EMI Shielding: Designing Spring Contacts and Cans That Pass Testing
Short answer: EMI shielding fails at gaps, not through metal. A stamped shield can or spring contact is only as good as the joints between its panels — the seams where covers meet frames, and the fingers that bridge them. Design those interfaces with enough contact points, contact force and plating, and a stamped enclosure passes radiated-emissions testing; leave a 1 mm gap or a relaxing spring finger and no amount of shield metal will save the result.
Every electronics product with a radiated-emissions limit — which in practice means everything sold commercially — relies on stamped metal parts to keep noise in and interference out. The shield can that covers a module, the spring fingers that ground a removable cover, and the clip that bonds a board to its chassis are all stampings. Their geometry is simple; their job is unforgiving. This guide covers the design rules that make stamped shielding pass the test the first time.
Why Shielding Fails at Seams
A solid metal box is a perfect shield — the problem is that no real product is a solid box. Covers must open, boards must slide in, cables must pass through. Every opening is a slot antenna that leaks at wavelengths related to its length. A gap of a few centimetres can leak at cell-phone frequencies; even a hairline seam leaks where the two surfaces do not actually touch metal-to-metal.
| Gap/seam type | Typical leak problem | Shielding fix |
|---|---|---|
| Cover-to-frame seam | Long slot antenna | Spring fingers every few cm |
| Board-to-shield contact | Intermittent grounding | Stamped clips, grounded edges |
| Connector cutouts | Aperture leakage | Gaskets, can extensions, screw spacing |
| Ventilation holes | Large openings | Perforation pattern below λ/20 |
| Screw points only | Gaps between fasteners | Add fingers or conductive gasket |
Takeaway: shielding effectiveness is decided by the worst gap, not the average panel. Designers who compute shielding at the metal and ignore the seams get a rude surprise at the compliance lab, which is why spring contacts and can geometry deserve as much attention as the enclosure material.
Stamped Shield Cans: The Base Structure
The shield can itself is a stamped sheet-metal box — typically tin-plated steel, nickel-silver, or aluminum — that sits over a noisy circuit and soldered or clipped to the board ground. Stamping makes cans cheap at volume, with the drawn walls, locking tabs and board-mount features formed in one die. The can's shielding effectiveness depends on its material, its seams, and how completely it connects to the ground plane.
| Can material | Typical use | Notes |
|---|---|---|
| Tin-plated steel | General-purpose cans | Good SE, solderable, low cost |
| Nickel-silver | Corrosion-sensitive areas | Good solderability, whiter finish |
| Aluminum | Weight-sensitive modules | Needs separate grounding path |
| Copper alloy | RF modules | Best conductivity, pricier |
Takeaway: material matters less than continuity. A steel can with fingers every 10 mm outperforms a copper can grounded at one screw, because the fingers close the seams that actually leak. Our stamped sheet-metal enclosures are designed with this seam logic first, wall thickness and material second.
Spring Contacts: The Gap Closers
Spring contacts and finger stock exist to press two metal surfaces together reliably, over thousands of open-and-close cycles, without soldering. They must deliver three things at once: enough contact force to keep resistance low, enough deflection to tolerate tolerance stack and wear, and a material that does not relax over life. That is why beryllium copper dominates the category, with phosphor bronze and hardened stainless as cost-driven alternatives.
| Material | Typical max service temp | Fatigue/relaxation | Typical use |
|---|---|---|---|
| Beryllium copper C172 | ~150 °C+ | Excellent | Board-level shields, fingers |
| Phosphor bronze C5210 | ~100–150 °C | Good | Lower-cost fingers, clips |
| Hardened 301 stainless | ~250 °C+ | Good | High-temp and harsh environs |
| Brass, nickel-silver | ~100 °C | Fair | Static grounding clips only |
Takeaway: choose the spring material by temperature and cycle life, then plate it for the corrosion and wear environment. A relaxing finger is a silent failure — the shield tests fine at first article and leaks a year later, which is why relaxation resistance matters more than raw conductivity in stamped contact springs.
