Forward voltage (VF) is not a fixed datasheet number; it is a function of current, junction temperature, and device construction. Design with the maximum VF at the operating temperature, and the loss math—conduction loss is simply IF × VF—stops producing surprises. This guide covers the physics behind the number, how it moves with temperature, and how to read it correctly.
Where Forward Voltage Comes From
VF is the voltage needed to push current across the rectifying junction, and it has two sources. The first is the potential barrier of the junction itself: for a Schottky that is the metal-semiconductor barrier; for a PN diode it is the built-in potential of the p-n junction. The second is the ohmic resistance of the silicon die, the metallization, and the package connections, which behaves like a series resistor.
The two sources respond to current differently. At low to moderate current densities, the barrier term dominates and VF rises logarithmically with current—slowly. At high current densities, the series-resistance term takes over and VF rises more steeply, which is why the same part can look very efficient at 5 A and much less efficient at its 40 A headline. Reading VF without its current and temperature conditions is meaningless; the number only exists at a stated IF and TJ.
The series-resistance term deserves a second look because it changes the shape of the VF curve at high current: a part optimized for low VF at 5 A can show a noticeably steeper VF slope at 40 A, and the efficiency comparison between two parts depends on which point of the curve the application actually occupies.
Why VF Falls as Temperature Rises—With Conditions
Over most of a power diode’s operating range, VF falls as junction temperature rises, at roughly a few millivolts per degree Celsius. The barrier height decreases with temperature, so the same forward current needs less voltage at a hotter junction. The AMBRP10H100 illustrates the size of the effect with datasheet-published numbers: 0.76 V typical at 10 A and 25 °C, versus 0.62 V typical at 10 A and 125 °C—about 1.4 mV/°C across that 100 °C span.
The trend is not a law. At high current densities, the series-resistance term rises with temperature, and a large enough resistive component can flatten or reverse the coefficient. Whether VF falls, holds, or rises depends on the current density and the device structure, so the correct habit is to read the VF curves of the selected part at the operating current and temperature rather than applying a blanket “VF always falls” rule.
The practical reading is a two-point check: read VF at the operating current at both 25 °C and the expected hot junction, and note the direction of the change. If the coefficient is negative, the conduction term will shrink at temperature—but the leakage term will grow, so the thermal picture is never complete with VF alone.
The Loss Math: IF × VF at the Operating Point
Conduction loss is the product of forward current and forward voltage at the working condition: PD = IF × VF. Because both IF and VF are operating-point values, the calculation must be done at the real current and the real junction temperature.
Worked check on the same 10 A part: at 25 °C and 10 A, 0.76 V typical gives about 7.6 W of conduction loss; at 125 °C, 0.62 V gives about 6.2 W. The conduction term shrank with temperature—but the reverse leakage term grew, and leakage power adds heat to the same junction. The correct thermal picture adds both: forward loss at the hot VF plus leakage power at the hot junction, then compares the total with the heat-removal capability of the assembly. Conduction loss alone is only half the story; the leakage interaction is covered in the thermal-runaway guide.
The same arithmetic works for the freewheeling path: a diode that carries current during the off-time dissipates the same IF × VF product, and the duty cycle decides how the loss averages over the switching period. The formula never changes—only the current and temperature at which VF is read do.
Typical vs Maximum: Which Number to Design With
Datasheets publish VF twice: a typical value and a maximum value at the same conditions. Typical describes the center of the distribution; maximum describes the guarantee. Design budgets use the maximum at the operating temperature.
The AMBR40250S shows why the distinction matters: at 20 A and 25 °C, VF is 0.83 V typical and 0.90 V maximum; at 20 A and 125 °C, 0.69 V typical and 0.77 V maximum. At 20 A, the difference between typical and maximum is about 1.4 W of conduction loss—enough to change a marginal thermal design. Read both columns, use the maximum for the maximum budget, and keep the typical number for efficiency estimates that are explicitly labeled as typical.
The maximum column also absorbs lot-to-lot variation, which is why it is the right input for bounding thermal design: two units from different lots can sit at opposite ends of the typical-to-maximum band, and the thermal budget must close for both.
