PCB Thickness Guide: Standard 1.6mm and When to Go Thinner or Thicker

Open any PCB ordering form and the thickness field defaults to 1.6 mm. Most designers click past it without a thought, and most of the time that is the right call. But "most of the time" hides a meaningful minority of projects where 1.6 mm is wrong — wearables that need 0.6 mm to fit the housing, backplanes that need 3.2 mm for press-fit connectors, power converters that need 2.4 mm for heavy copper layers, edge-connector cards where 1.6 mm ±10% will not seat in the slot.

This guide explains where the 1.6 mm convention came from, what the full range of practical thicknesses looks like, when to deviate in either direction, and the tolerance, warpage, and cost consequences of doing so. Linked Electronics Co., Limited fabricates boards from 0.4 mm to 6.0 mm finished thickness as standard capability, with thinner and thicker available on engineering review.

Why 1.6 mm Became the Standard

The 1.6 mm convention (0.062 inch — one sixteenth of an inch) dates to the early days of laminate manufacturing, when board material was supplied in imperial fractional-inch sheets and 1/16" was the workhorse thickness for the phenolic and early glass-epoxy boards of the era. The dimension carried forward into FR4, into connector designs (card-edge connectors are overwhelmingly designed around 1.6 mm), into test fixtures, and into every default value in every CAD tool and quoting system since.

The convention persists because 1.6 mm genuinely sits in a sweet spot:

• Stiff enough to survive assembly handling, conveyor transport, and connector insertion without flexing.
• Thin enough that drilling and plating remain easy — a 0.3 mm via in a 1.6 mm board is a comfortable 5.3:1 aspect ratio.
• Compatible with the vast majority of card-edge connectors, standoffs, enclosures, and fixtures designed over decades.
• Layer-flexible — 1.6 mm accommodates anything from 1 to roughly 10 layers with sensible dielectric thicknesses.

If your project has no constraint pushing you elsewhere, 1.6 mm is correct precisely because everything else in the ecosystem assumes it.

The Standard Thickness Ladder

Fabricators stock core and prepreg materials around a ladder of common finished thicknesses. Choosing one of these values keeps cost down and lead time short; choosing a value between rungs forces a custom stackup.

When to Go Thinner

Mechanical Fit

The dominant reason. Wearables, earbuds, smart cards, camera modules, and anything stacked inside a tight enclosure may simply not have 1.6 mm of vertical budget. Thin boards exist because products got thin first.

Flexibility Requirements

A 0.4 mm FR4 board has measurable compliance — some designs exploit this deliberately for snap-fit assemblies or slight curvature. Genuine repeated flexing, however, calls for polyimide flex or rigid-flex construction, not thin FR4. FR4 tolerates one gentle bend at install, not a lifetime of cycles.

Weight

Aerospace, drones, and portable devices count grams. Going from 1.6 mm to 0.8 mm halves the laminate weight per unit area. Across a large board or a fleet of drones, this is a real number.

Thermal Transfer Through the Board

When heat must pass through the board to a chassis or heatsink on the far side, thinner dielectric means lower thermal resistance. A 0.8 mm board conducts heat through its thickness twice as well as a 1.6 mm board of the same construction.

The Costs of Thin

• Handling and warpage — thin boards flex during paste printing and reflow, causing print misregistration and tombstoning. Assembly typically requires pallets or stiffener fixtures below 0.8 mm, which adds NRE and per-unit handling cost.
• Layer count limits — each copper layer plus its dielectric consumes thickness budget. Fitting 8 layers into 0.8 mm forces very thin dielectrics, which tightens impedance tolerance and raises cost.
• Depanelisation stress — V-scoring weakens thin boards severely; routed tabs are usually required, which uses panel space less efficiently.
• Connector compatibility — card-edge connectors designed for 1.6 mm will not grip a 0.8 mm board. Check every mating part.

When to Go Thicker

Press-Fit Connectors and Mechanical Retention

Telecom and backplane connectors with press-fit tails need barrel engagement length — the connector datasheet specifies a minimum board thickness, commonly 2.4 mm or more for large backplane connectors. The board thickness is dictated by the connector, not the electronics.

Heavy Copper and High Layer Count

A 6 oz copper layer is roughly 210 microns thick. Stack four of them with adequate dielectric and the board cannot physically be 1.6 mm. Similarly, a 20-layer board with sensible dielectric spacing lands at 2.4–3.2 mm by arithmetic alone. Thickness here is an output of the stackup, not an input.

Stiffness for Large Boards

Deflection scales with the cube of thickness. A 400 mm industrial board at 1.6 mm sags under its own component weight and vibrates in service; the same board at 2.4 mm is 3.4 times stiffer. Large-format boards in vibration environments (rail, automotive, industrial machinery) commonly step up thickness purely for rigidity.

The Costs of Thick

• Aspect ratio — this is the big one. A 0.3 mm via in a 3.2 mm board is a 10.7:1 aspect ratio, at the edge of routine plating capability. Thick boards force larger vias or accept the cost and yield consequences of high-aspect-ratio plating. Our aspect ratio guide covers this in depth.
• Drilling time and bit wear — drill cycle time scales with depth, and bit breakage rates rise in thick material.
• Material cost — more laminate, more prepreg, longer lamination cycles.
• Thermal mass at assembly — thick boards absorb reflow heat, requiring longer profiles and putting more thermal stress on components.
• Depanelisation — V-scoring becomes impractical above roughly 3.2 mm; routing is slower in thick material.

