Thermal Conductivity vs Thermal Resistance in PCB Design: What Engineers Need to Know

If there is one pair of terms that causes more confusion in PCB thermal design than any other, it is thermal conductivity and thermal resistance. These two properties are mathematically related — one is derived from the other — yet they describe fundamentally different things, get used at different stages of the design process, and are extracted from completely different sources. An engineer who conflates them will either under-specify substrate materials, misread component datasheets, or produce junction temperature calculations that are off by a factor of ten or more.

This guide untangles the two concepts from the ground up, shows how each one enters a real design calculation, and explains where you find them — in material datasheets, component datasheets, and IMS product sheets like the Bergquist Thermal Clad series discussed in detail on the Bergquist PCB reference page. By the time you finish reading, the relationship between thermal conductivity and thermal resistance in PCB design should feel as intuitive as the relationship between resistivity and resistance in circuit analysis.

---

The Electrical Analogy: Why It Actually Works for Thermal Design

The single most effective way to understand thermal conductivity vs thermal resistance in PCB design is to map them to their electrical equivalents. The analogy is not a loose metaphor — it is a precise mathematical parallel used in every thermal simulation tool in the industry.

In electronics, current flows through a conductor under a voltage difference. The material’s ability to oppose current flow is described by resistivity (Ω·m) — a material constant independent of geometry. When you apply geometry (length and cross-sectional area), you get resistance (Ω) — the actual opposition to current flow in a specific conductor.

The thermal equivalent is exact. Heat flows through a material under a temperature difference. The material’s ability to oppose heat flow is described by thermal resistivity — or equivalently, its ability to conduct heat is described by thermal conductivity (W/m·K), which is just the inverse of thermal resistivity. When you apply geometry (thickness and area), you get thermal resistance (°C/W or K/W) — the actual opposition to heat flow through a specific structure.

Table 1: Electrical vs Thermal Analogy in PCB Design

Electrical Domain	Symbol	Unit	Thermal Equivalent	Symbol	Unit
Voltage difference	ΔV	V	Temperature difference	ΔT	°C or K
Current	I	A	Heat flow (power)	Q	W
Electrical resistivity	ρ	Ω·m	Thermal resistivity	—	m·K/W
Electrical conductivity	σ	1/Ω·m	Thermal conductivity	k	W/m·K
Electrical resistance	R	Ω	Thermal resistance	R_θ	°C/W or K/W
Resistors in series	R_total = R₁ + R₂	Ω	Thermal resistors in series	R_total = R₁ + R₂	°C/W

The parallel is so clean that the thermal resistance network used to calculate junction temperatures — with resistors representing the junction-to-case path, the PCB dielectric, the substrate, the TIM, and the heatsink — is drawn and solved exactly like a DC resistor network.

---

Thermal Conductivity: What It Is and What It Is Not

The Definition

Thermal conductivity (symbol k, units W/m·K) is a material property that describes how readily a unit volume of material transfers heat. A high k value means heat flows easily through the material; a low k value means the material is a thermal insulator. It is a bulk material constant — independent of the shape, thickness, or area of the sample, just as electrical resistivity is independent of trace width or length.

Copper has a thermal conductivity of approximately 385–400 W/m·K, which is why copper traces and planes are so effective at spreading heat laterally across a PCB. FR-4 laminate has a thermal conductivity of approximately 0.25–0.35 W/m·K in the Z-axis (through the board thickness), making it an effective thermal insulator in the direction that matters most for component cooling. Aluminium substrate in an IMS PCB sits at approximately 150–200 W/m·K depending on alloy.

Where Thermal Conductivity Appears in PCB Design

Thermal conductivity is the right value to use when comparing materials before you have a geometry. When you are deciding between an aluminium MCPCB and an FR-4 board, comparing their substrate k values gives you a valid order-of-magnitude sense of the thermal advantage. When a dielectric supplier lists 3.0 W/m·K for Bergquist HPL-03015 versus 1.0 W/m·K for MP-06503, that ratio tells you the HPL grade conducts heat three times as well through a unit thickness of material.

What thermal conductivity does not give you directly is the temperature rise across a specific structure at a specific power level. For that, you need thermal resistance.

