Full-Bridge Converter

The four-switch isolated buck — two switch legs drive the transformer with the full ±Vin while each switch still blocks only Vin. Full drive voltage at half-bridge stress: that is why it owns the top of the power range, from a few hundred watts to many kilowatts.

Table of Contents:

Introduction — finishing what the half-bridge started

The half-bridge converter made one great trade: by driving the transformer from a capacitor mid-rail at Vin/2, it cut the switch stress to just Vin. But that trade has a hidden bill — at half the voltage, the primary must carry twice the current for the same power. Push a half-bridge past a few hundred watts and its switches and windings start cooking on conduction loss.

The full-bridge converter collects the best of both worlds. It replaces the half-bridge’s two divider capacitors with a second switch leg, so the transformer primary hangs between two switching nodes. Turning on a diagonal pair of switches applies the full +Vin across the primary; the opposite diagonal applies −Vin. The primary gets the full drive voltage — so the current halves back down — yet each switch still never sees more than Vin. Full voltage and low stress. The price: four switches, two of them floating high-side drives.

This page builds the full-bridge from that idea: the two-leg power stage, what each diagonal pair does, the waveforms on a scope, the series inductor Ls and the duty-cycle loss it causes, the phase-shift drive that turns this topology into the soft-switched workhorse of modern multi-kilowatt supplies, the formulas, and a worked example that lines the whole isolated-buck family up side by side.

What is a Full-Bridge Converter?

A full-bridge converter (or H-bridge converter) is an isolated switch-mode DC-DC converter that behaves like a buck converter fed through a transformer driven at the full input voltage in both directions. It is built from these parts:

  • Four switches (MOSFETs) in two legs. Leg 1 is QA (high-side) over QB (low-side), meeting at node A; leg 2 is QC over QD, meeting at node B. Both legs sit directly across Vin.
  • A single-primary transformer between nodes A and B. The bridge voltage vAB drives it at ±Vin — both magnetic directions, so no reset winding is needed.
  • A series inductor Ls in the primary loop — partly the transformer’s own leakage, often topped up with a small discrete inductor. It looks like a parasitic, but it is a design element: it sets the duty-cycle loss and (in phase-shifted designs) provides the energy for zero-voltage switching.
  • A four-diode bridge rectifier on the single secondary winding. One diagonal diode pair (D1+D4) conducts when the secondary is positive, the other (D2+D3) when it is negative.
  • An output L-C filter (Lo, Co) that smooths the rectified pulses into clean DC, exactly as in a buck.

The heart of it is the diagonal drive. QA+QD on together connects node A to Vin and node B to ground: vAB = +Vin. Half a period later QB+QC reverse it: vAB = −Vin. The transformer sees a symmetric AC square wave at the full rail — twice the half-bridge’s swing — and whichever switches are off are clamped by their own leg to exactly Vin.

The one-sentence version. A full-bridge converter uses two switch legs to drive a transformer with the full ±Vin — diagonal pairs firing alternately — so the primary current is half a half-bridge’s at the same power while each switch still blocks only Vin; four switches buy you the top of the power range.

Block Diagram

Block diagram of a full-bridge converter: a DC input feeds a four-switch bridge of two legs driven by a PWM controller, then a series inductor and single-primary isolation transformer, then a four-diode bridge rectifier, then an L-C output filter, then the load, with an isolated feedback path returning to the controller.
Figure 1: Full-bridge converter block diagram — a four-switch bridge drives the transformer at the full ±Vin into a bridge rectifier and L-C filter

Read it left to right as a power path: the DC input is chopped by the four-switch bridge, driven through the series inductor and transformer, rectified by the four-diode bridge, smoothed by the L-C filter, and delivered to the load. Two things mark it as a full-bridge. First, there is no capacitor divider — the second switch leg replaces it, so the transformer swings the full ±Vin. Second, the controller fires diagonal pairs (or, in the phase-shifted variant, slides one leg’s phase against the other). As always, feedback crosses the isolation barrier through an opto-coupler.

