Diode Reverse Recovery Time Explained: Waveforms, Datasheets, and Selection

Diode turn-off behavior affects the performance of circuits that repeatedly transfer current between a diode and a switching device. Reverse recovery time describes part of this transition, but it does not independently determine switching loss or component suitability. t_rr, Q_rr, and peak recovery current depend on test conditions such as forward current, current slew rate, reverse voltage, and temperature. This article explains recovery behavior, its circuit effects, and the interpretation of diode specifications.

Catalog

What Diode Reverse Recovery Time Means

Diode reverse recovery time, t_rr, is a specified interval during the transition of a previously forward-conducting diode toward reverse blocking. In a pn junction diode, forward conduction stores minority carriers, and charge removal and recombination allow reverse current to flow briefly when the circuit commutates the diode. [1]

The reported t_rr value depends on the manufacturer’s measurement definition and the stated test conditions, including quantities such as forward current, current-change rate, reverse voltage, and temperature. [2, 7]

Understanding the Reverse Recovery Waveform

Figure 2 separates recovery duration, peak reverse current, recovery subintervals, and the current-time area used to describe recovered charge. The 10% endpoint shown is illustrative and does not define every manufacturer’s t_rr measurement; the APT30DQ120BG datasheet instead uses an extrapolated zero-current endpoint based on I_RRM and 0.25 × I_RRM. [7]

In Figure 2, i_D represents instantaneous diode current, and I_F represents the forward current before commutation. Positive current flows in the forward direction. The curve crosses zero, becomes negative, reaches a peak reverse-current magnitude identified as I_RRM, and then approaches zero.

The negative sign on −I_RRM indicates current direction. The datasheet may report I_RRM as a positive magnitude even though the corresponding waveform lies below the zero-current axis. Other manufacturers use symbols such as I_rr or I_RM, so the parameter description should be read together with its symbol.

Recovery time measures horizontal distance along the time axis. Peak reverse current measures vertical magnitude. Reverse recovery charge, Q_rr, describes the amount of charge represented by the current waveform over a defined integration interval. Charge is obtained from current integrated over time, so its unit is the coulomb, often expressed as nanocoulombs for diode specifications. [7]

These quantities cannot substitute for one another. A narrow pulse may have a large peak, while a wider pulse may have a smaller peak. A time value alone does not establish either pulse area. Similarly, the peak current does not establish how long the current persists.

The waveform’s return toward zero also matters. A gradual recovery tail and an abrupt termination can have similar headline timing values while presenting different current slopes to the circuit. Some manufacturers divide reverse recovery into t_a, from the current zero crossing to I_RRM, and t_b, from I_RRM to the specified recovery endpoint, and describe recovery softness with S = t_b / t_a. The exact definitions and measurement method should always be checked in the applicable manufacturer documentation.

Calculated Triangular-Pulse Example

Consider an ideal triangular reverse-current pulse with a peak magnitude of 2 A and a full base duration of 50 ns. These are chosen teaching values, not measured data or diode specifications.

For this defined shape, the charge equals the triangle’s area.

Q_triangle = ½ I_peak t_base

I_peak is the peak current magnitude in amperes. t_base is the full triangular duration in seconds. Q_triangle is the calculated charge in coulombs.

Q_triangle = ½ × 2 A × 50 ns = 50 nC

The units are consistent because amperes multiplied by seconds give coulombs. The calculation assumes straight pulse edges, zero current at both ends, and no additional tail.

The 50 ns base duration must not automatically be substituted for a manufacturer’s t_rr. A threshold-defined endpoint may occur before the waveform returns to zero, while an extrapolated endpoint may not lie on the measured curve. The triangle also differs from Figure 2’s curved tail. Its purpose is to demonstrate waveform area, not to estimate the recovery charge of an unspecified diode.

How Reverse Recovery Affects Circuit Performance

Figure 3 shows how current transfers from a conducting diode to a switch in a conceptual inductive stage. During recovery, an additional current component passes through the diode in reverse and increases the switch current.

In the left panel, the switch S is open and the inductor L carries an already established current through diode D. The diode’s anode A connects to the switching node, and its cathode K connects to the positive rail.

When S turns on, the inductor current transfers toward the switch. During diode recovery, an additional current component can flow from the positive rail through D in the reverse direction and then through S. The DC-link capacitor forms part of the local transient-current loop.

The switch therefore handles more than the continuing inductor current during this interval. The figure separates the current components to make their paths visible. It does not imply that the red current flows through an independent physical wire or that it persists after recovery.

