For exposed-pad packages, vias are the bridge from the top-layer pad to the inner planes: a small pad with no vias concentrates all the heat in the surface layer, while a via array spreads it through the board thickness. The via count and copper weight are design parameters, not afterthoughts, and they show up directly in the measured case temperature.
Derating: What the Curve Is Telling You
The derating curve on a rectifier datasheet plots maximum forward current against case temperature. It is the manufacturer’s statement of how much current the part can carry at a given TC while keeping the junction within its limit—and it is the number that corrects the “40 A headline” impression. A part rated 40 A at a low case temperature can carry far less at 100 °C case, and the curve shows exactly how much.
Read the curve at the case temperature your design will actually reach, then apply your own margin for lot variation and transient peaks. Continuous duty at or below 80% of the curve value at the real case temperature is a common engineering starting point, with the final number confirmed by the junction calculation and a prototype measurement.
A concrete reading: a 40 A part whose derating curve drops to 25 A at a 100 °C case is telling you that 25 A is the ceiling at that case temperature—regardless of the headline rating. Design the load and ambient to stay inside the curve with your own margin, and treat any violation as a thermal problem even if the average current looks low.
When Heat and Leakage Interact
The conduction-loss calculation assumes VF only, but a Schottky’s reverse leakage also generates heat, and leakage grows with the temperature it helps create. At the hot working condition the leakage power joins the loss budget, and in the worst case the two terms feed each other in a positive feedback loop. The full mechanism and the stability criterion are covered in the thermal-runaway guide; for the thermal design here, the rule is simple: include the leakage term at the maximum junction temperature and reverse voltage before signing off the margin.
A 10-Item Thermal Design Checklist
| # | Check | Pass |
|---|---|---|
| 1 | Loss computed with maximum VF at the operating temperature | ☐ |
| 2 | Leakage power added at the hot junction and reverse voltage | ☐ |
| 3 | RθJC read from the datasheet for the actual package | ☐ |
| 4 | Interface resistance (TIM, torque, solder) included | ☐ |
| 5 | Board or heatsink-to-ambient resistance sized for the enclosure | ☐ |
| 6 | Derating curve read at the real case temperature | ☐ |
| 7 | Continuous duty derated, not rated | ☐ |
| 8 | Worst-case ambient and sealed-enclosure conditions tested | ☐ |
| 9 | Junction temperature calculated with margin to TJ(max) | ☐ |
| 10 | Case temperature measured on the first prototype | ☐ |
Engineering note. The chain TJ = TC + (PD × RθJC) and the RθJC/RθJA distinction follow the rating structure used in Good-Ark rectifier datasheets—RθJC 2.0–3.0 °C/W and RθJA 40–62.5 °C/W for the datasheet-published parts. The 3.8 W/11 °C/106 °C example uses the AMBRP5100’s maximum VF and RθJC, and it is an estimate pending the leakage iteration or a prototype measurement; no single-pass typical-value calculation should be treated as the final thermal result.
Frequently Asked Questions
What is the difference between RθJC and RθJA?
RθJC is the temperature drop from the die to the case, a property of the device and package. RθJA adds the mounting—board, copper, airflow—and is only valid for the stated conditions. The gap between them shows how much the heat path outside the part matters.
How do I calculate junction temperature?
Use TJ = TC + (PD × RθJC), where PD is the total loss at the working condition and TC is the case temperature. For a complete picture, include the interface and heatsink-to-ambient resistances and iterate or measure, because VF and leakage change with temperature.
Why does my rectifier run hot even though the current is below its rating?
The IF(AV) rating is valid at a stated case temperature and assumes a working heat path. High ambient, continuous duty, a small copper area, or a dry interface can push the case temperature beyond the rating point even when the average current looks fine.
How much should I derate for continuous operation?
Treat the derating curve value at the real case temperature as the ceiling, not the target. Keeping continuous current at or below 80% of that value is a common starting point, with the final margin confirmed by the junction calculation and a prototype measurement.
Does leakage need to be in the thermal budget?
Yes, especially for Schottky rectifiers at high junction temperature. Leakage power (VR × IR) adds heat at the hot working condition and can feed the temperature rise that increases it; add it to the loss budget and confirm the stability criterion.
Conclusion
Rectifier thermal design is one chain and one number: walk the heat from junction to ambient through every resistance, compute TJ with the maximum loss at the operating temperature, and confirm the margin on a prototype. Read RθJC and RθJA for what they are, use the derating curve at the real case temperature, and include the leakage term before signing off.
Compare thermal ratings across the Schottky rectifier diodes category on the Good-Ark site, and contact Good-Ark with your load, ambient, and mounting conditions for a thermal review and sample support.