Circuit Diagram & Construction

Circuit diagram of a full-bridge converter: four MOSFETs QA, QB, QC, QD in two legs across the DC input Vin form switching nodes A and B; a series inductor Ls and the single transformer primary connect node A to node B; the single secondary winding feeds a four-diode bridge rectifier D1 to D4 whose output feeds an output inductor Lo, capacitor Co and load R producing Vo.
Figure 2: Full-bridge converter power circuit — two switch legs drive the single-primary transformer through Ls; a four-diode bridge rectifies into the Lo-Co filter

This is the power stage as the TI seminar’s Figure 6 presents it, redrawn and re-labelled. Tracing it across:

Switch legs QA/QB and QC/QDTwo half-bridge legs directly across Vin, making switching nodes A and B. Within each leg the two switches are complementary (never both on); across the bridge, diagonal pairs conduct together. QA and QC need floating high-side gate drives; each switch blocks only Vin.
Series inductor LsThe transformer’s leakage inductance plus, usually, a small discrete “shim” inductor. It slows the primary-current polarity reversal (causing duty-cycle loss, below) and stores the energy that makes zero-voltage switching possible in phase-shifted designs.
Transformer (single primary Np, single secondary Ns)Driven between nodes A and B at ±Vin. Symmetric AC drive — the core swings both ways, no reset winding, no air gap. Simple two-winding construction: the best copper utilisation in the whole isolated family.
Bridge rectifier D1–D4Four diodes on the single secondary. Diagonal pair D1+D4 conducts when the secondary voltage is positive, D2+D3 when negative; all four share the current during the dead time. Each diode blocks only n·Vin (= Vo/Deff).
Output filter Lo, CoA buck-type L-C filter. Fed twice per period, so the inductor current ripples at twice the switching frequency.
Never let the two switches of one leg overlap. QA+QB (or QC+QD) on together would short Vin straight through that leg — destructive shoot-through. Each leg gets a guaranteed dead time, which caps each diagonal pair below 50 % duty. Note also that this circuit, unlike the half-bridge, has no series DC-blocking capacitor in its basic form — so a persistent volt-second imbalance between the two half-cycles can walk the core toward saturation. Practical designs prevent it with peak-current-mode control, or by adding a small series blocking capacitor.

Principle of Operation

The four gates come from one controller. In the classic hard-switched drive, the two diagonal pairs are fired alternately, each for a fraction D of the full period (D < 0.5), 180° apart. Over one period there are four intervals: pair 1 on, dead time, pair 2 on, dead time:

  • When QA+QD are on, node A is at Vin and node B at ground, so vAB = +Vin. The secondary produces +n·Vin, diagonal D1+D4 conducts, and energy passes straight through to the Lo-Co filter. The output inductor charges.
  • When QB+QC are on, the bridge reverses: vAB = −Vin. The secondary produces −n·Vin, the other diagonal D2+D3 conducts, and the filter is fed again. The core magnetises the opposite way — resetting itself.
  • In the dead times (all four off), the transformer is quiet but the output inductor keeps its current flowing: all four rectifier diodes conduct, the secondary is effectively shorted, and the inductor freewheels.

So the load is fed twice per switching period and the transformer sees a symmetric AC square wave at the full rail. The output voltage follows from volt-second balance on the output inductor, exactly as in a buck — the transformer just scales the pulse height by n.

Modes of Operation

A full-bridge has two driven states — one per diagonal pair — separated by dead times in which the output inductor freewheels. Trace each state on the power circuit of Figure 2 above; the current path is described in words below.

Mode 1 — QA + QD ON (vAB = +Vin)

  • Current flows from the + rail through QA to node A, through Ls and the primary to node B, and down through QD to the − rail. The full Vin sits across the Ls-primary loop, and the core magnetises in the “positive” direction.
  • On the secondary, +n·Vin forward-biases the diagonal D1 (secondary top → + rail) and D4 (− rail → secondary bottom). The inductor Lo sees (n·Vin − Vo) and its current ramps up, delivering power to the load.
  • QB and QC are off; each is clamped by its own conducting leg partner to exactly Vin. D2 and D3 block n·Vin.