In hard-switched operation, diode recovery can increase the turn-on loss of the associated transistor. The loss is not confined automatically to the diode. Its distribution depends on the circuit and the simultaneous voltage and current waveforms. [3]

An abrupt end to reverse current can also excite parasitic inductance and capacitance, producing voltage ringing. Reducing the inductance of the commutation loop can reduce this effect, while a softer recovery characteristic changes the current transition that excites it. Texas Instruments discusses this interaction in its explanation of body-diode recovery. [3]

This gives two distinct reasons to inspect a switching waveform. The current pulse helps explain additional switch loading, while the voltage waveform shows whether the transition produces overshoot or oscillation. A current-only illustration cannot establish the voltage stress on a component.

Slowing the associated transistor’s turn-on can reduce the imposed current-change rate, and a suitable snubber can control parts of the transient. These measures involve circuit tradeoffs rather than a universal improvement. Renesas identifies switching-speed adjustment, snubbing, and soft-recovery devices among approaches to recovery-related noise. [2]

A recovery-charge figure is insufficient to determine the exact heat generated in each device. Multiplying charge by a circuit voltage can support an approximate loss model, but it does not replace the actual voltage-current relationship throughout the event. [3]

Reverse-Recovery Energy and Switching Loss

For a measured recovery event, the general energy relationship is:

E_rr = integral of [v_D(t) × i_D(t)] dt

If E_rr is known for each switching event, its approximate average power contribution can be written as:

P_rr = E_rr × f_sw

where f_sw is the switching frequency. The integration interval and sign convention depend on the manufacturer or measurement method, so the applicable definition should be used when comparing or calculating recovery loss.

Comparing Reverse Recovery Across Diode Types

Diode technology helps identify which switching behavior deserves attention. It does not provide a complete performance ranking without device specifications and operating conditions. Table 1 compares the charge mechanisms and the parameters that distinguish them. [1, 5]

Device / Recovery Class	Recovery Behavior	Specification Focus	Selection Limitation
Standard silicon pn rectifier	Stored-charge recovery follows forward conduction.	Available recovery data and intended operating conditions.	Static voltage and current ratings do not establish switching suitability.
Fast or ultrafast silicon pn rectifier	Designed to reduce or control recovery behavior.	trr, Qrr, peak reverse current, and recovery shape.	A speed label does not replace condition-matched data.
Silicon Schottky diode	Normal Schottky conduction avoids conventional minority-carrier storage recovery.	Capacitive behavior, forward voltage, and reverse leakage.	Absence of a recovery tail does not mean zero switching current.
SiC Schottky diode	Majority-carrier operation avoids the conventional pn recovery mechanism.	Capacitive charge, voltage rating, and thermal characteristics.	SiC technology alone does not determine total circuit loss.

Vishay’s VS-SC200FA120 datasheet illustrates the distinction by specifying capacitive charge for a SiC Schottky device. Such a quantity should not be relabeled as pn-junction recovery charge or omitted simply because the device is described as having negligible recovery. [5]

Voltage capability is device-specific. For example, Vishay’s silicon Schottky V20100SG has a maximum repetitive peak reverse voltage, V_RRM, of 100 V, while the SiC Schottky VS-SC200FA120 has a 1200 V cathode-to-anode voltage rating. These are product ratings, not universal technology limits or recommended operating targets. [4, 5]

Forward voltage, reverse leakage, and capacitance remain operating-condition-dependent tradeoffs and should be compared using the applicable datasheet values. [6]

Reading Reverse Recovery Specifications in Datasheets

A recovery specification belongs to a particular test condition. Table 2 reproduces typical dynamic values for Microsemi’s APT30DQ120BG, a TO-247 diode, from Revision D, March 2018. All three rows use I_F = 30 A and V_R = 800 V; T_C and di_F/dt vary as shown. Here I_F is forward current, V_R is reverse voltage, and T_C is case temperature. [7]

Case Temperature	Forward-Current Change Rate	Typical trr	Typical Qrr	Typical IRRM
25 °C	−200 A/µs	320 ns	545 nC	4 A
125 °C	−200 A/µs	435 ns	2100 nC	9 A
125 °C	−1000 A/µs	180 ns	2975 nC	28 A

At T_C = 125 °C, increasing the magnitude of di_F/dt from 200 A/µs to 1000 A/µs produces a shorter typical t_rr but larger Q_rr and I_RRM. Comparing the two −200 A/µs rows shows that the published recovery values also depend on temperature. [7]

A typical value describes characteristic behavior under the stated conditions. It should not be promoted to a guaranteed upper bound. A maximum entry provides a different kind of specification, while an absolute maximum rating defines a stress boundary rather than an intended operating target. Blank minimum or maximum cells do not become implied limits.

