Why is one decibel figure not enough?

A: Difficulty varies by an order of magnitude across the spectrum: low-frequency fields need permeable material, high-frequency plane waves reflect from a conductor. A bare 60 dB figure names no band, no clamp state and no method, so it can neither be measured nor enforced at goods-in.

How much should a gasket be compressed?

A: Between 25 and 40 percent. Below 20 percent the rough moulded surface makes contact intermittent; above 50 percent the gasket takes a set and the next closure clamps nothing. Set groove depth from the measured stiffness of the case, because a rotomoulded wall is softer.

Does a vent always need a waveguide?

A: If it must both vent and shield, honeycomb is the simplest answer: small cells, enough depth, real attenuation below cutoff. A shallow louvre stops dust and does almost nothing to a field. Where no airflow is required, closing the hole in the wall beats any treatment.

Why not coat first and drill afterwards?

A: Drilling cuts the conductive layer, and the hole edge is bare polyethylene, so neither gasket nor plate makes contact. Drill and trim first, then coat, so the film wraps the hole edge and the interface plate has a flat, clean surface to bond onto. ---

EMI Shielding Case Specifications: Attenuation, Seams and Interfaces

Quick answer: Write an EMI shielding case specification in three groups: attenuation as band plus decibels, gasket compression plus contact resistance, and interface loss plus mounting. A single figure such as 60 dB says little, because low-frequency magnetic fields and high-frequency plane waves differ by an order of magnitude, and a bare shell can beat a finished case by 20 dB. Compression runs 25 to 40 percent, contact resistance stays under 10 milliohms per point, and drilling comes before coating.

Attenuation: band and decibels together
Gaskets: compression paired with contact resistance
Interfaces: filters versus waveguides
Process order: drill first, coat second
Five rows into the purchase clause
FAQ

1. Attenuation: band and decibels together

Shielding difficulty in an EMI shielding case changes with frequency. A low-frequency magnetic field needs a permeable path to divert flux, and a thin coating does nothing for it; plane waves reflect from a conductor. A bare 60 dB figure says nothing about either end.

Filtered connectors on a bolted metal interface plate

Separate the bare shell from the finished case. A shell alone reaches 80 dB, yet unfiltered connectors and one under-clamped seam drop the unit to 30 dB. Buy the complete EMI shielding case, measured as clamped, with margin for poor repeatability.

2. Gaskets: compression paired with contact resistance

In an EMI shielding case the gasket is the only conductive path between lid and body. Filled silicone is stable between 25 and 40 percent compression; below 20 percent a rough moulded surface makes contact intermittent, and above 50 percent the gasket takes a set.

Cavity coating with a braided ground strap along the wall

Hold contact resistance under 10 milliohms per point, continuous around the perimeter. The groove floor must conduct, since a gasket pressed onto release agent sits on plastic. Silver filler corrodes against aluminium, so wet service takes nickel-graphite.

3. Interfaces: filters versus waveguides

Every conductor crossing the shell of an EMI shielding case needs treatment. A filtered connector returns 40 to 60 dB above 100 MHz, with high density and simple mounting, but fixed capacitance and a high unit price. A waveguide tube works only below cutoff.

Closed latch with even compression marks on the gasket

Power takes filtered connectors, ventilation honeycomb, fibre a metallised sleeve. On an EMI shielding case large holes weaken the wall, so group connectors on a bolted plate, gasket it all round and let it reinforce.

4. Process order: drill first, coat second

Get the order wrong on an EMI shielding case and nothing downstream recovers. Drill, prepare, mask, prime, topcoat, cure, then fit gasket, plate and terminal. Coat before drilling and every hole edge shows bare polyethylene, so nothing conducts.

Finished case bonded to a ground bar in a test room

The polyethylene of an EMI shielding case has a surface energy near 31 millinewtons per metre, too low for paint, so flame or plasma treat it to 38. Decorative paint insulates, so a sprayed terminal is a disconnected terminal.

5. Five rows into the purchase clause

Collected into one table for an EMI shielding case, the specification becomes a purchase clause. Every row needs an acceptance method, or the number is an adjective: a chamber for shielding, probes for resistance, a cut section for compression.

Wall thickness is deliberately absent from an EMI shielding case sheet. It governs drop and stacking, not shielding. What deserves its own list is the interface schedule: how many holes, of what type, filtered or not, at what loss.

Parameter	Typical value	What it governs	How to check
Case shielding	30 MHz to 1 GHz, 40 dB or more	the final result	chamber, as clamped
Surface resistance	0.1 ohm per square or less	current spreading	four-point probe, five readings
Bond resistance	10 milliohms or less	ground return quality	micro-ohmmeter, terminal to shell
Gasket compression	25 to 40 percent	contact against rebound	cut section under a projector
Interface loss	40 dB or more above 100 MHz	the cable-entry leak	network analyser, port by port

Conclusion

Specify an EMI shielding case in order: band and case-level figure first, gasket and interface detail second, process sequence last in the work instruction. The band picks the material route, the gasket closes the longest seam, the interface covers the path everyone forgets, and the order decides whether the first three survive the shop floor. Only that last item can be checked piece by piece at the bench, and it is the one most often left out of the annex.

This article covers container-side specifications and design implications only, not the handling, loading or use of the contents.