Contact Us
  • Home
  • BLOG
  • How Do You Stop Tin Whiskers on a Plated Shield?

How Do You Stop Tin Whiskers on a Plated Shield?

You cannot stop tin whiskers, and any coating supplier who says otherwise is selling a film. What you can do is shorten them until the longest one in the batch is shorter than the gap it has to cross, and that is a decision made in three places: the layer under the tin, the heat treatment after plating, and the thickness of the coating over it. Do all three and the historical worst case drops from 800 um to under 100 um.

The reason it matters is that a whisker is a short circuit waiting for a gap. A bright tin plated lead has grown a 0.635 mm whisker in about four months at room temperature, and longer ones are reported in field environments. On a shield, a connector shell or a tin plated can, that length is the same order as the clearance to the adjacent pad, the chassis or the next shield finger. Nothing in the electrical test of a new assembly finds it, because the whisker has not grown yet.

Where the stress comes from

Tin whiskers are a stress relief mechanism. Tin plated onto copper forms an intermetallic layer of copper and tin at the interface, that layer grows at room temperature, and because it occupies less volume than the tin it consumes, the plating is left in compression. The tin relieves the compression by extruding a single crystal filament from a grain boundary. That is why the effect is worse on a copper base than on a nickel base: nickel does not form the same intermetallic at room temperature, so the driving stress stays low.

Two more factors decide the length. Mechanical stress raises it, which is why a press fit, a bent tab, a punched edge or a screw thread produces more whiskers than a flat surface. And time raises it, slowly, over years. The test method that quantifies it is JESD22-A121, which measures length aer defined storage and thermal cycling. Any incoming inspection that claims a whisker number without naming that method is quoting an opinion.

The three levers that actually work

A nickel barrier under the tin. Nickel blocks the copper tin intermetallic at the interface and removes most of the driving stress. The published historical maximum for matte tin over copper is 800 um; for matte tin over nickel it is under 100 um. That single change is worth roughly a factor of eight in the tail of the distribution, and it is a line on the drawing rather than a process hope.

Fusing or annealing after plating. Melting the tin, above 232 C, or holding it at 150 C for an hour within a day of plating, relaxes the stress before it can build a filament. Reflowed tin and hot dipped tin both sit under 100 um in the same historical data, and annealed matte tin over copper also comes in under 100 um, which shows how much of the effect is stress rather than the base metal.

Coating thickness, not coating presence. A NASA study on bright tin plated brass with a urethane coating found whiskers that had grown under the film and penetrated it after two years where the coating was 6.35 um. Where the coating was 50 um, whiskers formed underneath and none broke through. The film does not stop growth, it stops the short. Parylene C and silicone perform best in that work, acrylic the worst, and the reason is hardness: a whisker has to push through a film that resists it.

Finish or treatmentHistorical maximum lengthPractical note
Matte tin over copper800 umThe case to avoid on a fine pitch board
Matte tin over nickelUnder 100 umBarrier layer on the drawing
Reflowed or hot dipped tinUnder 100 umStress relaxed by melting
Annealed matte tin over copperUnder 100 um150 C for one hour within 24 hours of plating

Three habits that leave the risk in place

Counting a thin coating as mitigation. A 25 um acrylic coat over a tin plated shield is a moisture barrier, not a whisker barrier. The number that matters is the thickness over the tin, measured, at the thinnest place, which on a shield edge or a pressed finger is often half of what the drawing says.

Accepting the word matte without the layer under it. Matte is a grain structure, not a mitigation. Matte tin straight on copper is the 800 um case. Ask for the underlayer, the plating thickness, the fusing or annealing record and the date of that treatment relative to plating.

Assuming assembly reflow fuses the component plating. Reflow melts the paste and the component termination finish at the joint, not the plating on a shield, a can or a connector shell. Those surfaces go into the field exactly as they arrived from the plater.

What SUPERB controls

The board finish is specified per surface rather than per order: ENIG with nickel at 100 to 200 micro inches and gold at 2 to 3 micro inches, ENEPIG, immersion silver, immersion tin and lead free HASL are available, and a tin finished board is called out with its thickness so that the 6 um floor is not left to chance. Selective coating runs on Nordson ASYMTEK equipment with thickness verified on the first article, and the coating choice follows the hardness argument rather than the price list.

For the mechanical parts that whisker worst, the shield, the can and the connector shell, the drawing carries the underlayer and the heat treatment, and incoming inspection checks the finish rather than the label. Where a part cannot be changed, the gap is the fallback: keep tin plated surfaces at least 0.5 mm from an adjacent conductor, and coat the assembly.

The coating scope is on conformal coating selection, the board side including finish and Tg is on high reliability PCB, and the long life builds where this question arrives are on defense and radar PCBA.

What to send with the question

The finish on each tin plated surface with its thickness and underlayer, the date of plating and any heat treatment after it, the coating type and the measured thickness at the thinnest point, the gap between the tin surface and the nearest conductor, the storage and field temperature, the expected service life, and whether the part is bent, pressed or threaded in assembly. With those seven items the answer is a stack of finish, treatment and coating with a length behind it, instead of a promise.

Need a quote or a review of your build against these limits? pcba@superb-tech.com · WhatsApp +86 133 9608 1443