How to Choose Between Unidirectional and Bidirectional TVS Diodes

A unidirectional TVS diode is usually the first choice for a positive DC power rail. A bidirectional TVS is usually more appropriate for an AC line or a signal that intentionally swings above and below ground. The decision is based on the normal voltage window and the required current path for each transient polarity, not on a general claim that one structure is stronger.

If a unidirectional device is placed directly across a bipolar signal, the normal negative swing may forward-bias it. If a bidirectional device is placed on a positive DC rail, it can clamp a positive surge, but it will not provide the same low-voltage forward path for a negative pulse that a unidirectional part provides.

How do the two TVS structures conduct?

A unidirectional TVS operates in avalanche when reverse voltage reaches its breakdown region. In the opposite polarity, it behaves much like a conventional forward-biased diode. On a positive rail, a positive surge is handled in avalanche, while a negative spike can be diverted through the lower forward voltage.

A bidirectional TVS operates through an avalanche path in either polarity. A positive or negative voltage must reach the corresponding breakdown region before strong conduction begins. It therefore avoids clipping the normal negative half-cycle of a bipolar waveform at an ordinary diode forward drop.

Polarity does not define pulse-power rating. Both structures can be produced in SMA, SMB, SMC, or much larger packages. Surge capability still depends on die size, package, pulse duration, current waveform, temperature, and derating.

Why do DC power rails commonly use unidirectional TVS diodes?

The normal voltage has one polarity

A 12 V, 24 V, or 48 V DC bus normally remains positive relative to its return. A unidirectional TVS is reverse-connected across the rail and remains off during normal operation. Its forward path can respond to a short negative transient at a lower voltage than a bidirectional avalanche path.

For a 24 V example, the ASIM SMA04J24V is a unidirectional device with a VRWM of 24 V, a VBR range of 26.7 to 29.5 V, and a maximum VC of 38.9 V at 10.3 A. The SMA04J24B is the bidirectional version with a similar voltage and pulse-power class, but the two parts behave differently during a negative event.

The highest continuous voltage must be known

A nominal 24 V supply is not always 24.0 V. Supply tolerance, charging state, ripple, hot-plug overshoot, and regenerative energy can raise the bus. If the maximum sustained value exceeds 24 V, selecting a 24 V VRWM part from the nominal label alone can produce leakage or heating.

Reverse-polarity protection is a separate function

A unidirectional TVS will conduct forward during sustained reverse battery or wiring reversal. If a fuse, current limiter, or electronic switch does not interrupt the fault quickly, the TVS may overheat. Long-duration reverse polarity should be managed by a MOSFET or another dedicated circuit; the TVS is primarily a transient device.

Which lines normally need a bidirectional TVS?

AC inputs, audio lines, bipolar sensor signals, and some communication buses operate on both sides of 0 V. If their normal negative excursion can exceed a unidirectional TVS forward voltage, a bidirectional structure or a purpose-built low-capacitance protection array is required.

Check positive and negative peak voltage

VRWM must cover the maximum sustained peak in both directions, not merely an RMS value. A 24 V RMS sine wave reaches about 34 V peak before tolerance or switching overshoot is included. A 24 V VRWM device would therefore be unsuitable across that waveform.

Include common-mode range

A differential signal may have a small line-to-line amplitude while each conductor moves through a much larger common-mode range relative to ground. Line-to-line and line-to-ground TVS connections create different voltage windows. CAN and RS-485 interfaces require the bus fault voltage and transceiver common-mode range to be included.

Test both transient polarities

A bidirectional junction may be electrically symmetrical, while the PCB return path, chassis connection, and cable coupling are not. Apply positive and negative test pulses and measure residual voltage at the protected pin for both polarities.

A practical selection sequence

First, draw the normal voltage envelope. Include startup, shutdown, charging, hot plug, regeneration, and any expected reverse condition. A waveform that persists during normal use should not drive the TVS into significant conduction.

Second, identify transient polarity, source impedance, and duration. A unipolar DC connector can face a positive surge, a negative pulse, and sustained reverse wiring. One protection component may not safely handle all three.

Third, check VRWM, VBR, VC at the target current, and the downstream absolute maximum rating. VBR is not a substitute for VC, and the table's maximum IPP is not automatically the current in the finished system.

Fourth, examine package and return path. Place the TVS near the energy entry point. Keep the discharge path short and wide, and do not route surge current through a sensitive digital-ground region.

Common application mistakes

Can a large bidirectional power TVS protect a high-speed signal?

It may clamp a surge, but its capacitance and package parasitics can degrade the signal. High-speed lines normally need a low-capacitance ESD device selected for the interface's working voltage, common-mode range, and channel layout.

Is a bidirectional TVS always wrong on a DC rail?

No. A bidirectional device can provide avalanche clamping for a positive surge, and some systems benefit from more symmetrical behavior. Its negative clamping path is different, so the decision must match the fault model.

What happens if a unidirectional TVS is installed backward?

It may be forward-biased by the normal positive supply, causing current limiting, heating, or fuse operation. PCB polarity marks and production inspection should make orientation unambiguous.

Common questions

Can a bidirectional TVS replace two series-opposed unidirectional parts?

The circuit concepts are related, but parasitics, power sharing, and clamping curves may differ. Verify the actual devices rather than substituting by symbol alone.

Must the TVS return connect to digital ground?

No. The best return may be chassis ground, protective earth, connector ground, or another low-impedance discharge reference. The goal is to keep transient current away from sensitive circuits.

Is surge testing still required after polarity is selected?

Yes. Datasheet ratings do not include every PCB inductance and system current path. Test both polarities and all relevant operating states on the final layout.

ASIM begins this choice by defining the normal voltage envelope, transient direction, and downstream withstand limit. Once those boundaries are clear, unidirectional versus bidirectional selection becomes a circuit decision rather than a package-label guess.