ESD Protection in PCB Assembly: Standards & Best Practices

ESD protection is the set of controls — grounded workstations, the right packaging, and on-board clamping devices — that stops electrostatic discharge from damaging components during PCB assembly and in the field. You generally cannot feel a discharge below about 3,000 V, yet a modern IC can fail below 100 V, and the worst damage is latent: the board passes every test on the line, then dies weeks later in a customer’s hands. Effective ESD protection works on two fronts — controlling the assembly environment to the limits in ANSI/ESD S20.20, and designing protection into the board itself with TVS diodes. This guide covers both: the standards that govern an ESD workstation, the resistance numbers that actually matter, how to add board-level protection without wrecking signal integrity, and the mistakes that quietly sink reliability.

ESD Protection in PCB Assembly: Key Takeaways

• ESD protection has two halves: control the assembly environment (an EPA built to ANSI/ESD S20.20) and design protection into the board (TVS diodes at the connectors).
• The numbers that matter: a wrist-strap system under 35 MΩ (with a 1 MΩ resistor), worksurfaces and floors under 1×10⁹ Ω, and operator body voltage under 100 V.
• Most ESD damage is latent — the board passes test and fails in the field — so a clean test yield does not prove your ESD controls work.
• Humidity helps but is not a control method; you can still destroy a part at 50% RH, which is why grounding and dissipative materials do the real work.
• On high-speed lines the protection device is also a signal-integrity device: keep TVS capacitance under about 0.5 pF on USB 3.x and HDMI, and place it right at the connector.

What Is ESD and How Electrostatic Discharge Damages PCBs

Electrostatic discharge is the sudden transfer of charge between two objects at different potentials — the same spark you feel touching a doorknob after walking across carpet. In PCB assembly that spark passes through a semiconductor, and the energy that is harmless to you punches through gate oxides, melts bond wires, and degrades junctions. The industry models the event three ways, because charge reaches a device by more than one route.

Model	What it simulates	Key parameters	Status
HBM (Human Body Model)	A charged person touching a device	100 pF through 1.5 kΩ	Primary model (JS-001)
CDM (Charged Device Model)	A charged device discharging to ground	Sub-nanosecond, very high peak current	Often more damaging (JS-002)
MM (Machine Model)	A charged tool or machine	200 pF, near-zero series resistance	Largely withdrawn since ~2015

ANSI/ESD S20.20 sets its default program to protect devices rated 100 V HBM and 200 V CDM or higher; anything more sensitive needs a tailored program. The Charged Device Model deserves attention because it is faster and frequently more destructive than the human-body event — a board or component picks up charge on a conveyor or in a feeder, then dumps it in under a nanosecond when a grounded pin makes contact.

The failure that costs the most is the one you cannot see. A catastrophic hit kills the part on the spot and shows up at functional test. A latent hit only weakens it — increased leakage, reduced margin — so the board passes every inspection and then fails intermittently in the field. That is why a clean test yield is not evidence that your ESD controls are working; the damaged units shipped.

ESD Standards for PCB Assembly: ANSI/ESD S20.20, IEC 61340-5-1, and the MIL Legacy

The governing program standard in North America is ANSI/ESD S20.20, published by the ESD Association, which tells an organization how to build and verify an ESD control program rather than dictating specific products. IEC 61340-5-1 is the technically equivalent international standard. On the acceptance side, IPC-A-610 flags ESD-related defects on finished assemblies, and the device-level test methods live in the joint ANSI/ESDA/JEDEC documents JS-001 (HBM) and JS-002 (CDM).

Do not confuse those handling standards with IEC 61000-4-2, which is a product immunity test: it zaps a finished product with ±8 kV contact and ±15 kV air discharge (Level 4) to prove the design survives real-world events. S20.20 keeps charge away from bare boards on the line; IEC 61000-4-2 proves the sealed product can take a hit in a user’s hands. You need both mindsets.

If you work in aerospace or defense, you will still see the military documents referenced. MIL-STD-1686 was the mandatory DoD ESD control program standard — the “what you must do.” Its companion, MIL-HDBK-263, was the non-mandatory handbook explaining the science and the “how.” Both were canceled by the DoD on January 12, 2021 and superseded by ANSI/ESD S20.20 and the ESD TR20.20 handbook, so a modern program builds to S20.20 while legacy contracts may still cite the MIL numbers.

How to Set Up an ESD Workstation (EPA Requirements)

An ESD Protected Area (EPA) is any zone — a single bench, a line, or a whole room — where charge is controlled and every conductor, including the operator, is bonded to a common ground point. Build the workstation in this order.