For Table 2, t_rr extends from the current zero crossing to the zero-current intercept of a straight line through the recovery tail at I_RRM and 0.25 × I_RRM. Figure 2 uses an illustrative 10% endpoint instead, so the two endpoint conventions must not be treated as interchangeable. [7]

Selecting a Diode for the Switching Circuit

Start with the actual commutation conditions: forward current before turn-off, reverse voltage and transient stress, temperature, switching frequency, and di_F/dt. First eliminate candidates that do not meet the required voltage, current, thermal, and package limits; average, repetitive peak, and non-repetitive surge ratings are not interchangeable. [6]

Among electrically suitable candidates, compare t_rr, Q_rr, I_RRM, and recovery shape at test conditions close to the application. A shorter t_rr alone does not prove lower switch-current stress or lower loss, and room-temperature data should not be treated as equivalent to elevated-temperature data.

If changing from a pn diode to a Schottky device, also evaluate capacitance, leakage, forward conduction loss, and thermal behavior. The preferred device is the one whose complete electrical and thermal behavior matches the circuit objective, not simply the one with the smallest isolated recovery-time value. [4–6]

When published data do not establish the expected benefit, validate the candidate with a suitable fixture and correctly rated instrumentation under representative operating conditions. Interpret current spikes and ringing together with the circuit and measurement arrangement rather than attributing them solely to the diode.

Conclusion

Reverse recovery is a property of the diode’s transition within a circuit, and its time specification represents only one part of that behavior. A defensible component comparison connects recovery duration, current, charge, and waveform shape with the stated test conditions and operating limits. The useful result is a diode that meets the circuit’s combined requirements, rather than the device with the shortest isolated recovery-time value.

Technical References:

[1] Toshiba Electronic Devices & Storage Corporation. What Is Diode Reverse Recovery Time (t_rr)? Undated educational webpage.

[2] Renesas Electronics. Power Diodes, Descriptions of Electrical Characteristics and Usage. R07AN0042EJ0100, Revision 1.00, November 13, 2024.

[3] Steve Widener, Texas Instruments. GaN to the Rescue! Part 1, Body-Diode Reverse Recovery. SSZTC00, October 2015.

[4] Vishay General Semiconductor. V20100SG, VI20100SG, High-Voltage Trench MOS Barrier Schottky Rectifier. Document 89153, technical revision November 9, 2017.

[5] Vishay Semiconductors. VS-SC200FA120, SOT-227 Silicon Carbide Schottky Barrier Diode, 1200 V, 200 A. Document 97150, revision April 15, 2024.

[6] Travis Eichhorn, Texas Instruments. Schottky Diode Selection in Asynchronous Boost Converters. Application Report SNVA762, September 2016.

[7] Microsemi. APT30DQ120BG, Ultrafast Soft Recovery Rectifier Diode. Datasheet, Revision D, March 2018. Hosted by Microchip Technology.

[8] Infineon Technologies. Understanding and Interpreting the CoolSiC MOSFET 1200 V Datasheet. AN2025-10, Version 1.1, July 15, 2026, Section 6.1.

[9] Tektronix. Making Accurate Current Measurements on Power Supplies with Oscilloscopes. Application note, accessed September 28, 2026.

Frequently Asked Questions [FAQ]

1. Does a SPICE Model Always Predict Reverse Recovery Accurately?

No. A model may reproduce some switching characteristics without accurately representing recovery behavior across temperature and operating conditions. Before relying on its predictions, confirm the model’s documented scope and compare its recovery response with manufacturer data or suitable measurements. [3]

2. Can Dead Time Affect Body-Diode Reverse Recovery?

Yes. In a half-bridge, dead time can determine how long a MOSFET body diode conducts before commutation, which can affect subsequent recovery losses. Shortening dead time must still prevent simultaneous conduction of the two switches, so the applicable device and gate-driver guidance should determine the timing limits. [8]

3. Does a SiC MOSFET Body Diode Behave Like a SiC Schottky Diode?

No. A SiC MOSFET’s intrinsic body diode can exhibit minority-carrier reverse recovery, unlike normal majority-carrier operation in a SiC Schottky diode. Confirm the body-diode recovery characteristics and reverse-conduction conditions in the applicable MOSFET datasheet rather than applying Schottky-diode assumptions. [8]

4. Why Should Voltage and Current Probes Be Deskewed for Recovery-Loss Measurements?

Voltage and current probes can introduce different propagation delays, making the displayed waveforms appear misaligned. Multiplying those misaligned signals can produce an inaccurate switching-energy result. Deskew the channels using the instrument manufacturer’s procedure before evaluating voltage-current overlap. [9]

5. Is Forward Recovery Time the Same as Reverse Recovery Time?

No. Forward recovery describes the diode’s turn-on voltage transient, while reverse recovery concerns its transition from forward conduction toward reverse blocking. Where turn-on behavior affects the circuit, check the separately specified forward recovery time and forward recovery voltage under the applicable datasheet conditions. [2]