The Bridge Diode: Full-Wave Rectification Without the Center Tap
A bridge diode (commonly called a bridge rectifier) is a four-diode semiconductor package that converts alternating current (AC) into direct current (DC) using full-wave rectification. Unlike a two-diode full-wave setup, it does not require a center-tapped transformer, making it the standard choice for everything from 5V USB wall warts to 48V industrial motor drives.
If you are standing at the parts counter or browsing Mouser and need a safe default, here are the workhorses you should stock:
- KBPC5010: 50A, 1000V. Metal case, spade terminals. Ideal for high-current bench supplies. (~$3.00)
- GBU808: 8A, 800V. Inline SIP package, PCB mount. Perfect for embedded power rails. (~$0.80)
- W10M: 10A, 1000V. Inline SIP, compact. Great for general-purpose 12V/24V LED drivers. (~$0.60)
Pinout, Symbol, and Internal Routing
The schematic symbol for a bridge diode looks like a diamond with four diodes pointing inward toward the positive output and outward from the negative output. On the physical package, you will find four pins or spade terminals.
For the ubiquitous square KBPC metal-can package and the flat inline SIP packages (like the GBU series), the pinout is standardized:
- AC Inputs (~, ~): Two pins marked with a sine wave symbol or "AC". Polarity does not matter here; you can swap the transformer wires without affecting the DC output.
- Positive Output (+): The DC high-side. Internally, this connects to the cathodes of the two upper diodes.
- Negative Output (-): The DC ground/return. Internally, this connects to the anodes of the two lower diodes.
Inline SIP packages (4 pins in a row) typically follow the sequence: AC (+) AC (-) or (+) AC (-) AC. Always verify with the silkscreen on the specific manufacturer's datasheet, as Vishay, Diodes Inc., and ON Semi occasionally swap the physical pin order on inline packages.
Operating Regions and Thermal Limits
To bias and select a bridge diode correctly, you must understand its three operating regions. The primary selection criteria are Peak Inverse Voltage (PIV) and Forward Current ($I_F$).
| Operating Region | Bias Condition | Typical Voltage / Current | Design Implication |
|---|---|---|---|
| Forward Conduction | AC input forward-biases two diodes in series | $V_F$ = 1.0V to 1.2V per diode (2.0V - 2.4V total drop). $I_F$ up to rated max. | Causes power dissipation ($P = V_F \times I$). Dictates your heatsink size. |
| Reverse Blocking | AC input reverse-biases the other two diodes | $V_R$ up to PIV rating. Leakage current ($I_R$) in microamps ($\mu A$). | PIV must exceed the peak AC voltage ($V_{RMS} \times 1.414$) plus a 20% safety margin. |
| Avalanche / Breakdown | Reverse voltage exceeds PIV rating | $V_R$ > PIV. Current spikes to amps (destructive thermal runaway). | Catastrophic failure. The bridge shorts AC directly to the DC bus. |
For a 12VAC transformer secondary, the peak voltage is $12 \times 1.414 = 16.9V$. A 50V PIV bridge would technically work, but standard practice dictates using a 400V or 1000V (the "10" in KBPC5010) bridge because the price difference is pennies and the higher voltage silicon is generally more robust against transient spikes.
Bench Scenario: Why a 35A Bridge Rectifier Melted at 10 Amps
Early in my career, I built a 12V 10A power supply for a halogen lighting rig. I selected a KBPC3510 (35A, 1000V) bridge rectifier. The datasheet said "35 Amps," so I assumed it could handle a 10A load easily. I bolted it to the chassis, wired it up, and turned it on. Within three minutes, the solder on the AC spade terminals melted, the bridge cracked, and the 15A mains breaker tripped.
The Setup: 12VAC transformer feeding a KBPC3510, driving a 10A resistive load. No external heatsink, just the bridge mounted flat against a painted steel enclosure.
The Numbers: During forward conduction, two diodes inside the bridge conduct simultaneously. The typical forward voltage drop ($V_F$) is 1.1V per diode, meaning a 2.2V total drop across the bridge.