1. Establish a common ground point. Bond the worksurface, wrist strap, tools, and fixtures to one earth-referenced point rather than daisy-chaining them, so a single broken connection cannot float an item.
2. Ground the operator with a wrist strap. The cord carries a 1 MΩ current-limiting resistor, and the whole person-plus-strap-plus-cord path must measure under 3.5×10⁷ Ω (35 MΩ). Test straps daily, or use continuous monitors for high-value builds.
3. Install a static-dissipative worksurface. It should read under 1×10⁹ Ω to the groundable point — dissipative, not bare metal, so charge bleeds off without a fast, damaging discharge.
4. Add a dissipative floor and footwear for mobile operators. Qualify the floor-plus-footwear system to under 35 MΩ and confirm body voltage stays under 100 V with an actual walking test, not a resistance reading alone.
5. Neutralize what you cannot ground. Insulators (most plastics) and isolated conductors hold charge that no wrist strap reaches; keep isolated conductors under 35 V and use ionizers on process-required insulators near the board.
6. Package ESDS items for transport. Anything leaving the EPA goes into static-shielding (Faraday-cage) packaging per ANSI/ESD S541, not ordinary plastic or cardboard, which generate charge.

Here are the limits in one place, drawn from ANSI/ESD S20.20.

ESD control element	S20.20 limit	What it does
Wrist-strap system (person + strap + cord)	< 3.5×10⁷ Ω (35 MΩ)	Drains body charge; 1 MΩ resistor limits current
Worksurface (to groundable point)	< 1×10⁹ Ω	Bleeds mat charge slowly, no hard discharge
Flooring + footwear system	< 35 MΩ and < 100 V body voltage	Grounds mobile operators; verified by a walking test
Static-dissipative floor	1×10⁶ – 1×10⁹ Ω	Standard range for SMT and assembly areas
Conductive floor	2.5×10⁴ – 1×10⁶ Ω	Faster drain where the process requires it
Isolated conductor	< 35 V	Limit on ungrounded metal near an ESDS item
Insulator	≥ 1×10¹¹ Ω	Cannot be grounded — treat with ionizers

Two values surprise people. The worksurface and floor are meant to be dissipative, not conductive — a near-zero-ohm path would discharge a charged device in a fast, damaging spike and create a shock hazard. And the floor-plus-footwear requirement is not a resistance number alone; S20.20 added a walking body-voltage test in 2014 after data showed an operator can exceed 100 V even with a system that passes the resistance check.

Designing ESD Protection Into the PCB: TVS Diodes and Layout

Handling controls protect the bare board on the line. Once the product ships, the only defense against a user’s charged finger or cable is what you designed onto the board — transient voltage suppression (TVS) diodes that clamp the surge and shunt it to ground. The rules are simple and unforgiving.

• Place the TVS at the connector, hard against the external pin. ESD current takes the shortest path to ground; route it a centimeter inward “for convenience” and the surge finds a shorter path — straight through your transceiver.
• Give it a low-inductance ground. Multiple stitching vias and wide copper to the ground plane, with short, symmetric stubs on differential pairs.
• Match the diode to the signal. Unidirectional TVS for DC-biased digital lines that never go negative; bidirectional for signals that swing both ways, such as RS-485 or analog audio. Use power TVS on rails like VBUS and low-capacitance arrays on data pairs.

A consumer-device maker we worked with passed ±8 kV contact testing in the lab but saw USB ports dying in the field after users plugged in cheap, charged cables. The low-capacitance array was a centimeter from the connector to ease routing — far enough that the surge found a shorter path through the transceiver. Moving the array hard against the connector pins and adding ground stitching vias fixed it with no part change. Placement is where most ESD “mystery failures” begin.

Why TVS Capacitance Matters on High-Speed Lines

On a high-speed interface the protection device is also a signal-integrity device. Every TVS adds junction capacitance to the line, and on a multi-gigabit pair that capacitance loads the channel, rounds edges, closes the eye, and adds insertion loss. A general-purpose TVS at 10 pF or more will wreck a USB 3.x pair; you need a low-capacitance array sized to the data rate.

Interface	Typical data rate	Target TVS capacitance
USB 2.0	480 Mbps	up to ~5 pF
Gigabit-plus data lines	> 1 Gbps	< 1 pF (ideally < 0.5 pF)
USB 3.2 / HDMI 2.1 / Thunderbolt	10–40 Gbps	~0.1 – 0.5 pF

If a single low-capacitance part still loads the line too much, add a small series resistor of 2–3 Ω to damp reflections, keep the tap-point stub short and symmetric, and hold the controlled impedance through the protection node. At multi-gigabit rates, choosing and placing the ESD device becomes part of the high-speed routing job, not an afterthought.

Common ESD Protection Mistakes in PCB Assembly

Hand this list to anyone running an assembly line or laying out a new board.

