A dual-source circuit diode OR-ing configuration is the simplest, most robust method for power multiplexing. It allows two independent DC power supplies to share a single load without backfeeding, using the unidirectional conduction of diodes to automatically select the higher-voltage source. If your primary wall adapter fails or drops below the backup battery voltage, the load seamlessly transitions to the backup source with zero switching transients. This guide breaks down the exact topology, component selection, failure modes, and breadboard verification steps for a 12V, 1A system.
The Diode OR-ing Topology: Nodes, Paths, and Component Selection
The passive OR-ing topology relies on two diodes arranged in parallel with their cathodes tied together. To design this properly, we must define our circuit nodes and select components based on forward voltage drop ($V_f$) and thermal dissipation.
Node Definitions
- Node_V1 (Primary Source): Connected to the anode of Diode 1 (D1). Typically a 12V DC wall adapter.
- Node_V2 (Backup Source): Connected to the anode of Diode 2 (D2). Typically a 12V sealed lead-acid (SLA) or LiFePO4 battery.
- Node_OUT (Common Load): The tied cathodes of D1 and D2. This node feeds the positive terminal of your load.
- Node_GND (Common Ground): The shared negative return path for V1, V2, and the load. Never isolate the grounds in a passive OR-ing circuit.
Component Selection: Why Schottky?
For a 12V system drawing up to 1A, do not use standard silicon rectifiers like the 1N4007. A silicon circuit diode has a $V_f$ of roughly 0.7V to 1.0V at 1A. This wastes power and drops your load voltage to 11V. Instead, select a Schottky diode like the 1N5819 (1A, 40V). The 1N5819 exhibits a $V_f$ of approximately 0.35V at 1A, keeping your Node_OUT voltage near 11.65V and cutting power dissipation in half.
Behavior Matrix and Failure Mode Extremes
Understanding how the circuit reacts to source variations and component failures is critical for reliable deployment. The table below maps the behavioral states of the OR-ing multiplexer.
| System Condition | Node_V1 (Primary) | Node_V2 (Backup) | D1 State | D2 State | Node_OUT (Load) |
|---|---|---|---|---|---|
| Normal Operation | 12.0V | 11.5V | Forward Biased (Conducting) | Reverse Biased (Blocking) | ~11.65V (Sourced by V1) |
| Primary Fails/Drops | 10.0V | 11.5V | Reverse Biased (Blocking) | Forward Biased (Conducting) | ~11.15V (Sourced by V2) |
| Both Sources Equal | 12.0V | 12.0V | Conducting (Shared) | Conducting (Shared) | ~11.65V (Current shared) |
| D1 Fails Short | 12.0V | 11.5V | Shorted (0V drop) | Reverse Biased | 12.0V (V1 backfeeds into V2) |
| D1 Fails Open | 12.0V | 11.5V | Open (Infinite resistance) | Forward Biased | ~11.15V (V1 isolated, V2 takes full load) |
Step-by-Step Breadboard Testing Procedure
Before soldering this into a permanent enclosure, validate the OR-ing behavior on a solderless breadboard. Breadboard contact resistance can skew $V_f$ measurements, so follow this exact sequence.
- Prepare the Board: Ensure the breadboard is de-energized. Insert D1 and D2 into the same row on the cathode side (the silver band) to tie them to Node_OUT. Place their anodes in separate, isolated rows for Node_V1 and Node_V2.
- Establish Ground: Run a jumper wire connecting the ground rails of both power supplies and the load to a common Node_GND bus.
- Connect the Load: Attach your load (e.g., a 12Ω, 2W power resistor to simulate a 1A draw at 12V) between Node_OUT and Node_GND.
- Energize V1 Only: Power up Node_V1 to 12.0V. Leave Node_V2 disconnected. Measure the voltage at Node_OUT. You should read approximately 11.6V to 11.7V.
- Measure $V_f$ Directly: Place your multimeter's red probe on the anode of D1 and the black probe on the cathode. A healthy 1N5819 under 1A load will show 0.30V to 0.45V. If you read >0.6V, you likely have a counterfeit part or high breadboard contact resistance.
- Introduce V2: Power up Node_V2 to 11.5V. Measure Node_OUT; it should remain at ~11.65V. Measure the voltage across D2 (anode to cathode). It should read a negative voltage (approx -0.15V), confirming D2 is reverse-biased and blocking V2.
