Schottky thermal runaway is a positive feedback loop: leakage current grows with junction temperature, leakage power heats the junction, and the heating raises leakage further. Whether the loop stops or runs away depends on one comparison—total loss versus the heat the assembly can remove. This article explains the mechanism step by step and the stability check that prevents it.
The Self-Heating Loop, Step by Step
Trace the loop one turn at a time. A Schottky under reverse bias conducts a leakage current IR, which depends on the junction temperature and the applied reverse voltage. That leakage generates power: PD = VR × IR. The power raises the junction temperature through the thermal resistance of the assembly, and the higher junction temperature increases IR. The loop has completed one turn, and the question is what happens on the next.
If the heat-removal path is strong, each turn adds less temperature than the last, and the junction settles at a stable equilibrium. If the leakage term grows faster than the heat path can shed, each turn adds more, and the junction climbs until the part fails—thermal runaway. The mechanism is specific to Schottky diodes because their low forward drop and higher leakage share the same barrier physics: the low barrier that conducts well also leaks more.
Reading the Leakage Curve at Operating Temperature
The leakage curve is the input to the loop, and the 25 °C number is the most misleading one on the datasheet. The AMBRP10H100 shows the scale: 10 µA maximum leakage at 25 °C, rising to 2.5 mA maximum at 100 °C—a 250× increase over 75 °C. The ASGC051BS curve shows the same shape across a finer grid: 5 µA maximum at 25 °C, 100 µA at 85 °C, and 5 mA at 125 °C.
Two reading rules matter. First, leakage depends on reverse voltage, and the datasheet states IR at a defined VR; re-read or interpolate the IR–VR–TJ curve at the actual operating bias rather than assuming linear scaling. Second, the curve is read at the junction temperature the loop will create, not at the temperature the design hoped for. A part that leaks microamps at 25 °C can leak milliamps at the hot working condition, and at a few tens of volts reverse that is a meaningful fraction of a watt feeding the loop—enough to matter, and more at higher voltage classes.
The Stability Check: Loss Curve vs Heat-Removal Line
The stability check is a two-line comparison. The loss curve plots total power—forward plus leakage—against junction temperature; it rises with temperature because leakage grows. The heat-removal line plots the power the assembly can remove, which rises linearly with the junction-to-ambient temperature difference divided by the total thermal resistance.
Where the two lines meet is the operating equilibrium. If the loss curve stays below the heat-removal line at the working condition, the junction settles there and the device is stable. If the loss curve crosses above the line—or the two lines do not meet at all—the junction climbs to failure. The check is performed with the maximum IR at the maximum reverse voltage and ambient, the assembled thermal resistance, and the real enclosure conditions.
Graphically, the check is also a tangent test: if the slope of the loss curve at the working condition is steeper than the slope of the heat-removal line, the equilibrium is unstable even if the two lines currently meet—a small temperature disturbance grows instead of decaying. The stable design keeps the loss curve flat enough and the removal line steep enough that the working condition sits well inside the stable region.
Runaway Margins: How Much Headroom Is Enough
The margin between the loss curve and the heat-removal line at the working condition is the design’s runway. A common engineering target is a design junction-temperature limit derived from the mission profile and the reliability target—part tolerance, aging, transient duty, and the allowed failure rate each move the number—with the leakage term included, plus a check that the loss curve does not approach the heat-removal line at the worst ambient. The design limit is set by the program’s reliability math, not by a fixed percentage of the absolute maximum.
Two margins deserve explicit attention. First, the temperature margin: a junction that runs near TJ(max) leaves almost no room for the leakage feedback to settle. Second, the voltage margin: reverse voltage changes leakage along the datasheet curve, so a design that raised the bus voltage late in the program can move the loss curve upward without any change to the part. Re-run the stability check whenever the operating conditions change, not just at the original selection.
Reverse voltage deserves the same re-check as temperature: re-read or interpolate the datasheet leakage curve at the new reverse voltage and junction temperature instead of scaling the old value proportionally. Record the reverse voltage in the stability worksheet alongside the ambient.
The Sealed-Enclosure Hot-Start Scenario
The classic runaway scenario is a sealed enclosure starting hot: a module at 80 °C ambient, a Schottky under continuous reverse bias, and a load low enough that leakage—not forward current—dominates the loss budget. The part starts warm, the leakage is already elevated, and in the worst case the loop climbs over minutes or hours, long after the bench test at 25 °C has passed.
The verification for this scenario is specific: soak the assembled unit at maximum ambient with the reverse bias applied and the load at its real profile, and watch the case temperature after it appears to stabilize. If the case keeps climbing after the load has settled, the loop is not at equilibrium. The full diagnostic protocol for overheating belongs to the overheating guide; this article’s job is the mechanism that protocol detects.
The test is most informative at low load: with the reverse bias applied and the load near idle, leakage dominates the budget and the loop is easiest to see. A case temperature that stabilizes within the soak window is the pass criterion; one that keeps climbing is the alarm.
Engineering note. The stability analysis above follows the leakage-temperature relationship published on Good-Ark Schottky datasheets—IR rising with junction temperature and reverse voltage, as illustrated by the AMBRP10H100 (10 µA at 25 °C to 2.5 mA at 100 °C) and ASGC051BS (5 µA at 25 °C to 5 mA at 125 °C) maximum figures—and the standard loss-curve-versus-heat-removal-line stability criterion. A fixed 80–90% junction band is not a universal rule—at the top of that band the margin to the absolute maximum is already thin; the design junction limit is set from the mission profile, part tolerance, aging, transient duty, and the reliability target, and confirmed with the part’s maximum leakage at the operating reverse voltage and the assembled thermal resistance.
Frequently Asked Questions
What is Schottky thermal runaway?
A positive feedback loop in which reverse leakage grows with junction temperature, leakage power heats the junction, and the heating raises leakage further. If the loss curve exceeds the heat-removal capability, the junction climbs until the part fails.
Why are Schottky diodes more exposed to thermal runaway?
Their low forward drop and higher leakage come from the same low barrier physics: the barrier that conducts well also leaks more, so the leakage term is larger than in a PN diode at the same voltage and temperature.
How do I check thermal stability?
Compare the total loss curve—forward plus leakage at the operating reverse voltage and temperature—with the heat-removal line of the assembled heat-flow route. If the loss curve stays below the line at the working condition, the device is stable; if it crosses, it is not.
What margin should I keep?
Set the design junction limit from the mission profile, part tolerance, aging, transient duty, and the reliability target—well below TJ(max)—with the leakage term included, and check that the loss curve stays below the heat-removal line at the worst ambient. A fixed 80–90% band is only a rough starting reference, not a rule. Re-run the check whenever voltage or ambient changes.
How do I test for it in a sealed product?
Soak the assembled unit at maximum ambient with reverse bias applied and the real load profile, then watch the case temperature after it appears to stabilize. A case that keeps climbing after the load settles means the loop is not at equilibrium.
Conclusion
Schottky thermal runaway is prevented by one comparison, not by hope: total loss at the hot working condition versus the heat the assembly can remove. Read the leakage curve at the real reverse voltage and junction temperature, keep continuous operation well below TJ(max), and re-run the stability check whenever voltage, ambient, or the heat-flow route changes.
Review the leakage curves on the AMBRP10H100 product page and datasheet, compare parts in the Schottky rectifier diodes category, and submit your reverse voltage, leakage, ambient temperature, thermal resistance, and measured temperature-rise data to Good-Ark for a stability review before finalizing the design.