USB-C Power Delivery Hardware Design (Sink, Source, DRP)

USB-C power is deceptively simple: two pins (VBUS/GND) and you are done. In practice, reliable designs live or die on three things: CC correctness, inrush / bulk capacitance, and a power path that survives real cables and real ports.

Start here: what are you building?

Before picking a connector footprint, decide what you want from the port. The hardware blocks (and the failure modes) are very different between a basic 5V sink and a dual-role PD port.

CC pins in 2 minutes (the 80/20)

USB-C uses CC1/CC2 (Configuration Channel) to detect attach, orientation, and role. USB Power Delivery (USB PD) also uses the CC line for communication. If CC is wrong, everything else becomes random: ports that do not turn on, charging in only one orientation, or disconnects under load.

• Sink (power consumer): presents Rd on CC.
• Source (power provider): presents Rp on CC and advertises the available current at 5V.
• Orientation: only one of CC1/CC2 is connected through the cable; the other CC pin may be used as VCONN to power e-marked cables.

CC resistor values (Rp/Rd)

If you implement Type-C attach with discrete resistors (no PD messaging), these are the nominal values you will see in many designs. Always verify against the USB Type-C specification and your controller datasheet (tolerances and details matter).

When you need USB PD (vs CC resistors only)

A "Type-C only" port (no PD messages) can still be valid for a 5V sink, as long as CC is correct and you respect the current advertisement. USB PD becomes necessary when you need higher voltages, more power, or source/dual-role behavior.

You need USB PD when:

• You want higher voltage than 5V (for example 9V/15V/20V), or programmable voltage via PPS.
• You want more than 3A on VBUS (requires USB PD and an e-marked 5A cable).
• You are a source (your product powers other devices) or a dual-role port (DRP).
• You want predictable behavior across laptops, docks, chargers, and power banks.

Even with PD, the attach/boot sequence still starts at 5V. Do not assume a higher voltage is present until your controller confirms the contract is active. A lot of "mystery resets" are simply loads turning on too early.

PD negotiation & controller choices

USB-C Power Delivery is not "USB data". PD messages travel on the CC wire (not on D+/D-), so power-only devices can be fully valid without implementing any USB data stack.

PD in 90 seconds

• The source advertises what it can provide (its power data objects / capabilities).
• The sink requests one option (and optionally asks for PPS when supported).
• The source accepts and transitions VBUS to the new voltage.
• Once the contract is active, the sink can enable heavier loads.

Terminology cheat sheet

• Sink / Source: power consumer / power provider (VBUS direction).
• UFP / DFP: USB data device / USB data host.
• DRP: Dual-Role Port (can be sink or source, depending on negotiation).
• VCONN source: the side that powers the cable electronics via the unused CC pin.

Controller categories you will encounter

• Standalone PD sink/source controllers: the policy engine is inside the IC. You configure profiles through straps, NVM, or a simple I2C/SPI register map.
• TCPC-style controllers: the IC handles the physical layer and Type-C signaling, but a host MCU (policy engine) decides contracts and behavior. Common in DRP and complex systems.
• Controllers with integrated power path: some devices include a load switch / discharge / current limit, reducing external parts (but increasing dependency on the reference design).

Selection checklist (hardware-first)

• Role: sink vs source vs DRP (and whether you need role swaps).
• Power features: fixed voltages only vs PPS; do you care about EPR (beyond 20V)?
• Power path: external hot-swap/eFuse vs integrated switch; reverse-current needs.
• Interface: standalone vs a firmware-managed controller; fault reporting and logging.
• Compliance effort: how closely can you follow the controller datasheet layout and BOM?

Architecture patterns (schematic-level)

Think of a USB-C power port as a small power subsystem: connector + CC/PD brain + a controlled, protected power path. The next sections break down common patterns.

Pattern A: Sink-only, 5V input (no PD)

• USB-C receptacle
• Correct Rd on CC1 and CC2 (or a Type-C sink controller)
• VBUS protection (TVS + OVP strategy) and a controlled load switch if your input capacitance is high
• DC/DC or LDO rails as needed

This is a good fit for low-power devices that can live within the current the source advertises at 5V. If your design draws a big current spike at plug-in, strict ports may shut down.

Pattern B: PD sink (fixed profiles)

• USB-C receptacle
• PD sink controller (policy engine inside the IC, or TCPC + MCU policy engine)
• Power-path switch / hot-swap / eFuse (sometimes integrated in the PD controller)
• DC/DC sized for your negotiated VBUS range (for example 5V to 20V)

Use this when you want stable power above what CC-only can guarantee, or you want to request specific voltages.

Pattern C: PD sink + battery charging

If you are charging a battery (or feeding a power-hungry load), PD lets you trade cable current for voltage. Typical high-level options:

• Negotiate a higher fixed voltage and buck down into your charger / system bus.
• Use PPS (if supported by the source and your controller) to request a voltage closer to the instantaneous need.

Pattern D: Source-only (you power other devices)

• USB-C receptacle
• PD source controller
• Current-limited high-side switch (port protection) and VBUS discharge
• VCONN capability if you want to support e-marked cables

As a source, you are responsible for over-current protection, safe behavior on shorts, and handling cable events cleanly.