Mode 2 — QB + QC ON (vAB = −Vin)

  • Half a period later the opposite diagonal conducts: current flows through QC to node B, backwards through the primary and Ls to node A, and down through QB. The primary sees the full Vin in the opposite polarity, so the core magnetises the other way — this is what resets the flux without any reset winding.
  • The secondary now produces −n·Vin, forward-biasing the other diagonal D3+D2, so the output inductor charges exactly as in Mode 1. From the output’s point of view the two modes are identical.
  • QA and QD are now off, each blocking Vin.

Dead time — all four off (freewheel)

  • Between the driven states, all four switches are off. The transformer delivers no power, but Lo’s current cannot stop.
  • With no secondary polarity enforced, all four rectifier diodes conduct together, splitting the inductor current between the two diagonals. The secondary is effectively short-circuited, so the inductor sees −Vo and its current ramps gently down until the next pair fires.
  • Because the shorted secondary also clamps the primary voltage to zero, the magnetising current simply circulates (reflected into the secondary loop) and the switching nodes settle toward Vin/2. This freewheel is the same job a buck’s freewheel diode does.

Waveforms Explained in Detail

Full-bridge converter waveforms over two switching periods: the two diagonal-pair gate signals firing alternately; the bridge output voltage swinging plus and minus the full input voltage; the small bipolar magnetising current; the bipolar primary current of about one ampere; the smooth output-inductor current rippling at twice the switching frequency; the two rectifier-diagonal currents which always sum to the inductor current; and the voltage across one switch, never exceeding the input voltage.
Figure 3: Full-bridge waveforms — diagonal-pair gates, vAB (±Vin), magnetising and primary currents, inductor current, rectifier diagonals and switch voltage (peak Vin)

Read the traces together and the whole converter is visible in one picture:

  • Gates. The two diagonal pairs fire alternately, each for D of the full period, 180° apart, never overlapping. Within each leg, the two switches are complementary with a dead-time guard.
  • Bridge voltage vAB. A symmetric AC square wave at the full ±Vin (±100 V here) — twice the half-bridge’s swing. This is the whole point of the second leg.
  • Magnetising current im. Ramps up under one pair, down under the other, swinging symmetrically about zero — the core resets itself every period.
  • Primary current ipri. The reflected load current n·iL plus magnetising — only about ±1 A here. A half-bridge doing the same 80 W carries about ±2 A. Halving the primary current (and its conduction loss) is the full-bridge’s scaling advantage.
  • Inductor current iL. A smooth triangle rippling at twice the switching frequency, because the filter is topped up twice per period. Its average is the load current.
  • Rectifier diagonals. Each diagonal pair carries the full inductor current while its switch pair drives, and half during the dead-time freewheel. At every instant they sum to iL — the output never breaks.
  • Switch voltage vQB. Zero while QB conducts, about Vin/2 in the dead time, and Vin = 100 V while QA holds node A high. Full drive voltage, half-bridge stress.

Series Inductor Ls & Duty-Cycle Loss

Look back at the circuit: between the bridge and the transformer sits the series inductor Ls — the transformer’s leakage plus, in most real designs, a deliberate extra few microhenries. It creates the full-bridge’s most distinctive non-ideality, and understanding it is what separates a paper design from a working one.

Three stacked waveforms showing duty-cycle loss in a full-bridge converter. Top: the bridge voltage vAB as a clean bipolar square wave. Middle: the primary current, which ramps through zero with a finite slope at each polarity change because of the series inductor. Bottom: the transformer secondary voltage, whose pulses start late and are narrower than vAB; the lost slice at each edge is shaded and labelled duty-cycle loss.
Figure 4: Duty-cycle loss — while Ls reverses the primary current, the transformer sees zero volts, so each secondary pulse is narrower than vAB (reversal drawn exaggerated for clarity)

Here is the mechanism, one step at a time. When the bridge flips vAB from one polarity to the other, the primary current must also reverse — but it flows through Ls, and an inductor’s current cannot jump. So for a short interval the entire applied Vin sits across Ls, ramping its current through zero at a slope of Vin/Ls, while the transformer itself sees zero volts. During that interval the secondary delivers nothing. The result: each secondary pulse (vsec) starts late and is narrower than the vAB pulse that commanded it. The difference is called duty-cycle loss, and the effective duty that actually reaches the output is Deff < D. The larger Ls (or the heavier the load current it must reverse), the larger the loss.

