Vibration damping methods for a PCB in an enclosure

Short answer: mount the board rigid and glue what still moves. On a PCB under 100 g, skip elastomer isolators and snap-in rubber guides; at their 15-30 Hz resonance they gain 5 or more and turn a 2 g input into a 10 g load. Adhesive standoffs at three or four corners hold the board itself, a polyurethane dot holds every part heavier than about 5 g, and anything that dissipates more than about 0.5 W gets a metal path to the wall instead of a bed of adhesive.

How do the damping methods compare side by side?

Ranked for a board under 100 g, adhesive standoffs and polyurethane dots come first, foam tape survives only as a 20 g indoor prototype fix, spring washers mind the screws, and elastomer isolators usually amplify the very shock you bought them to absorb.

The rows below are the five options that come up in every such build; the numbers assume a board of about 35 x 50 mm and 90 g, the size of a typical retrofit job. Rigid connections win because they keep the transmission ratio near 1: the board follows the case, so nothing gets amplified. Anything soft introduces a natural frequency, and a natural frequency inside the input band is gain you did not order.

Why do elastomer isolators make a light board worse?

A soft mount only isolates above 1.41 times its natural frequency, and the small rubber guides that suit a 100 g board resonate at 15-30 Hz, right inside the 1-30 Hz band a car body lives in, where their gain of 5 or more multiplies the input.

A mount's natural frequency follows f_n = (1/2π)√(k/m): soften the mount or add mass and the frequency drops, which is why a light board is hard to isolate. A 20 Hz mount corresponds to about 0.62 mm of static sag, so a genuinely soft mount under a 90 g board sags, wobbles and bottoms out.

Below the natural frequency the transmission ratio sits near 1; it peaks at the resonance gain Q, and it drops below 1 only past about 1.41 times the natural frequency. Rubber mounts run a damping ratio near 0.05-0.10, so Q = 1/(2ζ) lands at 5-10. For a 20 Hz guide, isolation starts at 1.41 × 20 ≈ 28 Hz. Road input in a car puts most of its energy between about 1 and 30 Hz (suspension and body modes), so the guide spends its working life at or near resonance, where a 2 g input arrives at the board as 10 g.

A shock is crueler still: the mount stores the blow and hands it back ringing at its own 20 Hz, so the board takes extra cycles on top of the peak. Hard-drive makers went through the same lesson: as drive assemblies lost mass, their elastomer shock mounts disappeared in favor of rigid fixing. On a board under 100 g there is too little mass to pull a mount's resonance below the input band, so rigid wins.

When do adhesive standoffs and polyurethane dots become the fix?

Fix the board with adhesive standoffs at three or four corners, then put a 4-6 mm polyurethane dot under every part heavier than about 5 g; that dot measures 28 mm² and carries roughly 28 N in shear at a conservative 1 N/mm².

Adhesive standoffs are molded bases a few millimetres tall, glued to the case and holding the board; they make a rigid mount that costs little and slices off with a knife when the layout changes. Power-supply builders have globbed polyurethane over capacitors, ferrites and relays for decades; the same dot holds a retrofit part floating loose in a case. A dot skins in 15-30 minutes and reaches full strength in 48-72 hours, so test after the cure. Solid polyurethane carries a dielectric constant of about 4 against air's 1, so keep dots off high-impedance sense nodes where a few picofarads shift a reading.

Worked example: three floating power parts. Take a 10 g resistor dissipating 0.4 W, a 5 g diode at 0.3 W and a 2 g TVS that clips 1 kW-class transients but averages milliwatts, all loose in a case that sees the 15 g, 11 ms half-sine shock of IEC 60068-2-27.

• Force on the resistor at 15 g: 0.010 kg × 15 × 9.81 m/s² ≈ 1.5 N.
• The dot's face bonds 3.14/4 × 6² ≈ 28 mm². At a deliberately conservative 1 N/mm² of lap shear — replace this with your adhesive's datasheet figure — one dot carries 28 N, a margin of about 19.
• Thermal check: a 2 mm bed at 0.25 W/m·K measures 283 K/W through the dot; the resistor sheds most of its 0.4 W from its bare body, and 0.4 W sits under the encapsulation threshold from the heat section.

