A PCB coil (planar inductor) replaces traditional wire-wound magnetics with etched copper traces on the printed circuit board. You should use a planar coil when Z-height is strictly constrained (under 2mm), when manufacturing at high volumes to eliminate discrete component placement, or for high-frequency applications like 13.56 MHz RFID and Qi wireless charging. You must avoid it when you need high inductance (above 10 µH) or high DC current (above 5A) without resorting to massive, expensive copper pours.

The core challenge of designing a pcb coil is balancing trace resistance, skin effect at high frequencies, and spatial constraints. Unlike a discrete inductor where the manufacturer handles the magnetics, a planar coil forces you to manage the physics of the copper directly.

Designing the PCB Coil: Trace Width, Copper Weight, and Current Limits

The most common failure in planar inductor design is underestimating the DC resistance (DCR) of the traces, leading to excessive voltage drop and thermal runaway. To determine what trace width a specific current needs, you must reference the IPC-2152 standard, which updated the older IPC-2221 charts with more accurate thermal modeling for external and internal layers.

Because PCB coils are almost always routed on external layers to maximize cooling and allow for tighter spacing, the table below uses external layer IPC-2152 data for a standard 10°C temperature rise above ambient.

External Trace Width vs. Current Capacity (10°C Rise, IPC-2152)
Target Current (A) 1 oz Cu (1.37 mil) Width 2 oz Cu (2.74 mil) Width 3 oz Cu (4.11 mil) Width Est. DCR (mΩ/in, 1 oz)
1.0 A 11 mils 6 mils 4 mils 0.94 mΩ
2.0 A 25 mils 13 mils 9 mils 0.41 mΩ
3.0 A 42 mils 22 mils 15 mils 0.24 mΩ
5.0 A 85 mils 45 mils 31 mils 0.12 mΩ

Note: Data derived from standard external layer calculations. For precise modeling, use the Saturn PCB Toolkit.

The High-Frequency Skin Effect Trap

If your PCB coil operates at RF frequencies (e.g., 13.56 MHz for NFC/RFID or 6.78 MHz for A4WP wireless power), DC current capacity is only half the story. At high frequencies, current travels only on the surface of the conductor due to the skin effect.

At 13.56 MHz, the skin depth in copper is approximately 0.7 mils. A standard 1 oz copper layer is 1.37 mils thick, meaning the bottom half of your trace is essentially dead weight. Upgrading to 2 oz or 3 oz copper provides almost zero AC resistance benefit at these frequencies. Instead of thicker copper, you must use wider traces or route the coil across multiple layers, stitching them together with vias to increase the effective surface area.

Breadboard to PCB: Migrating Inductive Loads Without the Magic Smoke

When prototyping, you typically use a discrete wire-wound inductor (like a Coilcraft or Murata part) on a breadboard or dev kit. Migrating from that discrete part to a custom etched pcb coil introduces severe parasitic shifts. The mistakes that survive this transition usually stem from assuming the planar coil behaves identically to its wire-wound predecessor.

Which Mistakes Survive the Migration?

1. The Ground Plane Trap (The Fatal Error): On a breadboard, the inductor sits in free air. When moving to a 4-layer PCB, beginners often pour a solid ground plane on Layer 2 directly beneath the Layer 1 coil. This creates a shorted secondary winding in a transformer. The alternating magnetic field induces eddy currents in the ground plane, which generate an opposing magnetic field. This can drop your inductance by 50% or more and turn your ground plane into a localized heater. Fix: Hollow out all copper pours (ground, power, and signal) on all layers directly beneath and slightly beyond the coil footprint.

2. Decoupling Resonance Shifts: Wire-wound inductors have a specific Equivalent Series Resistance (ESR) and Equivalent Parallel Capacitance (EPC). A planar PCB coil has significantly lower ESR and different EPC due to the wide, flat trace geometry interacting with the FR-4 dielectric. The 100nF decoupling capacitor that perfectly damped ringing on your breadboard might now form an underdamped LC tank with your new coil, causing voltage spikes that destroy your driver IC.

Breadboard-to-PCB Migration Checklist

  1. Clear the Keep-Out: Ensure a copper-free keep-out zone extends at least 2mm beyond the outer edge of the coil on all layers.
  2. Recalculate the Matching Network: Do not copy-paste the breadboard's capacitor values. Use an LCR meter to measure the exact inductance and Q-factor of the etched coil, then recalculate your series/parallel matching caps.
  3. Verify Trace Feeds: Ensure the traces routing power to the coil are sized for the peak current (refer to the IPC-2152 table above), not just the coil traces themselves.
  4. Simulate Parasitics: Run a 2D field solver (like Saturn PCB Toolkit or your EDA suite's built-in solver) to extract the parasitic capacitance to the reference planes.

First-Spin Board Verification and Workshop Safety

When your first-spin PCBs arrive, do not just populate the board and apply power. Planar coils are highly sensitive to manufacturing tolerances; a 10% variation in etch width from your fab house will shift your inductance and alter the resonant frequency.

How to Test a First-Spin Board

Step 1: DMM Continuity and Isolation. Before soldering a single component, use a digital multimeter to check the DC resistance of the raw coil. It should match your calculated DCR (usually between 0.1Ω and 2Ω). Verify infinite resistance between the coil pads and any adjacent ground vias to ensure no etch defects caused a short.

Step 2: LCR Meter Profiling. Measure the inductance using an LCR meter. Crucially, you must measure at your operating frequency, not the default 1 kHz. A coil that reads 2.5 µH at 1 kHz might read 2.1 µH at 6.78 MHz due to parasitic capacitance. Measure the Q-factor simultaneously; a Q below 15 at RF indicates excessive trace resistance or severe eddy current losses in nearby copper.

Step 3: Thermal Profiling Under Load. Once populated, drive the coil at 50% of its target power. Use a thermal camera (like a FLIR One) or a K-type thermocouple taped to the board with Kapton tape. If the trace temperature exceeds 60°C above ambient, your copper weight is insufficient, or your skin effect calculations were overly optimistic.

Workshop Safety: Soldering and Fume Extraction

Assembling the driver ICs and high-Q NP0/C0G matching capacitors around a PCB coil requires precise thermal management. If you are using SAC305 lead-free solder (melting point 217°C), set your soldering station to 350°C and use a chisel tip to maximize thermal transfer to the high-mass copper pads. If you prefer Sn63/Pb37 eutectic solder (melting point 183°C), a tip temperature of 320°C is sufficient.

⚠️ Workshop Safety Note: Hand-soldering with rosin-core flux (RMA) releases colophony fumes, which are a known respiratory sensitizer and can trigger occupational asthma. You must use a localized fume extractor (such as a Hakko FA-400 or a BOFA system with a HEPA/carbon filter) positioned within 2 inches of the soldering iron tip. Never rely on passive room ventilation when assembling high-density RF boards that require extended dwell times on large copper pads.

By respecting the physics of trace geometry, eliminating ground plane eddy currents, and rigorously testing at the target frequency, your custom pcb coil will perform just as reliably as a discrete component, while keeping your BOM cost and Z-height to an absolute minimum. For deeper reading on magnetics layout, refer to the IPC-2152 standard documentation for advanced thermal derating curves.