So why keep Ls at all? Because that same stored energy is what discharges the MOSFETs’ output capacitance during the dead time and makes zero-voltage switching possible (next section). Ls is the classic engineering trade: more of it means easier soft switching but more lost duty (which the designer buys back with a slightly bigger turns ratio). A few percent of duty-cycle loss is the typical price.

Phase-Shift Control & ZVS — the modern way to drive it

Everything above used the classic drive: fire the diagonal pairs together, vary their width. Modern high-power designs almost always use a smarter scheme — the phase-shifted full-bridge (PSFB):

  • Each leg runs at a fixed ~50 % duty (complementary top/bottom switches with dead time), so every node swings rail-to-rail every period, always.
  • The control variable is the phase between leg 1 and leg 2. When the legs are in phase, the two ends of the primary move together and vAB = 0 — no output. As leg 2’s timing slides against leg 1’s, an overlap opens up in which one node is high while the other is low, and vAB carries ±Vin for that overlap. The phase shift is the effective duty cycle.
  • Zero-voltage switching (ZVS) comes free. In the dead time before each switch turns on, the current stored in Ls keeps flowing and discharges that switch’s output capacitance (Coss), pulling its drain-source voltage to zero — the current briefly flows backwards through the switch position — before the gate rises. The switch then turns on with no voltage across it, so the turn-on switching loss essentially vanishes.

This is why the phase-shifted full-bridge dominates the multi-kilowatt world: it keeps the full-bridge’s low stresses and full drive voltage, and removes most of the switching loss that would otherwise cap the switching frequency. A representative modern design point: a 54 V, 3 kW server/telecom rectifier stage switching at 140 kHz. The catch is light load: with little current in Ls, there is not enough energy to discharge Coss, ZVS is lost, and the controller must manage it (burst modes, adaptive dead time). The PSFB deserves — and will get — its own page; here it is enough to see that it is this same power stage, driven by phase instead of pulse width.

Output Rectifier Options

Figure 2 shows the four-diode bridge, but the full-bridge primary works with any full-wave secondary. The three standard choices, and when each wins:

RectifierDiodes / FETsDevice voltage stressBest suited for
Full-bridge (Fig 2)4Vo/Deff = n·Vin (lowest)Higher output voltages — two devices conduct in series, but each blocks only half what a centre-tap’s would.
Centre-tapped22·Vo/Deff = 2·n·VinLow output voltage / high current — only ONE diode drop in the path, but each device blocks double and the winding needs a centre tap.
Current doubler2 (+ two output inductors)Vo/DeffVery high current outputs — each inductor carries half the load current and the secondary is a simple single winding; the trade is Deff capped at 50 %.

In high-current designs all of these are usually built with synchronous rectifiers — MOSFETs in place of diodes, gated to conduct when their diode would — recovering the diode drop as efficiency. Our worked example keeps ideal diodes for clarity.

Key Formulas

All from volt-second balance on the output inductor, with n = Ns/Np and D the duty of each diagonal pair (D ≤ 0.5). Deff = 2D is the fraction of the period the secondary carries voltage.

Output voltage (CCM). The filter is fed n·Vin for 2D of the period and 0 for the rest:

(n·Vin − Vo)·2D·T = Vo·(1 − 2D)·T  →  Vo = 2 · n · D · Vin  (= n · Deff · Vin)

Twice the half-bridge’s Vo = nD·Vin at the same turns ratio — the full ±Vin drive restores the factor of 2.