A 3 mm dot would bond 7 mm² and carry 7 N, a margin of about 5 before aging, which is why the working answer is the 4-6 mm size. If the case belongs to a panel that has been running, lock it out and verify absence of voltage before gluing anything inside.

Where do tape, spring washers and doing nothing still work?

Tape, spring washers and even nothing extra all have their place: foam tape holds a prototype board up to about 20 g indoors, washers keep screws tight through vibration, and a rigidly screwed board in a stationary cabinet already moves with its case at a transmission ratio near 1.

High-bond foam tape grips hard at 23 °C and creeps in peel as the temperature climbs; a car cabin at 70-85 °C finds the weak spot within months, so tape belongs on the bench, not under the dashboard. Conical spring washers keep clamp load on a screw through vibration; they damp nothing, so pair them with a rigid standoff or clamp. A rigidly screwed board in a stationary cabinet needs nothing more, and production vehicle electronics are mounted the same way — rigid, designed and tested to the profile, not left floating. The one failure mode worth avoiding is a board carried only by its connector pins, where the solder joints take every shock the case sees.

How much heat does a bed of adhesive trap under a power part?

An adhesive bed insulates: solid polyurethane conducts about 0.25 W/m·K, ten times stagnant air at 0.025-0.027 W/m·K but roughly 800 times less than aluminum, so keep full encapsulation off any part dissipating more than about 0.5 W and bed only its base.

The measured values behind that: stagnant air conducts 0.025 W/m·K at 10 °C and 0.027 W/m·K at 40 °C, and solid polyurethane sits near 0.2-0.3 W/m·K in handbook tables. Full burial is what cooks a part: the 2 mm bed in the worked example measures 283 K/W, so half a watt pushed entirely through it adds 141 K. Three rules follow:

• Encapsulate only parts with near-zero average dissipation; the TVS in the example clips kilowatt transients and averages milliwatts, so a bed cannot heat it.
• Below the threshold, dot the base and leave the body bare; the exposed faces shed the heat.
• Above it, bolt the part through a metal standoff or clip it to the wall; a conduction-cooled sealed enclosure is that idea taken to its limit.

A 50 µm conformal film adds only about 2.5 K per watt through each square centimetre, a few degrees at board-level fluxes, so coating stays on boards whose hotspots sit at ordinary dissipation; the field practice for conformal coating on outdoor boards covers the condensation problem a film actually solves. In a sealed box that must shed its own 10-15 W through its walls, the small enclosure heat dissipation arithmetic tells you how much glue coverage the thermal budget can afford.

What does a car installation demand that an indoor cabinet does not?

A passenger car is the harshest site on the list: design the mount to ISO 16750-3 random vibration profiles and the 15 g, 11 ms shock of IEC 60068-2-27, expect 70 °C cabin air in full sun, and keep every resonance out of the 1-30 Hz road band.

ISO 16750-3 is the road-vehicle standard that supplies random vibration profiles for electronics by mounting location, passenger cars included. Design the mount so the ratio stays near 1 and no rubber lands in the road band, and pick adhesives for the cabin: 70-85 °C in full sun means a polyurethane grade rated 90 °C or better.

Vehicle-mounted builds outside cars get specified the same way; the tractor display enclosure for vibration, rain and dust walks through one example. When the whole box lives on a shaking frame, fix the structure first — the enclosure mounting guide covers anchoring the box itself.

What should you put on the drawing or the order?

Put the mounting scheme on the drawing itself: 3.2 mm standoff holes at the board corners, a polyurethane dot drawn under each part above 5 g, a bolted metal path for anything over the thermal threshold, and a 48-72 h cure noted ahead of the first road test.

A custom enclosure takes the hole pattern you send, so treat the mounting as a feature, not an afterthought. Choosing the case before board layout saves re-cutting later, and the send-with-the-drawing list is simple: board outline, the 3.2 mm hole pattern, and the heavy parts so the dots land before wiring runs over them. For small runs we build one-off stainless boxes to your drawing — custom enclosures for small PCB runs shows the route, delivered in 10 business days from drawing approval. The same standoff and cutout discipline carries over to an LCD window with its standoffs and cutouts.

Catalog dimensions to compare

Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.

Questions