Secondary Circuit Design: Isolation, Rectification and PCB Layout
Calling everything after a transformer “the safe low-voltage side” is bad engineering shorthand. A secondary circuit may be isolated from the input, but that does not automatically make every secondary node touch-safe, noise-free or allowed to share ground with the chassis. Its classification depends on the transformer, insulation system, product standard, voltage, environment and intended connections.
This guide focuses on the secondary circuit of an isolated AC/DC or DC/DC power supply—the winding, rectification, filtering, regulation and output network after the isolation transformer. In industrial control, “secondary circuit” can also mean an auxiliary metering or relay circuit; that is a different search intent and should not be mixed into the same PCB rules.
Direct Answer: What Is a Secondary Circuit?
A secondary circuit is the circuitry powered from a transformer's secondary winding. In an isolated switch-mode supply, it usually includes the output rectifier or synchronous MOSFETs, inductor and capacitors, sensing and feedback components, protection and output connector. Its PCB layout must preserve the isolation barrier while keeping high-current switching loops short and controlled.
“Secondary” describes position relative to the transformer. It does not, by itself, define whether the circuit is SELV, grounded, floating, reinforced-isolated or safe to probe with an earth-referenced oscilloscope.
Why Secondary-Side Boards Fail After the Schematic Works
The common community question is: “Why does my flyback work at light load but ring, overheat or lose regulation at full load?” The schematic may be correct. The physical secondary current path is not.
The direct losses are familiar: hot rectifiers, cracked capacitors, EMI test failure, output ripple, cross-regulation problems, optocoupler instability and repeated board spins. The difficult parts are less obvious:
- the output current is pulsed, so a trace that looks electrically short can still have harmful inductance;
- “secondary ground” may be a switching return, a quiet sense reference or a protective-earth connection—those are not interchangeable;
- creepage and clearance requirements depend on the applicable safety standard and use conditions, not a universal internet number;
- the isolation boundary can be accidentally reduced by copper pours, mounting hardware, test pads, silkscreen assumptions or contamination paths;
- probing across an isolation barrier with ordinary grounded equipment can create a hazardous short.
The real decision barrier is knowing whether the problem belongs to topology, component rating, magnetic design, PCB geometry, control-loop compensation or the product's insulation system. A layout checklist helps, but safety-critical values still need an engineer working from the correct product standard.
Map the Secondary Circuit Before You Route It
Start by drawing the energy path, not by filling copper. For a typical isolated flyback supply, the path is:
secondary winding → rectifier → output capacitor → load → secondary return → winding
When the primary switch turns off, secondary current changes quickly. The loop containing the winding, rectifier and high-frequency output capacitor is therefore the critical secondary switching loop. Place these parts close together and keep the outgoing and return conductors adjacent. A wide loop produces more magnetic field, ringing and EMI even when every trace passes a DC-current calculator.
For forward, push-pull, half-bridge or full-bridge converters, the exact conducting path changes, but the method does not: identify each high-di/dt state and minimize its loop area. Do not copy a flyback placement pattern into another topology without tracing current during each switch interval.
Choose Rectification by Loss, Timing and Risk
A diode rectifier is simple and robust, but its forward drop can dominate loss at low output voltage and high current. A Schottky diode may reduce that loss, provided reverse-voltage rating, leakage at temperature and transient margin are adequate.
Synchronous rectification replaces the diode conduction path with a controlled MOSFET. It can improve efficiency, but adds timing, gate-drive and fault-mode questions. Analog Devices describes secondary-side synchronous rectification as a route to higher efficiency in isolated supplies; it is not a free upgrade. False turn-on, reverse current or excessive dead time can erase the gain.
| Option | Usually suits | Main benefit | Verify before release |
|---|---|---|---|
| Standard diode | Moderate current or higher output voltage | Simple control | Forward loss and recovery behavior |
| Schottky diode | Low-voltage outputs | Lower forward drop | Leakage and reverse-voltage margin at temperature |
| Synchronous MOSFET | High-current, low-voltage rails | Higher potential efficiency | Timing, body-diode interval, reverse current and driver supply |
Rate the component for actual waveform stress, not only nominal output voltage and current. Include transformer leakage, ringing, startup, short circuit and load release. If a snubber or clamp is required, position it at the device and current loop it is intended to control.
Output Filtering: Keep Power Return and Sense Return Honest
The high-frequency output capacitor belongs close to the rectifier and winding return. Bulk capacitance may sit nearer the output connector, but long shared copper between the power capacitor and feedback divider can make load current appear as regulation error.
Take the feedback sense point from the intended regulated output node and provide a quiet return to the controller or reference. Do not let rectifier current share a narrow section of that sense return. If remote sensing is used, protect and route the pair as a measurement path, not as spare copper attached wherever convenient.
Multiple secondary windings add another trap: cross-regulation. Regulating one rail does not guarantee every auxiliary rail remains within limits at all load combinations. Winding coupling, diode drops, load balance and post-regulation matter. Test no-load, minimum-load, asymmetric-load and transient cases rather than checking only the nominal operating point.
Feedback Must Cross the Barrier Without Defeating It
An optocoupler and secondary reference remain a common feedback arrangement. Other designs use an isolated amplifier, transformer-coupled signal or primary-side regulation. Analog Devices' isolated flyback reference design shows feedback crossing the barrier through an isolated error-amplifier path, while its secondary-side controller examples place sensing and control functions on the isolated output side.
Whichever method is used, the feedback component must have suitable insulation ratings for the required isolation system. Its pads, copper, solder mask and nearby routing must not reduce the designed creepage or clearance. Keep the quiet reference and divider away from the rectifier switch node, transformer fringe field and synchronous gate loop.
