USB-C shell

Should a USB-C shell go straight to ground, or through an RC?

Straight, unless an earthed box or cable shield touches it too. An 8 kV zap puts the same 30 A into the board’s ground either way. Through 1 MΩ beside 4.7 nF, the cap is left holding 249 V. Common 0805 caps are rated 50 V.

The zap, both ways

From 1 to 300 ns

Through the RC
Straight
A 50 V part’s rating

The test gun is 150 pF charged to 8 kV, behind 330 Ω. The RC is 1 MΩ beside 4.7 nF.

Two ways to tie a shell down

USB-C on Raspberry Pi’s CM5 IO board

A USB-C receptacle has a metal shell, soldered down at four tabs. Reach for the cable with a charged finger and the zap lands on that shell. Open boards tie it to ground one of two ways. Straight puts every tab on the ground net. An RC puts the tabs on a net of their own, tied to ground through 1 MΩ beside a 4.7 nF cap.

The circuit

The gun is the one IEC 61000-4-2 sets out for test labs. Each tab with its vias is about 1 nH.

On the board

The CM5 IO board’s USB-C power input, as the field model built it. The shell is seen through. Its four tabs are its only way into ground.

What the cap is left holding

8 kV
4 kV

One contact zap. Every RC has 1 MΩ beside the cap.

USB-C shell

4.7 nF is left holding 249 V

Field marks. The gun’s 150 pF at 8 kV holds 1.2 µC, and the zap delivers 1.17 µC of it. All of that lands in the cap, so the cap ends at that charge over its capacitance.

Look-alike. A smaller cap is worse. 1 nF is left at 1,169 V.

Range. At 4 kV, the usual EU test level, 4.7 nF still gets 124 V. The 1 MΩ drains it below 1 V in 23 ms.

Habitat. Arduino’s Uno R4 fits 4.7 nF rated 1 kV beside its 1 MΩ. Raspberry Pi and MNT Reform tie their shells straight.

Zap it yourself

5 ties, 1 to 8 kV

Tie

Zap

8 kV

USB-C shell

1 MΩ and 4.7 nF, 8 kV zap

Through an RC
Straight
Part ratings

Left on the shell: 249 V
Into ground, at its peak: 29.8 A

Every RC has 1 MΩ beside the cap. The circuit is linear, so each zap is the 8 kV run scaled.

Where the zap crosses the ground

Most charge
Less

An 8 kV zap on the HDMI0 shell, and the charge that crossed the ground in its first 100 ns. The CM5 module sits above the board on two connectors.

USB-C shell

With power on the far edge, 11 A goes through the CM5

Field marks. The zap leaves through whatever cable goes to ground. On the board as built, the power cable is on the zapped edge. The zap runs along that edge to it, and 2.9 A goes through the CM5.

Look-alike. Move the power input to the far edge and the zap runs round the board to it. 11 A goes through the CM5. That’s the current that can reset a board.

Range. Straight or RC, the zap spreads the same way. The field model ties every shell straight, and at 100 MHz 4.7 nF is 0.3 Ω.

Habitat. Raspberry Pi’s CM5 IO board, zapped on its HDMI0 shell. Every cable connector sits on one edge.

Which to use

Shell ties

The bump while the zap is on

Straight, on the 8 kV zap. Each path is a tab with its vias, about 1 nH. The bump is over within a few nanoseconds.

Key to shell ties

How should this shell be tied down?

1aNothing earthed touches itStraight, every tab, 2e
1bAn earthed metal box, or a cable shield earthed at the far end, touches itGo to 2
2aYou can fit a cap rated 1 kV or moreAn RC, 2c
2bOnly a 50 V part fitsStraight, 2c

With an earthed box, the RC keeps mains hum off your ground. Leave footprints for both, so a fix at the test lab is a part swap.

Conditions

Zap: IEC 61000-4-2 contact discharge, from the standard’s own formula. At 8 kV its first peak is 30 A, with a 0.8 ns rise.
Gun: 150 pF behind 330 Ω, run as a current source. That holds while the shell stays far below 8 kV, which it does here.
Straight: 4 paths to ground side by side, each a tab with its vias at about 1 nH
RC: 1 MΩ beside one cap with 0.8 nH and 0.05 Ω in series, like a 1206 high-voltage part
Circuit: ngspice, 0.02 ns steps to 400 ns
Board: Raspberry Pi’s CM5 IO board. Its two inner ground planes as one sheet of 1 mm cells, each shell a metal block tied down at its tabs, and the CM5’s ground 1.5 mm above on its 51 ground pins.
Bench: The gun’s rod on the HDMI0 shell, its source 80 mm out. The exit is a wire from the power input to a ground plane 60 mm under the board.
Field: openEMS, solved per ampere up to 1.2 GHz, then put through the 8 kV zap

Simulated only. Nothing on this page has been measured on a board yet.

Limits

The bench is smaller than the IEC’s 0.8 m table, and cables are wires. That changes how the outer loop rings, not which way the current runs on the board.
FR-4 is lossless and copper is perfect.
1 nH per tab is an estimate.
The field model has no RC case. At ESD speeds the cap is a short, so the current spreads as it does straight.

Files

esdsim/iec.pyThe IEC 61000-4-2 current
esdsim/bond.pyThe shell ties in ngspice
esdsim/model.pyThe CM5 IO board’s field model
esdsim/synth.pyFrom per-ampere runs to the 8 kV zap

Sources

Like this note?

Rev 1 · 3 Nov 2026

Rev 13 Nov 2026First published

Each fix gets a line here, with the reader's name if they want credit.

Did it answer your question?

Want more on this?

Report a problem

What should the next note be? Ask for one