The Design Numbers That Matter
Three numbers govern whether a spring contact passes: contact force, deflection range, and the plating at the interface. Force must stay high enough to break through surface films — oxides and contamination — typically tens of grams per finger at minimum, scaled by the duty. Deflection must cover the tolerance stack between the cover and the frame without overstressing the material. Plating, usually tin or gold over nickel, keeps the mating surface conductive over life.
| Design parameter | Typical value/rule | Failure mode if wrong |
|---|---|---|
| Contact force per finger | 30–150 gf typical | Low force = oxide film resistance |
| Working deflection | 20–50% of free height | Over-deflection = permanent set |
| Finger pitch | 5–15 mm typical | Wider pitch = seam leaks |
| Plating | Tin or gold flash over nickel | Bare metal corrodes at the wipe |
| Mating surface | Clean, flat, plated land | Paint or anodize kills the joint |
Takeaway: these numbers interact. Raise force and you need stiffer material or more deflection room; raise deflection and you need better fatigue life. The drawing should state force at a given deflection, not just a free height — that single spec is what makes a spring contact predictable on the assembly line and in the lab.
Plating and Grounding: The Electrical Half
A spring finger that touches an anodized or painted surface is not grounding anything. The mating land on the cover or frame must be bare, plated metal, and the finger itself needs plating matched to the environment — tin for cost, gold flash where the product sees corrosive atmospheres or very low signal levels. Grounding strategy matters too: fingers ground the cover to the frame, clips ground the board to the chassis, and both paths must land on the same ground reference the shield can connects to.
| Interface | Plating pair | Reason |
|---|---|---|
| Finger to cover land | Tin-tin or Au-tin | Low and stable contact resistance |
| Can to board pad | Solder joint | Permanent, high-integrity bond |
| Clip to chassis | Tin or nickel | Corrosion-resistant pressure joint |
| Board edge to can | Multiple clips in parallel | Redundancy for grounding |
Takeaway: treat every interface as an electrical joint with a plating spec, not as metal touching metal. The same plating logic that protects stamped electrical contacts applies to shielding, and the same inspection records should follow the parts into production.
Passing the Test the First Time
Compliance testing is expensive, so design shielding margins before the lab visit. Space fingers closer than the calculation demands to cover wear and tolerance, specify plating thickness with margin, and prototype the actual cover-and-frame assembly — a shield can tested alone tells you nothing about the seam between it and the cover. Sample builds that include the stamped fingers, the can and the mating land will find grounding problems while they are still CAD changes, not redesigns.
On the manufacturing side, consistency is the test-passing variable: fingers stamped from one coil lot behave like fingers from the next only if material temper and plating stay in spec. That is why we run stamped shielding parts with incoming coil inspection and batch plating records under ISO9001 — a shield that varies batch to batch is a test result that varies batch to batch.
Email sc@bquq.com or WhatsApp +86 137 1315 7787 with your PDF/DXF/STEP file. An engineer reviews it and replies with price, lead time and DFM notes on working days.
Frequently Asked Questions
Q: How many spring fingers does a shield cover need?
Enough that the spacing between contact points stays below the wavelength of the highest frequency you must contain — in practice, fingers every 5–15 mm around the cover perimeter. Fewer fingers mean longer slots and more leakage at high frequencies.
Q: Beryllium copper or phosphor bronze for EMI fingers?
Beryllium copper wins on relaxation resistance and cycle life, which is why it dominates board-level shielding. Phosphor bronze is acceptable for lower-cycle, cost-sensitive designs. If the environment runs hot, hardened stainless is the durable choice.
Q: Does the shield can need to be soldered all around?
No — cans are often clipped or tack-soldered at a few points. What matters is that the can connects to the ground plane at multiple points and that any openings stay small. Full perimeter solder is used only where very high frequencies demand it.
Q: Why did my shield pass bench testing but fail the compliance lab?
Because the bench test did not reproduce the real seams, tolerances and grounding. Covers warp, fingers relax, lands corrode. Test the full assembly with production-tolerance parts, and check contact force and plating on actual stamped samples, not ideal CAD geometry.
Q: Can stamped fingers be plated selectively?
Yes. Selective plating puts tin or gold only on the contact zone, which cuts precious-metal cost and keeps the rest of the strip clean for forming. The plating must be registered to the strip layout so the finished finger wipes on the plated band.
Authored by the BQUQ Engineering Team. BQUQ is an ISO9001-certified source factory in Dongguan, China, running CNC machining, metal stamping, custom springs, heat sink and collet lines under one roof. Send drawings to sc@bquq.com or WhatsApp +86 13713157787 for a quote within 12 working hours. www.bquq.com