Table 2: Thermal Conductivity of Common PCB Materials

Material	Thermal Conductivity k (W/m·K)	Notes
Copper (bulk)	385–400	Traces, planes, vias — lateral heat spreading
Aluminium 5052 (substrate)	~138	Common MCPCB base alloy
Aluminium 6061 (substrate)	~167	Higher strength alloy
Aluminium 1100 (substrate)	~222	High thermal, lower strength
Copper (substrate, OFC)	~385	Premium MCPCB, EV inverter applications
FR-4 (Z-axis)	0.25–0.35	Low; why FR-4 fails in high-power design
Bergquist HPL-03015 dielectric	3.0	Thinnest Thermal Clad grade
Bergquist HT-04503 / HT-07006 dielectric	2.2	High-temperature IMS grades
Bergquist MP-06503 dielectric	1.0	General-purpose IMS grade
Ceramic (Al₂O₃)	24–30	High-end power module substrate
Ceramic (AlN)	170–200	Top-tier power electronics substrate
Air	0.026	Worst-case: what a void becomes

The air value at the bottom of that table deserves particular attention. Any void in a PCB dielectric layer — a bubble trapped during lamination — has a thermal conductivity of 0.026 W/m·K. That is an order of magnitude worse than FR-4. A void under a high-power component pad behaves as a local thermal insulator, creating a hot spot regardless of how good the surrounding substrate is. This is why dielectric void inspection is a quality-critical process step in MCPCB fabrication.

---

Thermal Resistance: The Design Calculation Variable

The Definition and Formula

Thermal resistance (symbol R_θ, units °C/W or K/W) is the temperature rise per watt of heat flowing through a specific structure. The fundamental formula is:

R_θ = ΔT / P = t / (k × A)

Where ΔT is the temperature difference across the structure (°C), P is the power dissipated (W), t is the thickness of the material (m), k is the thermal conductivity (W/m·K), and A is the cross-sectional area through which heat flows (m²).

This formula is the direct analogue of Ohm’s law. ΔT is voltage. P (heat flow) is current. R_θ is resistance. The analogy is exact.

From Conductivity to Resistance: A Worked Example

Consider the dielectric layer in a Bergquist HPL-03015 IMS PCB: k = 3.0 W/m·K, thickness t = 38 µm = 38 × 10⁻⁶ m. A power device with a 5 mm × 5 mm thermal pad is mounted on this board, so the effective heat flow area A = 25 mm² = 25 × 10⁻⁶ m².

R_θ (dielectric) = t / (k × A) = (38 × 10⁻⁶) / (3.0 × 25 × 10⁻⁶) = 0.507 °C/W

At 3 W dissipation, the temperature rise across the HPL dielectric alone is 3 × 0.507 = 1.5 °C. This is negligible compared to the heatsink and TIM resistances in a typical system — which is exactly the point of using HPL-03015 for high-power LED and power electronics designs.

Now run the same calculation for standard FR-4 at 1.6 mm thickness: k = 0.3 W/m·K, same 5 × 5 mm area.

R_θ (FR-4) = (1.6 × 10⁻³) / (0.3 × 25 × 10⁻⁶) = 213 °C/W

At 3 W, the temperature rise across the FR-4 is over 640 °C. That is not a calculation error — it is why FR-4 is completely unusable as the thermal path for a 3 W surface-mount device without an elaborate thermal via array to bypass the laminate.

Specific Thermal Resistance vs Absolute Thermal Resistance

These two versions of thermal resistance appear in different contexts and are easy to confuse. Specific thermal resistance (sometimes called thermal resistivity, or reported as °C·in²/W or °C·cm²/W in IMS datasheets) is an area-normalised material property — it is the thermal resistance per unit area. Absolute thermal resistance (°C/W or K/W) is the actual resistance of a specific, sized structure.

When you read 0.02 °C·in²/W on a Bergquist HPL-03015 datasheet, that is the specific thermal resistance of the dielectric at 38 µm thickness. To get the absolute thermal resistance for your design, you divide by the area of your component pad. A 5 × 5 mm (0.039 in²) pad gives: R_θ = 0.02 / 0.039 = 0.51 °C/W — consistent with the calculation above.

Component datasheets use absolute thermal resistance: θJC (junction to case, °C/W), θJA (junction to ambient, °C/W), θCA (case to ambient, °C/W). These are properties of a specific package mounted on a specific test board. When you see θJA on an IC datasheet, it is only valid for the JEDEC standardised test PCB used during measurement — it cannot be transferred directly to your design.