VF Across Families: Schottky, FRD, and Silicon
| Family | VF at moderate current (class-typical) | Temperature behavior | Main tradeoff |
|---|---|---|---|
| Silicon Schottky | 0.3–0.6 V | Falls, but conditional on current density | Higher leakage |
| Fast recovery (PN) | 0.8–1.5 V | Falls slowly; recovery charge adds loss | Recovery at high frequency |
| Standard silicon | 0.7–1.1 V | Falls slowly | Slow recovery |
The ranges are class-level guidance for comparison, not datasheet claims for any specific part. The practical reading: the Schottky wins the conduction comparison at low-voltage outputs, the FRD is competitive only where its recovery speed justifies the higher VF, and the standard silicon diode is the low-frequency value choice. What the table cannot show is the leakage side of the Schottky’s low VF, which is the point of the next section.
The family ranges also shift with current density: a Schottky at 1 A can sit near 0.4 V while the same part at 40 A moves toward 0.9 V as the series-resistance term grows. Compare at the application’s current, not at each family’s best-looking point.
The Price of Low VF: Leakage and Surge Tradeoffs
A low forward drop is not free. In a Schottky, the low barrier that produces low VF is the same low barrier that produces higher reverse leakage, and leakage grows with junction temperature and reverse voltage. The same part that conducts beautifully at 10 A can contribute a meaningful leakage term at the hot junction in a sealed enclosure. The surge story is similar: high-voltage Schottky classes trade surge and leakage characteristics for blocking capability, so the low-VF advantage is always evaluated together with the reverse and transient budgets. Selection logic for these tradeoffs is covered in the Schottky selection guide; this article’s job is to make the VF side of the equation legible.
The surge side is the other half of the price: high-voltage Schottky classes trade leakage and surge behavior for blocking capability, so a low-VF part selected at the wrong voltage class can fail the very transient it was meant to survive. The full tradeoff logic belongs to the selection guide; the point here is that the number is never evaluated alone.
Engineering note. The VF analysis above follows the datasheet structure of Good-Ark rectifiers—VF stated at a forward current and junction temperature, with typical and maximum columns, and curves that show the temperature behavior of the selected part. The 0.76 V/0.62 V and 0.83 V/0.90 V numbers are datasheet-published typical and maximum values; the family ranges are class-level guidance. Confirm the temperature coefficient of the specific part at the operating current before relying on it in a loss budget.
Frequently Asked Questions
Why is forward voltage not a fixed number?
Because VF depends on forward current, junction temperature, and device construction. The datasheet value only exists at a stated IF and TJ, and both must be matched to the working condition before using it in a loss calculation.
Does VF always decrease with temperature?
Not unconditionally. The barrier term falls with temperature, but the series-resistance term rises, so at high current densities the trend can flatten or reverse. Read the part’s VF curves at the operating current and temperature.
Should I use typical or maximum VF?
Maximum, for the bounding loss budget; typical, for explicitly labeled efficiency estimates. The difference is real—at 20 A, 0.83 V typical versus 0.90 V maximum is about 1.4 W.
How do I calculate conduction loss?
Multiply the forward current by the forward voltage at the working condition: PD = IF × VF. Add the leakage term at the hot junction and compare the total with the assembly’s heat-removal capability.
Why do Schottky diodes have lower VF but higher leakage?
The low metal-semiconductor barrier that reduces forward voltage is the same low barrier that increases reverse leakage. The two tradeoffs come from the same physics and must be designed together.
Conclusion
Forward voltage is an operating-point function, not a constant: read it at the real current and junction temperature, use the maximum value for bounding budgets, and treat “VF falls with temperature” as a conditional trend rather than a rule. Conduction loss is IF × VF, but the thermal picture is only complete when leakage is added at the hot junction.
Compare low-VF parts side by side in the Schottky rectifier diodes category on the Good-Ark site, and contact Good-Ark with your operating current, voltage, and thermal conditions to confirm VF curves and samples for your design.