Copper Thickness for IoT PCB Designs: How to Choose
Copper thickness is one of the few PCB parameters that touches current capacity, thermal behaviour, signal quality and price at the same time, yet it is often fixed by habit rather than by calculation. On an IoT product the decision is sharper than on a rack mounted board, because the same small assembly has to run for years on a battery, radiate a radio signal and survive whatever enclosure the industrial designer chooses. Getting copper thickness right early avoids a costly respin later.
Why Copper Thickness Matters on IoT Boards
Copper foil on a board carries current and spreads heat at the same time, and on a compact IoT design both jobs matter. A sensor node that draws a few milliamps in sleep mode can still pull a short burst of hundreds of milliamps when the radio transmits, and every milliohm in that supply path turns the burst into a voltage drop and a local temperature rise. Thicker copper lowers resistance, and it also moves heat sideways away from the parts that generate it.
The other side of the balance is that heavier copper limits how finely the board can be etched. Fine lines and tight spacing become harder to produce as the foil gets thicker, because the etchant has to remove more material in every gap. That trade-off, rather than any single rule of thumb, is what decides the right copper thickness for most IoT PCB designs in practice.
Reading the Ounce Ratings
The industry expresses copper thickness in ounces, meaning the weight of foil per square foot. Half an ounce is about 17 micrometres, one ounce about 35, two ounces about 70, three ounces about 105 and four ounces about 140. More than eighty percent of consumer and IoT boards use one ounce, which is the default because it balances current capacity, fine line etching and price for the majority of designs.
Half an ounce suits dense, low power boards such as wearables, where line width and impedance control matter more than current. Two ounces and above appear in power conversion, LED drivers and motor control, where the current is high and the thermal load is real. Designers should also remember that inner layers are usually thinner than outer layers in the same stackup unless the fabricator is asked to do otherwise.
Current Carrying Capacity and Voltage Drop
Current carrying capacity is a thermal statement rather than an electrical one. A trace fails when its temperature rise exceeds what the laminate and the solder mask can tolerate, and the permitted rise depends on the copper cross section, the ambient temperature and how much of the heat can escape into the plane. Wider and thicker copper both reduce that rise, and a trace on an inner layer behaves differently from the same trace exposed to air.
Voltage drop is the electrical side of the same question and it is easy to neglect on battery powered products. A supply path with a few tens of milliohms costs almost nothing at one hundred milliamps and a great deal at two amps. Our guide to trace width and current calculation shows how to size a conductor against both limits, and why the result often justifies moving to heavier copper.
Heat Spreading in Small Enclosures
IoT products are usually sealed, which removes convection from the list of useful cooling routes. Heat then has to travel through copper and the enclosure wall, so copper thickness becomes part of the thermal design instead of a separate electrical decision. Heavier copper under a regulator or a power amplifier spreads heat into a larger area and lowers the hot spot without adding a single component.
Where a thermal pad and a via array carry heat to the opposite side of the board, the copper on that side decides how much area is available to dissipate it. A thick ground plane acts as a heat spreader, while a thin one concentrates the heat around the vias. This is one reason a two ounce board often runs measurably cooler than a one ounce board carrying the same circuit.
Impedance Control and Fine Line Density
Copper thickness affects controlled impedance because the trace geometry needed to hit a target impedance depends on the conductor cross section. On thin copper, a narrow trace over a close reference plane can reach fifty ohms in a practical width. On heavy copper the same impedance needs a wider trace and a thicker dielectric, which consumes routing area on a dense board.
Antenna feeds and RF matching networks are the places where this matters most, and there the copper thickness should be held constant across the whole RF section. Changing thickness changes impedance and loss, and a mixed stackup that uses one ounce for the radio and two ounces for the power section needs the transition modelled rather than assumed. Advice on dielectric constant and impedance applies directly here.
Manufacturing Limits and Cost
Every step up in copper thickness narrows the process window. Etching heavier foil produces more undercut, so minimum line width and spacing grow, and the fabricator may need to adjust plating and etching to hold tolerance. Very heavy copper can require stamped or etched features instead of ordinary chemical etching, which changes tooling and lead time as well as price.
Cost rises in steps rather than smoothly. A half ounce board is cheapest, one ounce is standard, and anything at or above two ounces attracts a premium because of the extra etching control and the difficulty of drilling heavy copper. The practical approach is to choose the lightest copper that satisfies the current and thermal requirements, then confirm that the fabricator can still hold the finest lines the layout needs.
Choosing Between Half an Ounce and Two Ounces
Start from the current the board must carry, add the temperature rise the enclosure allows, and check the resulting width against the space available. If the required traces fit comfortably, stay with one ounce, which keeps etching fine, impedance predictable and price low. Move to half an ounce only when routing density or impedance control forces it and the currents are genuinely small.
Go to two ounces when the currents are large, when the thermal path is short and the enclosure is sealed, or when the board has to survive repeated current surges. Document the reasoning in the design file, because the requirement is easy to lose during a later revision. The material requirements for PCB fabrication notes cover how copper weight interacts with laminate choice and stackup planning.
FAQ
Is heavier copper always better for thermal performance? It improves spreading, but it is not free. Heavier foil narrows the achievable line width, complicates impedance control and raises cost. Where heat is concentrated under one device, a via array and a larger copper area on a standard one ounce board often solve the problem at lower cost than moving the whole stackup to two ounces.
Can I mix copper weights in one stackup? Yes, and it is common. A board might use one ounce on the signal layers and two ounces on the power and ground layers. The fabricator needs to know before quoting, because mixed weights change lamination and drilling, and impedance targets on the signal layers have to be recalculated against the actual dielectric spacing.
How does copper thickness affect RF performance? Skin effect pushes high frequency current to the surface of the conductor, so beyond a certain point extra thickness does not reduce loss much. What thickness does affect is the trace geometry needed for a target impedance, and any variation across the RF section changes matching. Keep RF copper uniform and model the transition if the board uses mixed weights.