Power dissipated = $V_{drop} \times Current = 2.2V \times 10A = 22W$.
The junction-to-ambient thermal resistance ($R_{\theta JA}$) of a KBPC package without a proper heatsink is roughly $30^\circ C/W$.
The Outcome: Temperature rise = $22W \times 30^\circ C/W = 660^\circ C$. Since the maximum allowable junction temperature ($T_J$) for silicon is $150^\circ C$, the internal silicon literally cooked itself. The solder on the internal pigtails liquefied, causing an internal short between the AC input and the DC output, which dead-shorted the transformer secondary and tripped the primary breaker.
What Went Wrong: I confused peak surge current with continuous DC current. A 35A bridge can only deliver 35A if it is mounted to an infinite heatsink at $25^\circ C$ ambient. In the real world, you must derate heavily. For a 10A continuous load, you either need a massive finned heatsink with thermal paste, or you need to use active cooling. I replaced it with a KBPC5010 bolted to a dedicated aluminum finned heatsink, and it ran at a manageable $65^\circ C$.
Application Circuit: Building a Robust 12V DC Supply
Here is a complete, battle-tested application circuit for a 12V 5A DC power supply. This topology includes the necessary protection and filtering components that beginners often omit.
- Step-Down Transformer: 120VAC primary to 12VAC secondary, rated for at least 6A (72VA). Never use a transformer rated for exactly your load current; give it 20% headroom.
- Primary Fusing: 1A slow-blow fuse on the 120VAC primary side to protect against transformer shorts.
- Secondary Fusing: 6A fast-acting fuse on the 12VAC secondary side, placed before the bridge diode.
- Rectification: GBU808 (8A, 800V) inline bridge diode. Mount to a small U-channel heatsink if drawing continuous 5A.
- Bulk Filtering: 22,000$\mu F$ 25V electrolytic capacitor across the DC + and - outputs. Formula used: $C = I / (2 \times f \times V_{ripple})$. For 5A, 60Hz, and 2V max ripple, $C = 5 / (120 \times 2) = 0.0208F$.
- High-Frequency Bypass: 0.1$\mu F$ 50V ceramic capacitor in parallel with the bulk cap to shunt high-frequency switching noise.
- Bleeder Resistor: 10k$\Omega$ 1/2W resistor across the DC output. This safely discharges the 22,000$\mu F$ capacitor in about 5 seconds after power-off, preventing a nasty shock if you probe the board.
For a deeper theoretical breakdown of how the ripple voltage interacts with the filter capacitor, the All About Circuits textbook chapter on full-wave rectifiers provides excellent waveform visualizations.
Failure Modes and Multimeter Diode Testing
Bridge diodes fail in two primary ways: open circuit (one or more internal diodes blow, resulting in half-wave rectification and massive 120Hz ripple) or short circuit (thermal runaway melts internal junctions, shorting AC to DC and blowing your upstream fuses).
You do not need an oscilloscope to diagnose a dead bridge. A standard digital multimeter (DMM) in diode-test mode is all you need. According to Fluke's official testing guidelines, the diode test mode applies a small current and measures the voltage drop across the semiconductor junction.
Remove the bridge from the circuit (or at least disconnect the transformer and load) to prevent parallel paths from skewing your readings. Set your DMM to the diode symbol (usually shares a button with continuity).
The Testing Matrix (Red probe = Positive, Black probe = Negative):
- Red on (+), Black on AC1: Should read 0.4V to 0.7V. (Forward biasing the upper diode).
- Black on (+), Red on AC1: Should read "OL" or "1" (Overlimit). (Reverse blocking).
- Red on AC1, Black on (-): Should read 0.4V to 0.7V. (Forward biasing the lower diode).
- Black on AC1, Red on (-): Should read "OL". (Reverse blocking).
- Repeat all four tests for AC2.
By treating the bridge diode not just as a schematic symbol, but as a thermal component with real-world limits, you can design power supplies that survive the bench and the field. Always respect the $V_F$ power dissipation, size your filter capacitors mathematically, and verify your silicon with a DMM before applying mains power.