---

The Complete Junction Temperature Calculation: Putting It Together

The Thermal Resistance Chain

Real PCB thermal design requires tracking the entire heat path from the component junction to ambient as a series resistance network — exactly like series resistors in a circuit. The total temperature rise from junction to ambient at power P is:

T_junction = T_ambient + P × (R_θJC + R_θsolder + R_θdielectric + R_θsubstrate + R_θTIM + R_θheatsink)

Each resistance term represents one thermal interface or material in the heat path. The total junction temperature is T_ambient plus P times the sum of all those resistances.

Table 3: Typical Thermal Resistance Values in an IMS PCB System

Thermal Path Element	Symbol	Typical Value (°C/W)	Where to Find It
Junction to case (component)	R_θJC	1–20 °C/W	Component datasheet
Solder joint (SAC305, large thermal pad)	R_θsolder	0.1–0.5 °C/W	Application notes
IMS dielectric (HPL-03015, 5×5mm pad, 3W)	R_θdielectric	~0.5 °C/W	Calculated from IMS datasheet
IMS dielectric (FR-4 equivalent)	R_θdielectric	200+ °C/W	Calculated — shows why FR-4 fails
Aluminium substrate (2mm, 5×5mm)	R_θsubstrate	~0.1 °C/W	Usually negligible
Thermal interface material (phase-change)	R_θTIM	0.5–2 °C/W	TIM datasheet
Natural convection heatsink	R_θheatsink	3–15 °C/W	Heatsink datasheet

The heatsink is usually the dominant resistance in the chain for well-designed IMS boards. That observation contains a practical design insight: once you have selected an IMS dielectric that drops the dielectric thermal resistance to below 1 °C/W, further improvement in the dielectric grade has diminishing returns compared to improving the heatsink or TIM selection.

Table 4: Junction Temperature Comparison — IMS vs FR-4 for a 3 W LED

Assumptions: 3 W LED, T_ambient = 25 °C, R_θJC = 8 °C/W, natural convection heatsink R_θ = 7 °C/W, phase-change TIM R_θ = 1.0 °C/W.

PCB Type	R_θ dielectric	R_θ total	T_junction
Standard FR-4 (1.6 mm)	~213 °C/W	~230 °C/W	Off the chart
Aluminium IMS + MP-06503	~0.8 °C/W	~17 °C/W	76 °C
Aluminium IMS + HPL-03015	~0.5 °C/W	~16.5 °C/W	74.5 °C
Aluminium IMS + HPL-03015 + better heatsink (3 °C/W)	~0.5 °C/W	~12.5 °C/W	62.5 °C

The numbers show that switching from MP-06503 to HPL-03015 saves about 1.5 °C at this power level and heatsink combination. Improving the heatsink from 7 °C/W to 3 °C/W saves 12 °C. When the heatsink dominates the chain, optimising the dielectric grade is the wrong leverage point. When you cannot improve the heatsink — because of product geometry, cost, or IP67 sealing — then the dielectric selection becomes the only lever available.

---

How FR-4 Thermal Vias Fit Into This Picture

Thermal vias in FR-4 boards are an attempt to short-circuit the high Z-axis thermal resistance of the laminate. A plated copper via with 25 µm wall plating has a thermal conductivity of ~400 W/m·K along its axis — far better than the FR-4 surrounding it. However, the via represents a small fraction of the total board cross-section. The effective thermal conductivity of a via field through FR-4 rarely exceeds 1–2 W/m·K across the full pad area, even with aggressive via density.

The calculation for a single thermal via thermal resistance is R_θ = L / (k × A_copper), where A_copper is the annular copper area of the via wall, not the full via diameter. A single 0.3 mm finished via through 1.6 mm FR-4 with 25 µm copper plating has a thermal resistance of approximately 193 °C/W on its own. You need a dense array of such vias beneath every thermal pad to achieve useful cumulative resistance reduction — and even a well-designed array of 20–30 vias under a power device might achieve a combined thermal resistance of 8–10 °C/W through the PCB dielectric, compared to 0.5–1 °C/W for a high-grade IMS dielectric at the same component footprint.

This is why IMS PCBs do not need thermal vias. The dielectric handles the vertical heat path directly. Thermal vias are a workaround for FR-4’s fundamental thermal limitation — IMS eliminates the limitation at the substrate level.

