EMI Shield Cans vs Conductive Coatings: SMT DFM, Fixed vs Removable & RFQ

Factory methods guide: EMI shield cans vs conductive coatings — can materials and SMT DFM (NiAg vs SS, pick-up, lands), fixed vs removable frames, design factors (height, heat, gaps, BeCu), plastic housing paint mix/settle traps, can vs coating vs both matrix, and China PCBA RFQ fields.

Last updated
EMI shield cans vs conductive coatings: PCB can SMT vs plastic housing paint

Board-level EMI shield cans and conductive coatings on plastic housings are the two methods buyers most often argue after fundamentals are settled. A soldered can isolates a noisy island on the PCB; a conductive paint turns a plastic shell into a crude Faraday skin. Neither replaces continuous reference planes or stitch discipline — and China PCBA lines will invent pick-up, paste, and land defaults if the RFQ is silent. This is the methods deep dive: can materials and SMT DFM, fixed vs removable, design factors (height, power, frequency, gaps, BeCu springs), coating process traps, and a can-vs-coating-vs-both decision matrix.

EMI shield cans vs conductive coatings: PCB can SMT vs plastic housing paint

Answer first: can, coating, or both

QuestionFactory / procurement answer
Why shield at board / enclosure?Contain an aggressor island or harden a plastic shell so energy does not leak / enter — methods after planes and zoning exist
What is a shield can?Drawn / stamped metal cover soldered (or clipped) to PCB ground lands — SMT or hand place
Fixed vs removable?Fixed = one-piece soldered lid (cheapest, hard to rework). Removable = soldered frame + snap / spring lid (debug / rework access)
Where cans winDiscrete RF / clock / converter islands; defined keepout; need hard local attenuation on the board
Where coatings winLarge plastic housings; no room for a can; whole-product skin; cost-sensitive consumer shells
When both?Noisy PCB islands under cans plus painted / coated enclosure for system-level containment
Fear list (implicit)Can on wrong net; SS can that will not wet; settled paint with thin spots; lid gaps that radiate

💡 Procurement tip: Put can footprint / height / material / finish (NiAg vs SS), fixed vs removable frame, land net + mask openings, and coating coverage + mix/stir notes on the RFQ. “EMI shield” with no method is how quotes mix a can line and a spray line without matching the design.

Why you still need a method after fundamentals

Planes, stitch fences, and zone maps decide whether a barrier has a return. Methods decide where metal sits:

  • Board-level can — short Faraday box over one IC or block, tied to the board ground net through solder lands.
  • Enclosure coating — conductive film on the inside of a plastic housing, tied to chassis / PCB ground at defined contact points.
  • Both — common when the board has hot islands and the product shell is plastic with no metal die-cast body.

Skipping the method choice and ordering “add EMI” after layout freeze is how pre-scan fails twice: once for the island, once for the shell aperture.

Shield can — definition, materials, SMT

A shield can (EMI can / RF can / fence-and-lid) is a thin metal enclosure that sits on the PCB and covers a keepout zone. Production forms:

FormHow it attachesFactory note
One-piece soldered canEntire drawn box reflowed to perimeter landsFast SMT; lid is permanent
Two-piece (frame + lid)Frame soldered; lid snaps or springs onRemovable for debug / rework
Fence / wall onlyPerimeter wall without full lidPartial; often pairs with a lid later

Materials buyers actually quote

Material / finishWhy it shows upSMT / solderability
Tin-plated steel or nickel-silver (NiAg) classCommon SMT cans; solderable skirtsGood wetting on paste lands when finish is specified
Stainless steel (SS)Stiffness / corrosion storiesOften poor solder wetting unless plated or process-qualified — call finish explicitly
AluminumLight, conductiveUsually not a drop-in solder can without special attach
Copper / brass familyHigh conductivityCost / oxidation / finish must be on the traveler

Factory rule: Do not assume “metal can = solderable.” NiAg / tin-finish cans and SS cans are different paste and heat stories. Put material + surface finish on the BOM and assembly notes.

SMT pick-up and land pattern (China PCBA DFM)

What the SMT line needs called out:

DFM itemWhy
Pick-up surfaceFlat lid top for nozzle; tape/reel or tray packaging; no warped drawn parts that fail vision
Land patternContinuous or segmented ground pads matching the can skirt; clearance from signals under the wall
Mask openingsSolderable pads — masked “ground” under the skirt does not wet
Paste apertureEnough paste for fillet without bridging into the keepout or adjacent pads
Height / keepoutCan clearance above tallest part under the lid; nozzle and neighboring tall parts
Polarity / orientationAsymmetric cans need silkscreen or feeder notes

A can that looks fine in 3D and fails first article because the skirt sits on soldermask or the nozzle cannot grip a glossy lid is a drawing failure, not a mystery EMI fail.