Duty ratio for a target output. Rearranging:

D = Vo / (2 · n · Vin)   (must stay < 0.5)

Inductor-current ripple and average. Standard buck filter, ripple repeating every half period:

ΔiL = (n·Vin − Vo)·D·T / Lo   |   IL,avg = Io

Device stress. Each off switch is clamped by its own leg; each bridge-rectifier diode sees the secondary pulse:

Switch:  VQ,off = Vin   |   Diode:  VD,rev = n · Vin = Vo/Deff

Currents. The primary carries the reflected load current at full drive voltage:

Ipri ≈ n · Io   |   Iin,avg = 2 · D · n · Io = Po / Vin

Duty-cycle loss. The slice of each pulse spent reversing Ls’s current:

ΔD ≈ 2 · Ls · n · Io / (Vin · T)  per edge  →  use Deff = 2D − ΔDtotal in the gain formula

(Switch and diode drops, winding resistance and losses are neglected; a hard-switched full-bridge is typically 90–94 % efficient, and a well-designed phase-shifted version with synchronous rectifiers can exceed 96 %.)

Voltage Gain & Turns Ratio

Voltage gain of a full-bridge converter versus per-pair duty ratio for a turns ratio of 0.25. The gain M equals two n times D is a straight line from the origin, twice the slope of a half-bridge converter shown dashed at the same turns ratio. A vertical dashed line marks the hard duty limit at 0.5 per diagonal pair. The operating point at duty 0.4 gives a gain of 0.2.
Figure 5: Full-bridge gain M = 2n·D — twice a half-bridge’s at the same n, capped at the D = 0.5 shoot-through limit

The full-bridge’s gain is a straight line, M = 2n·D — easy to control, and twice a half-bridge’s at the same turns ratio, because the primary is driven at Vin instead of Vin/2. Like all its buck-derived relatives it is fundamentally a step-down through a transformer. Key points:

  • The turns ratio does the coarse scaling. Here n = 0.25 puts the 100 V → 20 V operating point at a comfortable D = 0.4 — the same n a push-pull would use, half what a half-bridge needs.
  • The duty is hard-capped at 0.5 per pair by the leg dead times, so the maximum output is n·Vin. Phase-shifted designs replace D with Deff, slightly smaller because of duty-cycle loss.
  • Continuous inductor current keeps it linear. Very light load can push Lo into discontinuous conduction, where the output rises above 2nD·Vin and becomes load-dependent; the inductor is sized to stay continuous over the working range.

Worked Example — and the family showdown

One consistent design, used by every figure above. Follow the numbers and the formulas check out.

GivenVin = 100 V, Ns:Np = 1:4 (n = 0.25), D = 0.4 per pair, f = 50 kHz (T = 20 µs), Lo = 100 µH, Lm = 1 mH
Primary swing   ±Vin±100 V (double a half-bridge’s ±50 V)
Output voltage   Vo = 2·n·D·Vin2 × 0.25 × 0.4 × 100 = 20 V
Load (for Io = 4 A)   R = Vo/Io, Po = VoIoR = 5 Ω, Po = 80 W
Secondary pulse voltage   n·Vin0.25 × 100 = 25 V (applied Deff = 2D = 0.8 of the period)
Inductor ripple   ΔiL = (n·Vin−Vo)·D·T/Lo(25−20) × 0.4 × 20µ/100µ = 0.4 A at 2f (Imax 4.2, Imin 3.8 → CCM)
Magnetising swing   Δim = Vin·D·T/Lm100 × 0.4 × 20µ/1m = 0.8 A bipolar (±0.4 A about zero)
Switch voltage stress   Vin100 V (same as half-bridge; a push-pull would be 200 V)
Rectifier diode reverse   n·Vin = Vo/Deff25 V (= 20/0.8 ✓; a centre-tap secondary would double this to 50 V)
Primary current (driving)   ≈ n·Io≈ 1 A (a half-bridge doing the same job carries ≈ 2 A)
Input current (avg)   2·D·n·Io2 × 0.4 × 0.25 × 4 = 0.8 A → Pin = 100 × 0.8 = 80 W = Po ✓

The last row is the lossless sanity check: 80 W in, 80 W out. And now the payoff of using the same 100 V → 20 V / 4 A job on all three double-ended pages — the whole trade-off landscape in one table:

Same job: 100 V → 20 V, 4 APush-pullHalf-bridgeFull-bridge
Primary drive±Vin = ±100 V±Vin/2 = ±50 V±Vin = ±100 V
Turns ratio needed0.25 (per half)0.50.25
Switch voltage stress200 V (2·Vin)100 V100 V
Primary current (approx.)≈ 1 A≈ 2 A≈ 1 A
Switches / floating drives2 / none2 / one4 / two
Sweet spotLow-voltage input, 150 W–1 kWOff-line, 100–500 WOff-line, 500 W and up

Each column pays for its advantage somewhere else: the push-pull’s simple ground drive costs double switch stress; the half-bridge’s low stress costs double current; the full-bridge pays in switch count and gate-drive complexity to get full voltage and low stress — which is exactly the trade you want once the power is high enough to make silicon cheaper than heat.

Advantages & Disadvantages

Advantages

  • Full drive voltage at low stress. The primary sees ±Vin yet each switch blocks only Vin — the best voltage economics in the isolated family, ideal off a rectified line or PFC bus.
  • Half the primary current of a half-bridge at the same power — lower conduction loss, so it scales to kilowatts.
  • Best transformer utilisation. A simple two-winding transformer, driven both ways (no reset winding, no centre-tap copper waste), fed twice per period into a small filter.
  • ZVS-capable. Driven phase-shifted, the same power stage soft-switches using Ls’s energy — high frequency and high efficiency together.
  • Isolation, multiple outputs, and a clean low-ripple output, like any transformer-isolated converter.

Disadvantages

  • Four switches and two floating gate drives — the highest parts count and drive complexity in the family; overkill below a few hundred watts.
  • Duty-cycle loss from Ls shaves the effective duty, forcing a slightly larger turns ratio; leakage energy must be managed.
  • No built-in flux balancing in the basic circuit (no series blocking capacitor) — volt-second imbalance can walk the core, so practical designs rely on current-mode control or add a blocking cap.
  • Shoot-through risk per leg demands reliable dead-time generation for two legs, not one.
  • ZVS is load-dependent in phase-shifted designs — it fades at light load and the controller must compensate.

Applications

  • High-power off-line supplies, ~500 W to tens of kW — the topology of choice once a half-bridge’s doubled current stops making sense.
  • Server and telecom rectifiers — e.g. a 54 V / 3 kW phase-shifted stage at 140 kHz behind a PFC front end.
  • EV on-board chargers and DC fast-charging modules, where the PSFB (and its resonant cousins) dominate.
  • Industrial equipment: welding supplies, plating, induction heating, high-power battery chargers.
  • Renewable-energy converters — isolated DC-DC stages in solar and storage systems.
  • The front half of many inverters, boosting and isolating a battery bus before the DC-AC stage.

The rule of thumb completes the family: flyback/forward below ~100 W, push-pull for low-voltage inputs to ~1 kW, half-bridge off-line to ~500 W, and full-bridge for everything above — hard-switched when simple, phase-shifted with ZVS when efficiency and frequency matter.

Frequently Asked Questions – FAQs

A full-bridge converter is an isolated DC-DC converter that uses four switches arranged in two legs across the input. The transformer primary hangs between the two leg midpoints, and diagonal pairs of switches fire alternately, driving the primary with the full input voltage in both directions — plus Vin, then minus Vin. The secondary is rectified (often by a four-diode bridge) into an L-C filter, giving buck-like regulation through a transformer. Each switch blocks only Vin, the primary current is half a half-bridge's, and the topology handles the highest power in the isolated family — from about 500 W to many kilowatts.

In continuous conduction the output voltage is Vo = 2 · n · D · Vin, where n = Ns/Np is the turns ratio and D is the duty of each diagonal pair (D less than 0.5). Equivalently Vo = n · Deff · Vin, where Deff = 2D is the fraction of the period the secondary carries voltage. This is twice a half-bridge's output at the same turns ratio, because the primary is driven at the full Vin rather than Vin/2. In a phase-shifted design, Deff is set by the phase between the legs and is slightly reduced by duty-cycle loss from the series inductor.