---

Anisotropic Thermal Conductivity in PCBs: The Axis Problem

One subtlety that catches engineers who are new to PCB thermal design is that FR-4 and most PCB laminates are anisotropic — their thermal conductivity is different in different directions. The Z-axis (through the board thickness) has a thermal conductivity of approximately 0.25–0.35 W/m·K for standard FR-4. The in-plane (XY) direction has an effective thermal conductivity closer to 1–2 W/m·K for a multilayer board with copper pours, because the copper layers conduct heat laterally far better than the FR-4 between them.

In PCB thermal modelling, the Z-axis conductivity is what matters for heat extraction from component to heatsink. The XY conductivity is what determines how heat spreads before it reaches the extraction point. A large copper pour under a hot component improves XY spreading and reduces the effective thermal resistance to the heatsink by presenting a larger area to the Z-axis path — even if the Z-axis conductivity of the dielectric has not changed.

For IMS PCBs, the metal base itself provides excellent in-plane spreading at ~150–400 W/m·K. Heat from a small component pad spreads laterally through the aluminium base before flowing into the heatsink. This spreading effect means the effective thermal resistance of the system is lower than a simple one-dimensional R = t/(k×A) calculation would suggest — an IMS board with a 50 mm × 50 mm aluminium base conducts heat from a 5 mm × 5 mm component pad not just through 25 mm² but through a much larger effective area at the heatsink interface.

---

Practical Design Decision Guide

When Thermal Conductivity Is the Right Metric

Use thermal conductivity to compare and select materials before you have finalised component footprints. Compare FR-4 vs IMS dielectrics, compare HPL vs HT vs MP Bergquist grades, or compare aluminium vs copper base metal. Conductivity comparisons give you the relative performance ranking without needing component geometry.

When Thermal Resistance Is the Right Metric

Use thermal resistance to calculate actual junction temperatures once you have component data. Extract R_θJC from the component datasheet, calculate R_θdielectric from the IMS dielectric spec and your pad area, add TIM and heatsink resistances, and solve the series network. This tells you whether your design keeps junction temperatures within specification at maximum operating power and ambient temperature.

Table 5: When to Use Conductivity vs Resistance in PCB Thermal Design

Design Stage	Use	Metric	Source
Substrate material selection	Conductivity	W/m·K	Material datasheets
IMS dielectric grade comparison	Specific thermal resistance	°C·in²/W	IMS product datasheet
Junction temperature calculation	Absolute thermal resistance	°C/W	Component datasheet + calculation
Via array design	Thermal resistance	°C/W per via, total	Calculation
Heatsink selection	Absolute thermal resistance	°C/W	Heatsink datasheet
TIM selection	Specific thermal resistance	°C·cm²/W	TIM datasheet

---

Useful Resources for PCB Thermal Design

Resource	What It Covers	Link
Bergquist Thermal Clad Selection Guide	Dielectric thermal resistance data, design guidelines, base metal selection	Digikey PDF
Texas Instruments TIDA030 — FR-4 vs IMS Thermal	Measured R_θJA comparison: FR-4 (61.56 °C/W) vs IMS (39.1 °C/W)	TI.com PDF
Fineline Global — PCB Temperature Management PDF	Thermal resistance formulas, via array calculations, LED module examples	Fineline PDF
Analog Devices — Thermal Resistance Glossary	θJA, θJC, θCA definitions with component datasheet context	Analog.com
PCBCart — PCB Thermal Dissipation Design	PCB thermal resistance calculation formulas, multilayer board analysis	PCBCart
Mbedded.ninja — Thermal Design for PCBs	Via thermal resistance calculation, TIM selection, practical layout notes	Mbedded.ninja
IPC-2221B Generic PCB Design Standard	Current capacity, thermal relief design, trace width guidelines	IPC.org

---

5 FAQs: Thermal Conductivity vs Thermal Resistance in PCB Design

Q1: Can I use W/m·K from a dielectric datasheet directly in my junction temperature calculation?

Not directly. Thermal conductivity in W/m·K is a material constant — it does not include the geometry of your specific dielectric thickness or component footprint. To calculate junction temperature, you need thermal resistance in °C/W, which you derive from the conductivity value using R_θ = t / (k × A). Many IMS dielectric datasheets, including the Bergquist Thermal Clad series, save you this step by listing specific thermal resistance in °C·in²/W alongside the W/m·K value. If only W/m·K is listed, apply the formula above with your actual dielectric thickness and the component’s thermal pad area.