Fixed vs removable — debug and rework decision

TypeStructureWins whenLoses when
Fixed (one-piece)Whole can solderedCost, cycle time, no lid logisticsMust desolder to probe / replace the IC under the can
Removable (frame + lid)Soldered frame; snap or spring lidBring-up, field rework, FA access, iterative EMC tweaksExtra part, lid retention QC, spring contact quality

Procurement framing:

  • Prototype / NPI / RF modules that will be probed → budget removable.
  • Locked consumer BOM with no under-can service → fixed is fine.
  • Mixed: solder a frame early, ship lids after EMC tune — only if the traveler allows staged assembly.

Removable is not “more EMI.” A loose lid with daylight at the seam can radiate worse than a well-soldered fixed can. Retention and seam continuity are part of the EMI design, not packaging cosmetics.

Design factors that decide can performance

Textbook “metal box” fails in production for mundane geometric reasons:

FactorWhat goes wrongBuyer / layout lever
HeightParts touch the lid (shorts / abrasion) or too-tall can fights the enclosureMax part height + clearance note; can height on RFQ
Power / heatCan traps heat on a hot regulator or PAThermal vias, vent slots (slots are apertures — trade carefully), or leave the thermal island un-canned
FrequencySlot / gap length approaches wavelengths of concernKeep seams short; stitch density under the skirt; no long ungrounded slots “for airflow” without EMC review
Gaps / seamsLid-to-frame gap, cutouts for connectors, incomplete solderSpec max gap intent; BeCu or spring fingers where lids remount
BeCu springs / fingersRemovable lids need reliable ground at the seamCall spring material / plating; inspect crush set and plating wear on rework cycles

Gaps dominate. A beautiful can with a 10 mm ungasketed connector cutout is an intentional aperture. Treat every opening as an EMC feature on the drawing.

Where cans win

Cans are the right first method when:

  • The aggressor (or victim) is a compact board island with a clear keepout.
  • You need local containment without painting the whole product.
  • SMT capacity and land nets already exist on the board.
  • Removable access to one RF / clock block matters more than enclosure paint QA.
  • The housing is already metal — coating adds little; the residual problem is on-PCB.

Cans lose when the shell is large plastic with many seams, the noise is system-level cable/aperture driven, or there is no space for a can footprint and height.

Conductive coating / paint on plastic housings

Conductive coating (nickel, copper, silver-filled paints and similar production coats — quote by generic chemistry class, not brochure brand) is sprayed or otherwise applied to the inside of plastic enclosures so the shell behaves more like a conductive barrier. Typical jobs:

  • Consumer / IoT plastics that cannot afford die-cast metal
  • Large volumes where a can cannot cover every aggressor
  • Closing the system Faraday story when the PCB already has local cans — or when cans are impractical

What the coat must do electrically

NeedFactory language
Continuous coverageNo thin spots at corners, ribs, or boss shadows
Ground tieDefined contact to PCB ground / chassis studs / EMI gaskets — floating paint is a floating hope
AdhesionPaint that peels after drop / humidity is an open seam
Thickness classCall the process window the coater qualifies — do not invent dB claims on the PO

Process traps — mixing, settling, coverage (qualitative)

Filled conductive paints settle. Pigment and binder separate in the can and in the spray pot.

TrapWhat happensControl
No stir / incomplete mixFirst parts rich, later parts starved — or the reverseTimed mix / stir before and during spray; pot agitation per process sheet
Settling in hose / gunPatchy conductivity; “looks coated” visuallyRecirculation / purge discipline; first-article conductivity check
ShadowingRibs, bosses, deep draws miss paintSpray angles, multiple passes, mask strategy
Over-thin / over-thickThin = weak barrier; thick = peel, orange peel, fit issuesQualified wet-film / process coupon — not eyeball
Ground contacts painted shut or left bare wrongOpens or shorts at studs / spring contactsMask map on the coating drawing

Failure path: paint looks uniform in photos → settled batch ships thin at corners → pre-scan fails → team adds a can on the PCB → second fail because the shell aperture still dominates. Coat process QA is EMI QA.

Can vs coating vs both — decision matrix

SituationPreferWhy
One noisy IC / RF block on a dense boardCanLocal, SMT-friendly, clear keepout
Need probe / rework under the shieldRemovable canFrame + lid; springs if specified
Large plastic consumer shell, few board islandsCoatingSystem skin without many can footprints
Hot PA + plastic IoT box + cable exitsBoth + aperture disciplineCan on PA; coat shell; gasket / cable entries still required
Metal enclosure already continuousCan (board) or nothing extraCoating on metal is usually wasted
No land net, no height, layout frozenCoating or enclosure gasket path — not a late can fantasyLate cans on wrong nets still fail

Both is not “belt and suspenders free.” It is two travelers: SMT can DFM and coating mix/coverage/ground-tie notes. Quote both lines or accept that one will be guessed.