The four switches form two legs: QA over QB (node A) and QC over QD (node B). Turning on QA and QD together connects node A to the positive rail and node B to the negative rail, so the primary sees plus Vin. Turning on QB and QC does the opposite, applying minus Vin. The two diagonal pairs fire alternately, 180 degrees apart, which drives the transformer with a symmetric AC square wave. The two switches within one leg must never conduct together — that would short the input — so each leg has a guaranteed dead time, which limits each pair to less than 50 percent duty.

Ls is the transformer's leakage inductance plus, usually, a small deliberate series inductor. When the bridge reverses polarity, the primary current must also reverse — but current through an inductor cannot jump, so for a short interval the whole input voltage sits across Ls while its current ramps through zero, and the transformer sees zero volts. During that interval the secondary delivers nothing, so each secondary pulse is narrower than the bridge pulse that commanded it. That difference is duty-cycle loss, and it grows with larger Ls or heavier load. Designers keep Ls anyway because its stored energy is what discharges the MOSFET capacitances and enables zero-voltage switching.

In a phase-shifted full-bridge, both legs run at a fixed near-50-percent duty and the controller regulates by sliding the phase of one leg against the other. The overlap when one node is high while the other is low determines how long the primary sees voltage — the phase shift acts as the effective duty cycle. The benefit is zero-voltage switching: during each dead time, the current maintained by the series inductor Ls discharges the turning-on MOSFET's output capacitance, pulling its voltage to zero before the gate rises, so the switch turns on with essentially no switching loss. This lets multi-kilowatt converters run at high frequency with small magnetics. ZVS weakens at light load, when Ls carries too little energy.

A half-bridge uses one switch leg plus two divider capacitors, so its primary is driven at only Vin/2 — which doubles the primary current for a given power. A full-bridge replaces the capacitors with a second switch leg, driving the primary at the full Vin, so the primary current halves back down. Both stress each switch to only Vin. The half-bridge wins on parts count (two switches, one floating drive) and suits roughly 100 to 500 W; the full-bridge pays four switches and two floating drives to win on conduction loss, taking over above roughly 500 W. For the same output, the full-bridge also needs only half the turns ratio.

Three full-wave options work. A four-diode bridge rectifier puts two devices in series in the current path but stresses each to only Vo/Deff, making it the choice for higher output voltages. A centre-tapped rectifier has just one diode drop in the path — best for low-voltage, high-current outputs — but each device blocks twice as much (2·Vo/Deff) and the winding needs a centre tap. A current-doubler uses two diodes and two output inductors, each inductor carrying half the load current, which suits very high current outputs. In high-current designs, all three are usually implemented with synchronous-rectifier MOSFETs instead of diodes to recover the diode drop.

Because the two switches of each leg sit in series directly across the input. When the top switch of a leg conducts, it clamps the midpoint node to the positive rail, so the bottom switch sees exactly Vin — and vice versa. No reflected transformer voltage can add on, because the node is pinned to a rail by the conducting switch. This is the same mechanism as the half-bridge, and it is what lets both bridge topologies run from high rectified-line buses with ordinary-voltage MOSFETs, unlike the push-pull, whose idle switch sees 2·Vin from the reflected half of its centre-tapped primary.

It can. The basic full-bridge drives the primary directly, with no series DC-blocking capacitor, so if the two half-cycles apply slightly unequal volt-seconds — from mismatched switch timing or on-state drops — a small DC flux increment accumulates each period and the core can drift toward saturation, like an uncorrected push-pull. Practical designs prevent this either with peak-current-mode control, which senses the asymmetry as a current difference and trims each pulse, or by adding a small series blocking capacitor that stops any DC. This differs from the half-bridge, whose coupling capacitor provides that balancing automatically.

The full-bridge takes over where the half-bridge runs out of headroom — from roughly 500 W up to tens of kilowatts. Typical applications include server and telecom rectifiers (for example a 54 V, 3 kW phase-shifted stage switching at 140 kHz), EV on-board chargers and fast-charging modules, welding and induction-heating supplies, high-power battery chargers, and isolated DC-DC stages in solar and energy-storage systems. Hard-switched drive is fine at modest frequency; above that, the phase-shifted variant with zero-voltage switching is the standard, keeping efficiency high while shrinking the magnetics.

Related Topics