ESD HBM Cdm
HBM and CDM are the two standard models used to qualify a chip's electrostatic discharge (ESD) protection, and they simulate two completely different real-world events: a charged person touching a pin versus a charged chip discharging through a pin as it touches a grounded surface.
These two models exist because ESD damage doesn't happen just one way, and a protection circuit tuned for one event can still be dangerously exposed to the other. A person walking across a carpet can build up a static charge and discharge it through a pin the moment they touch a chip or a board — that's the human body model, and it's been the industry's reference ESD threat for decades. A chip sliding down a shipping tube or across an assembly line can pick up its own static charge independent of any person, and then discharge that charge almost instantly the moment a pin touches a grounded surface — that's the charged device model, and it became a serious qualification concern specifically because modern packages and handling automation made this exact kind of self-charging event common.
Those two pulse shapes demand genuinely different protection circuit designs, which is why passing one model says very little about the other. HBM's slower, longer pulse gives a protection device time to turn on and conduct current away from sensitive circuitry, so an ESD clamp optimized for HBM can afford to be a bit slower to respond; CDM's pulse is over before a sluggish protection device can even finish turning on, which means CDM protection has to react in a fraction of the time and survive a much higher instantaneous current — a chip can pass a comfortable HBM qualification while still failing CDM, or vice versa, because the two events are stressing fundamentally different things about the same protection circuit.
| HBM | CDM | |
|---|---|---|
| What it simulates | A charged person touching a pin | A charged chip discharging through a pin |
| Discharge speed | Slower, roughly hundreds of ns | Extremely fast, under 2 ns |
| Peak current | Moderate | Very high |
| Governing standard | JEDEC JS-001 / ANSI-ESDA | JEDEC JS-002 |
| Typical qualification level | ~2 kV | ~500 V |
| Dominant risk source | Human handling | Automated assembly, shipping, handling |
CDM became a serious qualification concern specifically because the industry's own manufacturing automation created more opportunities for it, not because the physics changed. Older, more manual assembly processes gave parts more chances to pick up and bleed off static charge gradually, while modern high-speed automated handling — tape-and-reel feeders, pick-and-place equipment, shipping tubes — moves parts fast enough that a chip can build up significant charge on its own and then discharge it violently the instant a pin contacts a grounded surface, which is exactly the event CDM is built to simulate; design teams now treat CDM robustness as at least as important as HBM, and often harder to achieve given how little time the protection circuit gets to react.
Read HBM and CDM through a different-villain lens rather than a stricter-vs-looser lens: HBM asks whether a chip survives a charged person, CDM asks whether it survives its own charge discharging through automated handling, and because those are physically different events with different pulse shapes, a chip's ESD protection has to be designed — and qualified — against both separately before it's trusted to ship.