Electrical

ESD at the home bench: what actually matters

A sceptical look at ESD protection for hobbyist electronics repair: what damages chips, what protects them, and what's just cargo cult.

Published 2026-05-15

Search "do I need an anti-static wrist strap" and you'll find two camps shouting past each other. One says every modern IC is a delicate flower and working without a strap is electronic manslaughter. The other says they've reseated RAM in socks on a nylon carpet for twenty years and never killed a thing. Both are partly right, which is what makes the topic so annoying.

The truth sits in an awkward middle. Modern silicon is significantly more robust than the bare CMOS parts that gave ESD its terrifying 1990s reputation — almost every input pin you'll touch has on-die protection diodes designed to clamp small zaps. But "more robust" is not "immune", and the genuinely nasty thing about ESD damage is that it rarely shows up at the bench. You don't see a puff of smoke. The part keeps working. Then three months later the laptop you fixed develops random USB dropouts, or the audio gets crackly, or the RAM throws a single-bit error on cold boot. By then nobody connects it back to the fact that you reseated a DIMM in December wearing a fleece.

So the goal of this article isn't to scare you into a full clean-room setup. It's to tell you what actually causes damage, what the cheap effective fixes are, and where most internet advice is cargo cult.

What ESD damage actually does

The classic failure mode is CMOS gate oxide breakdown. The insulating layer between a MOSFET's gate and channel is incredibly thin — a few nanometres on modern processes — and a high-voltage discharge punches a hole straight through it. Once it's holed, the gate leaks, the transistor behaves oddly, and the device is degraded or dead.

The good news: nearly every modern IC has integrated ESD-protection structures on its external pins. Diode pairs clamp the pin voltage to within a diode drop of the rails, dumping the energy through the supply network rather than the sensitive logic. Most consumer parts are rated to around 2 kV on the Human Body Model (HBM) test, which represents a person discharging through a defined resistance and capacitance.

The bad news: a person walking across a synthetic carpet in dry winter air will routinely generate 5 to 15 kV. You don't feel it under about 3 kV. That means every time you reach for a board without grounding yourself, you may be firing well over the rated HBM voltage at whatever pin you touch first. The protection diodes usually survive — but they degrade. Repeated sub-catastrophic zaps create "walking wounded" devices that meet spec at first and fail early. This is the failure mode that makes ESD damage so easy to dismiss and so expensive to chase.

The realistic risk hierarchy

High risk. Anything with exposed high-impedance CMOS inputs: discrete MOSFETs, op-amps, audio codecs, USB PHYs and re-drivers, FPGAs with pins exposed on a breakout, RAM chips out of socket, bare microcontrollers on anti-static foam.

Medium risk. Consumer ICs still soldered to a board (the PCB itself dissipates some charge), most board-level repair, connector reseating on assembled hardware.

Low risk. Through-hole work, discrete passives, power-stage components, mechanical assembly. You can be sloppy here and get away with it indefinitely.

What you actually need at home

Bare minimum: a wrist strap, properly grounded. A £3-5 strap from Amazon with the legally-required 1 MΩ resistor in the lead, clipped to a real earth point. The earth pin of a mains socket via a dedicated grounding plug is correct; clipping directly to anything live or to a random metal object is not. The 1 MΩ resistor is there to limit current if you accidentally touch something live — never bypass it.

Better: a dissipative ESD mat. Rubber mat on the bench, popper-stud to the wrist strap, mat earthed via its own lead. This means everything you put on the mat — board, screws, removed RAM — sits at the same potential as you. Cheap and worth it if you do more than occasional work.

Pro: ioniser, ESD-safe tools, dissipative footwear. Worth it if your room sits below about 30% relative humidity, or you're working on a static-prone carpet, or you're doing BGA rework where one bad part costs more than the whole setup.

Cheap myths and shortcuts

"I touched the case first, I'm fine." Works for that one discharge. The moment you move, your trousers and the chair are recharging you. By the time you reach back to the board you're carrying voltage again. A wrist strap solves this for the price of a coffee.

"My workbench is wood, it's fine." Wood is reasonably non-static, but it doesn't help if you drag a jumper across it or rest your forearm on a polyester sleeve.

"I just don't wear nylon." Genuinely a real factor — synthetic fabrics in dry air are the worst offenders. But it's risk reduction, not risk elimination. Cotton plus a strap is still better than cotton alone.

When the workshop goes full ESD

On the Hark Tech bench we run a dissipative mat, wrist strap, dissipative-handled tools, and a fume extractor that doubles as humidity control for the immediate work area. Anything BGA-related or involving exposed FPGA or RAM gets the full protocol — ioniser on, controlled handling, parts staying in their shielding until the moment they're placed.

For 99% of consumer repairs — laptop reseats, console board swaps, soldering a new USB-C port on — the strap-on-mat-on-earth setup is plenty. The marginal benefit of going further isn't zero, it's just not worth it unless you're doing the kind of work where a single dead chip ruins your day.

If you'd rather hand the soldering iron to someone with the setup already built, we do board-level repair here.

See also