Q2: Component datasheets list θJA and θJC. Which one should I use for PCB thermal design?

Almost always θJC (junction to case), not θJA. θJA is the total thermal resistance from junction to ambient measured on a JEDEC standardised test board under specific conditions. It is not transferable to your design because your PCB, stackup, and thermal management will be different. θJC represents only the thermal resistance from the silicon junction to the package case — the point where your PCB takes over. You add your own PCB, TIM, and heatsink resistances to θJC to get the total thermal resistance from junction to ambient in your actual design. Using θJA from the datasheet as if it applies to your design will produce misleading junction temperature estimates.

Q3: If I double the thermal conductivity of my IMS dielectric, does the junction temperature drop by half?

No, because the dielectric is only one element in the series thermal resistance chain. Halving the dielectric thermal resistance reduces the total R_θ by the fraction of the chain that the dielectric represents. If the heatsink contributes 7 °C/W and the dielectric contributes 0.5 °C/W in a total chain of 16 °C/W, halving the dielectric resistance saves 0.25 °C/W — about 1.5 °C at 5 W dissipation. Doubling the dielectric conductivity has meaningful impact only when the dielectric thermal resistance represents a significant fraction of the total chain — which is the case on cost-optimised designs using lower-grade dielectrics, or on very large-area LED panels where the heatsink resistance per unit area is low.

Q4: My thermal simulation software asks for thermal conductivity but my IMS dielectric datasheet only lists thermal resistance in °C·in²/W. How do I convert?

Rearrange the relationship: k = t / (R_θ_specific × unit_area). For Bergquist HPL-03015 with specific thermal resistance of 0.02 °C·in²/W at 38 µm thickness: convert 0.02 °C·in²/W to SI units — 1 in² = 6.452 × 10⁻⁴ m², so 0.02 °C·in²/W = 0.02 × 6.452 × 10⁻⁴ °C·m²/W = 1.29 × 10⁻⁵ °C·m²/W. Then k = t / R_θ_specific = (38 × 10⁻⁶) / (1.29 × 10⁻⁵) = 2.95 W/m·K — consistent with the 3.0 W/m·K listed directly. Most simulation tools accept W/m·K as the material input; perform this conversion and enter the conductivity value directly.

Q5: How many thermal vias do I need under a 10 W component on FR-4 to match IMS performance?

The short answer is that you cannot fully match IMS performance with thermal vias on FR-4, regardless of via density. A practical thermal via array of 20 × 0.3 mm finished vias at 0.6 mm pitch beneath a 5 × 5 mm thermal pad might achieve a combined PCB dielectric thermal resistance of around 8–10 °C/W. An HPL-03015 IMS board at the same footprint achieves 0.5 °C/W. The 16–20× difference in dielectric thermal resistance cannot be closed with more vias — you run out of physical space in the pad footprint before the via array delivers IMS-comparable performance. At 10 W dissipation, the via array contributes 80–100 °C of temperature rise across the PCB dielectric alone. The IMS dielectric contributes 5 °C. At 10 W with a 25 °C ambient and a 150 °C junction limit, you have 125 °C of temperature budget across the entire thermal chain — that budget is consumed by the via-array FR-4 PCB before the heatsink even enters the calculation.

---

Summary: Thermal Conductivity Sets the Material Ranking, Thermal Resistance Decides the Junction Temperature

The practical takeaway from this entire discussion fits in two sentences. Thermal conductivity in W/m·K tells you how one material compares to another — it is the right tool for substrate selection. Thermal resistance in °C/W tells you whether your specific design, with its specific component footprints and power levels, stays within junction temperature limits — it is the right tool for thermal calculations.

Every other confusion in PCB thermal design — θJA vs θJC, specific vs absolute thermal resistance, anisotropic conductivity, why thermal vias have limits — reduces to this single conceptual distinction. Get comfortable moving between W/m·K and °C/W using the formula R_θ = t/(k×A), and the rest of PCB thermal design falls into place.

For material-specific thermal resistance data on Bergquist Thermal Clad IMS dielectrics and guidance on applying these values to LED and power electronics designs, visit the Bergquist PCB reference page.