- Simulate Primary Failure: Slowly dial down the voltage of V1 from 12.0V to 10.0V. Watch your multimeter on Node_OUT. As V1 drops below V2's voltage plus the diode drop, D1 will snap into reverse bias, and Node_OUT will seamlessly transition to being powered by V2 (reading ~11.15V).
Why Passive Diode OR-ing Over Active MOSFET Alternatives?
While the passive circuit diode OR-ing topology is cheap and instantaneous, modern designs often use "ideal diode" controllers (like the Analog Devices LTC4357 or TI's TPS2410) which drive N-channel MOSFETs to emulate diode behavior. Here is how they compare for a 1A, 12V application.
| Criteria | Passive Schottky Diode (1N5819) | Active Ideal Diode Controller + MOSFET |
|---|---|---|
| Forward Voltage Drop | ~0.35V (Fixed by semiconductor physics) | ~0.02V (Depends on MOSFET $R_{DS(on)}$) |
| Power Dissipation at 1A | 0.35W (Requires thermal management) | 0.02W (Negligible heat) |
| Bill of Materials (BOM) Cost | ~$0.15 (Two diodes) | ~$3.50 - $5.00 (Controller IC + 2x MOSFETs + passives) |
| Switching Speed | Nanoseconds (Inherent junction physics) | Microseconds (Depends on controller slew rate) |
| Quiescent Current | 0A (Zero leakage when reverse biased) | ~1mA to 5mA (Required to power the controller IC) |
The Verdict: Choose the passive circuit diode topology when your load current is under 2A, your BOM budget is strict, and your system can tolerate a 0.4V drop. As detailed in Electronics Tutorials, the simplicity of the PN junction provides inherent reliability. Choose the active MOSFET alternative only when you are designing high-current systems (>5A) where a 0.4V drop would cause unacceptable $I^2R$ heating, or when powering sensitive 3.3V/5V logic where every millivolt of headroom matters.
Circuit Diode FAQ: Troubleshooting and Design Edge Cases
Why is my circuit diode getting too hot to touch under load?
If your 1N5819 is exceeding 60°C, you are likely exceeding its continuous current rating or operating in a high-ambient environment. At 1A, the diode dissipates 0.35W. If you bend the leads tightly against the plastic body, you eliminate the copper lead's ability to act as a heatsink. Always leave at least 5mm of lead length between the DO-41 body and the PCB/breadboard to allow thermal conduction into the copper traces. If the load draws more than 1.5A, upgrade to a 3A Schottky like the 1N5822 or an SR360.
Can I use a standard 1N4007 silicon circuit diode for 5V logic OR-ing?
Technically yes, but practically no. A 1N4007 will drop roughly 0.7V to 0.9V at typical logic currents. If you are OR-ing two 5V USB supplies to power a Raspberry Pi, the 1N4007 will drop your Node_OUT to 4.1V, which is below the Pi's brownout threshold and will trigger a low-voltage warning. For 5V logic OR-ing, use a Schottky diode (0.3V drop) or, better yet, a dedicated USB power multiplexer IC like the TPS21110 which has an internal $R_{DS(on)}$ of just 70mΩ.
What happens to the circuit diode if the load suddenly shorts to ground?
In a dead short at Node_OUT, the current is limited only by the internal resistance of your power supplies and the diode's bulk resistance. A standard 12V wall adapter might output 3A to 5A into a short before its internal foldback protection trips. The 1N5819 is rated for a non-repetitive peak forward surge current ($I_{FSM}$) of 25A for one 8.3ms half-cycle. It will likely survive the initial short, but if the power supply lacks fast overcurrent protection and sustains 3A continuously, the diode junction will overheat, fail short, and permanently bridge V1 and V2. Always use a fast-blow fuse on the source lines.
How do I measure the exact forward voltage drop on my breadboard?
Do not measure $V_f$ by referencing the power supply's display and subtracting the breadboard Node_OUT voltage; breadboard contact resistance can easily add 0.1V to 0.2V of phantom drop. Instead, use two multimeters or a differential probe. Place the red probe directly on the diode's anode lead (as close to the plastic body as possible) and the black probe directly on the cathode lead. This Kelvin-style measurement isolates the semiconductor junction drop from the parasitic resistance of the breadboard contacts and jumper wires.