Primary-side regulation can remove the optocoupler, but it estimates output conditions through the transformer waveform. Its accuracy and transient behavior depend on topology, winding coupling, load range and controller method. Choose it because the performance fits—not because one component disappears from the BOM.
Preserve the Primary-to-Secondary Isolation Boundary
The isolation corridor is a system feature, not empty artwork. Texas Instruments recommends applying creepage and clearance from the specific application standard and notes that PCB slots or ribs may help increase the surface path. The shortest path can occur at an isolator footprint, transformer pin, connector, screw, heat sink, board edge or hidden inner-layer copper.
During PCB design and fabrication review, check every conductive layer and mechanical item:
- Copper: Remove unintended pours, tracks and via annular rings from the barrier on all layers.
- Components: Verify transformer, optocoupler and isolated-device package ratings against the required insulation level.
- Slots: Define routed slots in the mechanical data and confirm their finished width, location and fabrication tolerance.
- Hardware: Include mounting screws, heat sinks, shields and chassis metal in the shortest-path check.
- Contamination: Account for pollution degree, coating, altitude and material group as required by the governing standard.
- Test features: Do not drop convenient probe pads into a certified clearance corridor.
Do not publish a single creepage value as if it fits every product. Working voltage, transient category, insulation type, altitude and standard can change the requirement. CAM can verify the geometry against a stated rule; it cannot invent the safety classification.
Grounding: “Secondary Ground” Is Not Automatically Earth
Labeling the output return GND often causes accidental connections. Use names that show function: PRI_RETURN, SEC_RETURN, CHASSIS or PE, for example. Then document whether the secondary is floating, intentionally bonded to chassis, connected through an EMI capacitor, or referenced elsewhere in the system.
An isolation-rated Y capacitor may provide a controlled common-mode noise path across the barrier, but its type, value, location and leakage-current implications require safety and EMC review. A generic ceramic capacitor is not a substitute. Likewise, an oscilloscope earth clip can bond a floating secondary to protective earth; confirm the measurement setup before probing.
Thermal and Mechanical Details on the Secondary Side
Low voltage does not mean low heat. Secondary rectifiers, synchronous MOSFETs, inductors and capacitors can carry the highest current in the supply. Provide copper spreading without expanding the switching loop. Keep electrolytic capacitors away from rectifier and transformer hot spots, and check ripple-current rating at the actual internal temperature.
If a heat sink bridges near the barrier, include it in clearance analysis and confirm the semiconductor tab potential. Our practical heat sink selection guide explains why interface, airflow and mounting must be evaluated as one thermal path.
Secondary Circuit Validation Plan
Validate the assembled product with appropriate isolated instruments and qualified procedures. At minimum, test:
- startup at minimum and maximum input;
- no-load, light-load and full-load regulation;
- every relevant multiple-output load combination;
- ripple and ringing with a short probe loop;
- load-step overshoot and recovery;
- short-circuit, overload and recovery behavior;
- rectifier, MOSFET, transformer, inductor and capacitor temperatures;
- EMI and required dielectric/insulation tests under the applicable standard.
Save This Secondary Circuit Release Checklist
- The exact topology and every high-di/dt secondary loop are marked.
- Rectifier voltage, current, thermal and transient margins are verified.
- High-frequency output capacitance is inside the critical current loop.
- Power return and quiet feedback return do not share harmful impedance.
- Multiple outputs are tested under asymmetric loading.
- The feedback component has the required isolation rating.
- Creepage and clearance come from the correct product standard and conditions.
- All layers, slots, vias, test pads and mechanical hardware are checked.
- Primary, secondary, chassis and protective-earth nets use unambiguous names.
- Thermal testing includes rectifiers, magnetics and capacitor hot spots.
- Measurement equipment cannot unintentionally bridge the barrier.
- Gerber/ODB++, drill, stack-up, BOM and assembly drawing agree.
FAQ
Is a Secondary Circuit Always Low Voltage?
No. A transformer's secondary can be lower, equal to or higher than its primary voltage. “Secondary” identifies the winding or circuit receiving transferred energy; it does not guarantee extra-low voltage or touch safety.
Can Primary and Secondary Grounds Be Connected?
Only when the system architecture and safety design intentionally allow it. A direct connection defeats galvanic isolation. Controlled chassis bonds or isolation-rated EMI components may be used, but they must be designed and reviewed for the application.
What Components Are Normally on the Secondary Side?
Typically the secondary winding connections, rectifier or synchronous MOSFETs, output inductor and capacitors, feedback sensing, protection and output connector. The exact set depends on converter topology and control method.
Where Should the Output Capacitor Be Placed?
Place the high-frequency capacitor close to the rectifier and secondary return so the pulsed loop is small. Additional bulk capacitance may be placed for load support, but feedback sensing must avoid shared high-current voltage drop.
Does Solder Mask Increase Creepage or Clearance?
Do not assume ordinary solder mask automatically grants the required insulation credit. Whether coating affects the required distance depends on the standard, material, process control and qualification. Use the applicable safety rules.
When Should the PCB Manufacturer Review the Design?
Before release when the board contains an isolation slot, tight barrier geometry, high-current secondary copper, via-in-pad thermal structures or mechanical parts near the barrier. Submit the complete data package for PCB manufacturing and assembly review rather than sending Gerbers alone.
Final Takeaway
A reliable secondary circuit is not simply “the components after the transformer.” It is a pulsed power path, a feedback reference, an insulation boundary and a thermal assembly at the same time. Mark the current loops, preserve the barrier on every layer, separate power and sense returns, and validate the final enclosure under real load. Save the checklist and ask for a CAM/DFM review before committing a safety-critical isolated supply to production.