1. Trusting humidity as a control. Raising RH reduces charge generation but never eliminates it — you can still damage a part at 50% RH, and HVAC drifts. Ground and dissipate; do not rely on the weather.
2. Leaving isolated conductors floating. Ungrounded metal — a fixture, a tray, a connector shell — holds charge that no wrist strap reaches. Bond it or keep it under 35 V.
3. Assuming a wrist strap works because it is worn. Cords fail and skin dries out; test daily or monitor continuously, and remember the path includes the operator.
4. Qualifying a floor by resistance alone. A system that passes the resistance check can still let a walking operator exceed 100 V — run the body-voltage walking test.
5. Placing the TVS away from the connector. Surge current takes the shortest path; a distant diode lets it route through the IC instead. Diode first, then route inward.
6. Using a high-capacitance TVS on a fast line. A 10 pF part closes the eye on a USB 3.x or HDMI pair; pick a sub-0.5 pF array for multi-gigabit signals.
7. Keeping insulators in the EPA. Common plastics and packaging near the board generate charge; remove them or neutralize with an ionizer.
8. Skipping shielding packaging outside the EPA. Boards travel in static-shielding bags or conductive totes per ANSI/ESD S541, never bare in ordinary plastic.

Frequently Asked Questions About ESD Protection

What voltage of ESD damages a PCB component?

Many modern devices are damaged below 100 V, and the most sensitive can fail under 25 V — well beneath the roughly 3,000 V a person can actually feel. That gap is the whole problem: operators discharge damaging energy into parts without ever noticing a shock, which is why grounding is continuous, not occasional.

What is an ESD workstation?

An ESD workstation is a grounded bench inside an ESD Protected Area where every conductor — worksurface, wrist strap, tools, and operator — is bonded to a common ground point. A compliant station keeps the wrist-strap system under 35 MΩ and the dissipative worksurface under 1×10⁹ Ω to ground.

Why is there a 1 megohm resistor in an ESD wrist strap?

The 1 MΩ resistor limits current if the wearer touches a live voltage, protecting the person, while still bleeding static charge to ground in milliseconds. It is a safety element, not an ESD requirement on its own; the full person-plus-strap path must still measure under 35 MΩ.

What is the difference between HBM and CDM?

HBM models a charged person touching a device — 100 pF through 1.5 kΩ, a relatively slow event. CDM models a charged device discharging when a pin contacts ground — sub-nanosecond and very high peak current. CDM is often the more destructive in automated assembly, where components self-charge on feeders and handlers.

Does humidity prevent ESD damage?

It helps but does not prevent it. Higher humidity (above 40–50% RH) lowers charge generation, while dry air (10–20% RH) produces the highest voltages. Components can still be damaged at high humidity, and HVAC varies, so S20.20 treats grounding and dissipative materials — not humidity — as the controls.

Is ANSI/ESD S20.20 mandatory?

It is voluntary unless a customer or contract requires it, but it is the de facto standard for electronics manufacturing and is referenced by most aerospace, automotive, and defense programs. Facilities can be third-party certified to it, which many OEMs require of their assembly suppliers.

Do I need TVS diodes if my assembly line is ESD-controlled?

Yes. An EPA protects the bare board during manufacturing; it does nothing once the product is in a user’s hands. Any externally accessible port — USB, HDMI, buttons, antennas — needs on-board TVS protection to survive field ESD events tested under IEC 61000-4-2.

What ESD standards apply to PCB assembly?

ANSI/ESD S20.20 (or IEC 61340-5-1 internationally) governs the handling program; IPC-A-610 covers ESD-related defects on finished assemblies; JS-001 and JS-002 define device-level HBM and CDM testing; and IEC 61000-4-2 tests finished-product immunity. Legacy defense work may still reference MIL-STD-1686.

Building ESD Protection Into Your Assembly Process

ESD protection only works when both halves are in place: an assembly environment controlled to ANSI/ESD S20.20 and a board designed to survive the field. Get the EPA numbers right — a wrist-strap system under 35 MΩ, surfaces and floors under 1×10⁹ Ω, body voltage under 100 V — test them on a schedule rather than trusting that the controls still hold, and treat the on-board TVS as a signal-integrity decision, placed at the connector and sized to the data rate. Do that and you stop shipping the latent failures that a clean test yield hides.

Send us your schematic, BOM, and the interfaces that face the outside world, and we will review your ESD protection — handling and board-level — with a DFM check and quote.

Standards reference: ANSI/ESD S20.20, IEC 61340-5-1, IPC-A-610, ANSI/ESDA/JEDEC JS-001 and JS-002, and IEC 61000-4-2 are published by the ESD Association, IPC, JEDEC, and the IEC.