The Component Signal · Issue #5
The Component Signal #005 — A 25% Tariff Doesn't Change Physics, It Changes Geography
The 25% tariff on Korean-origin electronics is live. Plus: the standing-wave math behind shield-can resonance, and why a 45 mm can failed at 10 MHz.
By Mike Kwak, Director · POCONS USA · How we report
The Tariff Math Just Changed
A tariff is a tax on a country-of-origin stamp. It does not care what is inside the box — only where the box says it came from.
The 25% US tariff on Korean-origin electronics took effect March 1, 2026. It applies to finished components and assemblies of Korean origin. Critically, raw material processed through Korea can also trigger duty if it is not substantially transformed elsewhere — so the question is not "where did I buy it" but "where was it last substantially transformed."
The market response was immediate: within 48 hours of the effective date, multiple customers requested Vietnam-origin certificates on shield and stamping orders. The teams that asked that question in January re-quoted calmly. The teams that waited are now absorbing 25% or scrambling for documentation under deadline pressure.
Origin is determined by substantial transformation, not by the last warehouse. A part fabricated and plated in Vietnam carries Vietnamese origin even if it transits a Korean consolidation hub. Verify the manufacturing location on every PO with Korean content.
POCONS position: Our high-volume shield and stamping production runs in Vietnam, with engineering support from our San Diego headquarters. For tariff-exposed customers, Vietnam-origin manufacturing is not a workaround — it is the standing configuration.
Price Watch: Metals
Copper's structural driver is real: a single hyperscale AI datacenter consumes 30,000–60,000 tonnes of copper in busbars, power distribution, and interconnect. Goldman Sachs has raised its 2026 copper forecast on datacenter intensity. This is demand that does not respond to price.
Design Corner: Shield-Can Resonance Is a Cavity Problem
A metal shield can is not just a barrier — it is a resonant cavity. Enclose a volume in conductor and you have built a microwave resonator whose modes will amplify energy at specific frequencies instead of attenuating it.
For a rectangular cavity, the resonant frequency of the TE₁₀₁ mode is:
f = (c/2) × √[(1/L)² + (1/W)²]
For a can where one dimension dominates, this collapses to the familiar half-wave approximation f ≈ c / 2L:
- A 30 mm internal length resonates near 5 GHz
- A 50 mm internal length resonates near 3 GHz
At resonance, the cavity Q can reach hundreds. Energy from a noisy source inside couples into the mode and the can radiates more through its seams and apertures than an unshielded board would at that frequency. The shield you added to fix emissions becomes the emitter.
Design rule: Keep the longest internal dimension below λ/4 at the highest frequency of concern. If the geometry forces a larger can, partition it with internal walls (compartment shields) so each sub-cavity's longest dimension stays sub-resonant.
Bench Note
A customer placed a 45 mm shield can over a 2 MHz switching regulator. The can's fundamental cavity resonance was ~3.3 GHz — nowhere near 2 MHz, so the design looked safe. But the 2 MHz square-wave switching node is rich in odd harmonics; the 5th harmonic at 10 MHz, and higher-order content, coupled into the cavity's lower-order modes through the can's geometry and aperture loading. Measured result: an 8 dB performance loss versus an unshielded reference at the affected band. The fix was a partitioned dual-can design — two smaller compartments, each sub-resonant — at zero BOM cost increase, since the same sheet metal was simply re-formed.
One Thing
A 25% tariff doesn't change physics. It changes geography. The electrons behave identically in San Diego and Hanoi — but the customs invoice does not.
— Mike Kwak, POCONS USA