Thickness Tolerance: The Detail That Bites

Finished board thickness is not exact. The industry-standard tolerance per IPC-6012 is ±10% of nominal for boards 1.0 mm and above — meaning a 1.6 mm board legitimately arrives anywhere from 1.44 mm to 1.76 mm. The variation comes from prepreg resin flow during lamination, copper distribution across the panel, and plating build-up.

For most applications this is invisible. It bites in three places:

• Card-edge connectors — PCIe and similar gold-finger interfaces specify the finger-area thickness more tightly (typically 1.57 ±0.13 mm). State the connector requirement explicitly on the drawing.
• Press-fit zones — barrel engagement depends on actual thickness; specify tolerance at the connector region.
• Impedance — thickness variation is dielectric variation, which moves impedance. Controlled-impedance boards manage this through stackup control rather than overall thickness tolerance, but the two are linked.

Tighter overall tolerance (±5%) is available at a cost premium — specify it only where a mating part genuinely requires it. Linked Electronics Co., Limited holds ±10% standard and quotes tighter tolerance per project.

How to Decide: A Short Checklist

1. Does a connector, slot, or enclosure dictate thickness? If yes, that constraint wins. Check every mating part datasheet.
2. Does the stackup (layer count, copper weights, impedance dielectrics) add up to more than the target? If yes, thickness follows the stackup.
3. Is the board large or in a vibration environment? Step up for stiffness.
4. Is the product height-constrained or weight-constrained? Step down, and budget for assembly fixturing.
5. None of the above? Use 1.6 mm and move on.

And one cross-check before finalising: divide your target thickness by your smallest via diameter. If the result exceeds 10:1, revisit either number before sending the design out — or expect your fabricator to raise it at DFM review, as Linked Electronics Co., Limited does on every incoming design.

Frequently Asked Questions

Is 1.6 mm thickness measured with or without surface finish and solder mask?

The convention is finished board thickness over the laminate and copper, measured at the board surface excluding solder mask in most fabrication standards — but practices vary, and for tight-tolerance applications you should state explicitly on the drawing whether the dimension applies over mask, over copper, or at the gold fingers. Ambiguity here is a classic source of connector-fit disputes.

Can I get 1.2 mm or 1.4 mm boards?

Yes — 1.2 mm is a semi-standard rung stocked by most fabricators including Linked Electronics Co., Limited, common for memory modules and compact consumer products. 1.4 mm is less common and may price as a custom stackup. When the exact value is flexible, choosing the nearest standard rung saves cost.

What is the thinnest rigid PCB that can be made?

Production rigid boards go down to approximately 0.2–0.3 mm for simple 1–2 layer constructions (smart cards are the canonical example). Below that, the material is effectively a flex circuit and polyimide construction takes over. At Linked Electronics Co., Limited, 0.4 mm is the standard thin capability with thinner builds reviewed per project.

Does board thickness affect controlled impedance?

Not directly — impedance is set by the trace geometry and the dielectric thickness to the adjacent reference plane, not by overall board thickness. But total thickness constrains how the stackup can be arranged, which indirectly constrains the impedance options. A 0.8 mm 6-layer board forces thin dielectrics everywhere; a 1.6 mm 6-layer board gives the stackup room to breathe.

Why did my thick board quote come back with a larger minimum via?

Aspect ratio. The fabricator divided your board thickness by your specified via diameter and the result exceeded their reliable plating capability. The options are a larger via, a thinner board, or a price/leadtime premium for high-aspect-ratio plating. See our dedicated aspect ratio article for the full mechanics.

Does thickness affect which IPC class my board can meet?

No thickness is excluded from Class 2 or Class 3, but thick boards make Class 3 harder to achieve in one specific way: plating thickness minimums must be met at the centre of deep via barrels, which is exactly where high-aspect-ratio plating is thinnest. A Class 3 specification on a 4.0 mm board with small vias is a demanding combination that should be discussed with the fabricator before committing the design.

Can different areas of one board have different thicknesses?

Yes — this is called a stepped or cavity board, produced by controlled-depth routing or bonded sub-stacks. Common for boards that must be thin under one component (an RF shield, a display) and thick elsewhere. It is specialty work with real cost, but Linked Electronics Co., Limited supports cavity and stepped constructions on engineering review. Rigid-flex is the other common answer when sections of the product need to be thin and bendable.

Why Choose Linked Electronics Co., Limited Across the Thickness Range

Linked Electronics Co., Limited fabricates rigid PCBs from 0.4 mm through 6.0 mm finished thickness as standard capability in our Suining facility — spanning ultra-thin consumer builds, the 1.6 mm mainstream, heavy copper power boards at 2.4 mm and above, and telecom backplanes to 6.0 mm with the pulse-reverse plating capability that thick-board aspect ratios demand. Thickness tolerance is ±10% standard with ±5% available, and gold-finger areas are held to connector-specific tolerances on request.

Every incoming design is checked at DFM review for the thickness-related failure modes covered in this article: aspect ratio against our plating capability, press-fit zone thickness against connector requirements, depanelisation method against board thickness, and stackup arithmetic against the target dimension. If the numbers do not close, we tell you before fabrication — not after.

Whatever thickness your project needs — or if you are unsure whether to deviate from 1.6 mm at all — send your stackup, Gerber files, and mechanical constraints to [email protected]. Linked Electronics Co., Limited will return DFM feedback and a manufacturable quote within one to two business days.

← Back to Blog