Pattern E: Dual-role port (DRP)

DRP designs combine sink and source behavior. They often need a bidirectional power path and careful reverse-current handling. DRP is common in laptops and power banks and is the fastest way to enter "complex edge-case land".

Power path strategies

After negotiation, decide what to do with VBUS: pass-through (no conversion), conversion (buck/buck-boost), or a hybrid approach using PPS when available.

1) "No conversion" (pass-through)

In a pass-through design, you protect/switch VBUS, but you do not force it into a fixed internal bus. Your load sees the negotiated voltage directly (or through an ideal-diode / hot-swap style element).

• Pros: very efficient, fewer power components, less heat.
• Cons: everything downstream must tolerate the negotiated VBUS range; transient protection becomes critical.

2) Conversion (buck / buck-boost)

Treat USB-C as an input source and generate a fixed system bus (for example 5V, 12V, or a battery-charger input) using a DC/DC converter.

• Pros: stable internal rails, easier system integration, predictable behavior under source changes.
• Cons: conversion losses and heat; more BOM; layout becomes more sensitive.

3) Hybrid: negotiate to reduce conversion stress (PPS)

If your controller and the source support PPS, you can request a voltage that keeps the downstream DC/DC in a comfortable region (less current, lower losses) without over-volting the rest of your design.

Protection checklist

USB-C connects to the outside world, so plan for ESD events, hot-plug transients, and users doing strange things. Robustness is mostly protection + layout discipline.

• ESD protection for the pins you expose (VBUS, CC1/CC2, D+/D-, and high-speed pairs if present).
• VBUS surge/overshoot control (TVS and/or OVP strategy matched to your max negotiated voltage).
• Controlled attach / inrush limiting if you have meaningful capacitance behind VBUS.
• Current limiting and short-circuit behavior if you are a source (and often a good idea on sinks too).
• Reverse-current blocking / OR-ing if you have multiple power inputs or DRP behavior.
• VBUS discharge path (many controllers require or integrate this) so VBUS returns to a safe level after detach.

Inrush: the hidden failure mode

The most common "works on my wall charger but not on a laptop" failure is inrush. If your board presents too much capacitance directly on VBUS, the plug-in current spike can trip port protection or brown out the negotiation.

The Type-C ecosystem is designed around strict attach and inrush behavior. Instead of memorizing numbers, follow a robust pattern:

• Keep only small, necessary capacitance directly on the connector side of the switch.
• Place your bulk capacitance behind a controlled load switch / hot-swap / eFuse.
• Use a controlled VBUS slew rate / soft-start so the port never sees a current spike that looks like a short.
• If you need large hold-up energy, add it behind the switch and validate behavior on real hosts.

PD 3.1 EPR (28/36/48V): what changes

EPR extends USB-C beyond 20V. It enables 24V/36V/48V-class systems but raises the bar for component ratings, layout spacing, and transient protection.

• Component ratings: caps, FETs, TVS/OVP, eFuses, and connectors must have explicit voltage margin.
• Layout and spacing: clearance/creepage become more important, especially around the connector and switching elements.
• Cables matter more: higher power requires correct cable capabilities (for example e-marked 5A cables) and an EPR-capable ecosystem.
• Fault cases: unplug transients, shorts, and mis-negotiation are harder on the hardware at higher voltage.

If you are planning for EPR, pick a controller and power-path devices that explicitly support your target range, then validate with EPR-capable sources and cables (not just a random charger).

PCB layout checklist

Place ESD/TVS parts tight to the connector, keep CC short and quiet, and keep your VBUS path low inductance. If you route high-speed pairs, treat them like RF (controlled impedance, clean return paths).

• TVS placement: put ESD/TVS devices as close as physically possible to the connector with a short, wide return to ground (multiple vias).
• CC routing: keep CC traces short, keep them away from noisy switch nodes, and avoid stubs.
• VBUS path inductance: keep the VBUS loop tight; use wide copper and avoid long detours before the protection/switch.
• Sense and ground strategy: if you measure VBUS/current, route sense traces as a Kelvin pair and keep them away from high di/dt loops.
• If you route USB 3.x / Alt-Mode: keep pairs tightly coupled, match lengths within a pair, and keep return paths continuous.

If your design includes high-speed pairs, start with the USB-C Pinout Explorer to validate lane mapping, then use the PCB Stackup & Impedance Calculator and Controlled Impedance Designer to set trace geometry.

Bring-up & debugging

Validate CC first (both orientations), start at 5V, then scale up. Test with a strict laptop port early and instrument VBUS during plug/unplug and load steps.

1. Validate CC first: confirm attach is detected in both orientations.
2. Start at 5V: ensure your board boots or stays safe at default current before you request more.
3. Negotiate gradually: request a conservative PD profile first, then step up as you gain confidence.
4. Test a matrix: at least one strict laptop port, one reputable charger, a power bank, and multiple cables.
5. Instrument: log controller events, and scope VBUS during plug/unplug and load steps.

Common pitfalls (and what they really mean)

References & further reading

• Primary source: USB Implementers Forum (USB-IF) specifications and compliance resources (usb.org).
• USB Type-C Cable and Connector Specification (USB-IF).
• USB Power Delivery Specification (USB-IF).
• Your Type-C/PD controller datasheet + reference schematic + PCB layout guidelines.

FAQ