Can vs coating vs both decision: board island can, plastic shell paint, combined

China PCBA / SMT DFM checklist for cans

Hand this to the assembly house with the BOM:

  1. Footprint — land pattern, segment vs continuous skirt, clearances.
  2. Height — can ID height vs max component under lid + clearance.
  3. Material + finish — NiAg / tin-plate vs SS (and plating if SS).
  4. Fixed vs removable — one-piece or frame+lid; lid retention method.
  5. Pick-up — packaging, top flatness, polarity.
  6. Paste / mask — openings on can lands only; paste volume note if non-standard.
  7. Net — can lands on the intended ground; no signal under the wall.
  8. Keepouts — no vias/solder beads that lift the skirt; no tall neighbors in the nozzle path.
  9. BeCu / springs — if removable, spring count, plating, and inspection after rework.
  10. AOI / FA — solder fillet criteria for the skirt; lid seated check for two-piece.

Coating RFQ fields (when the shell is in scope)

If XFPCB or a partner quotes PCBA and you own the plastic coat:

  • Chemistry class (Ni / Cu / Ag-filled — generic) and qualified applicator
  • Coverage map (faces, ribs, exclusions)
  • Mix / stir / pot-life notes on the process sheet
  • Ground-tie locations to PCB / studs
  • First-article conductivity or continuity method (qualitative acceptance — no invented SE numbers)
  • Mask-offs for bearings, optics, logos, mating surfaces

Buyer RFQ checklist (methods → factory)

  1. Method intent: can / coating / both — not “EMI TBD.”
  2. Can: footprint, height, material, finish, fixed vs removable.
  3. Can lands: net, mask openings, paste, pick-up, keepout.
  4. Removable: frame + lid + spring/BeCu notes if used.
  5. Coating: coverage map, mix/stir discipline, ground ties.
  6. Apertures: connector / display / cable exits called as EMC features.
  7. Do not invent shielding-effectiveness dB limits on the PO — attach the test plan or customer annex.
  8. Align with plane / stitch fundamentals already on the fab notes; a can on a broken return still fails.

What this post is — and what it is not

Covered hereCovered elsewhere
Board-level cans: materials, SMT DFM, fixed vs removableEMC/EMI/EMS vocabulary, reflection/absorption physics
Design factors: height, heat, frequency, gaps, BeCuFlex silver-paste / carbon / foil constructions
Conductive coatings on plastic + mix/settle trapsQuantitative SE tables and legal limit values
Can vs coating vs both matrix + RFQ fieldsCable braid / gasket encyclopedia

This slug is the methods companion to fundamentals and flex constructions: rigid-board cans, enclosure coatings, and the China PCBA notes that make either buildable.

Shield cans and conductive paints are tools. Continuous returns, honest land nets, and stirred paint are how the method you paid for is the method that ships.

EMI shield cans and conductive coatings FAQ

When should I use an EMI shield can vs a conductive coating?

Use a board-level can for a compact noisy or sensitive island with a defined keepout and ground lands. Use conductive coating when the product shell is plastic and you need a system-level skin. Use both when hot PCB islands and a plastic enclosure both drive emissions or immunity — and budget two travelers (SMT can DFM plus coat mix/coverage/ground ties).

Fixed vs removable shield can — which for debug and rework?

Fixed one-piece cans are cheapest and permanent: desolder to access parts underneath. Removable designs solder a frame and snap or spring a lid for bring-up, FA, and field rework. Removable is not automatically better EMI — lid seam continuity and BeCu/spring contact quality decide whether the removable can still shields.

What SMT DFM notes does a China PCBA line need for shield cans?

Call footprint and land net, mask openings, paste intent, can height vs max part under the lid, pick-up packaging, material and finish (NiAg/tin vs stainless with plating), and fixed vs frame-plus-lid. Silent RFQs invite wrong-net lands, masked pads that will not wet, and SS cans that refuse solder.

Why does conductive paint fail even when the housing looks coated?

Filled paints settle. Incomplete mix or pot settling produces thin corners and ribs that look painted but lack conductivity. Missing ground ties to the PCB or chassis leave a floating film. Require mix/stir discipline, coverage maps, and first-article continuity — do not invent shielding-effectiveness numbers on the PO.

What RFQ fields should buyers list for cans and coatings?

Method intent (can / coating / both); can footprint, height, material, finish, fixed vs removable; land net and mask openings; BeCu/spring notes if removable; coating coverage map, mix/stir notes, and ground-tie locations; plus aperture callouts for connectors and cables. Attach the EMC test plan rather than fake dB limits.