PCB Thermal Analysis Back
Thermal Analysis

PCB Thermal Analysis · Junction Temperature Calculator

Real-time PCB junction temperature calculation via thermal resistance network Tj = Ta + P × (θjc+θcs+θsa). Supports BGA, QFP, TO-220, and MOSFET packages with combined thermal via and heatsink design evaluation.

Parameter Settings
Component Type
Power Dissipation P
W
θjc (Junction–Case)
°C/W
θcs (Case–Spreader)
°C/W
θsa (Heatsink–Ambient)
°C/W
Ambient Temperature Ta
°C
Thermal Via Count N
pcs
Board Thickness t
mm
Airflow Velocity v (forced)
forced
0 = natural convection
Results
Tj Junction Temperature [°C]
Tc Case Temperature [°C]
PCB Surface Temperature [°C]
ΔT = Tj − Ta [K]
θja Total [K/W]
Status (125 °C limit)
Visualization
Tj
Theory & Key Formulas

The thermal path of an electronic component from junction to ambient is treated as a series connection:

$$T_j = T_a + P \times \theta_{ja}$$ $$\theta_{ja}= \theta_{jc}+ \theta_{cs}+ \theta_{sa}$$

Thermal via parallel resistance: $\theta_{via}= \dfrac{t}{N \cdot k_{Cu} \cdot \pi r^2}$ (r=0.15 mm, k=385 W/m·K)

Natural convection heatsink: $\theta_{sa}\approx \dfrac{1}{h_c \cdot A_s}$ (h_c ≈ 5–15 W/m²K)

Forced convection: $h_c \approx h_0 + c \cdot v^{0.6}$ (velocity power law)

What is PCB Junction Temperature?

🙋
What exactly is "junction temperature," and why is it so important for a chip on a PCB?
🎓
Basically, the junction is the tiny, active silicon area inside the chip where the transistors work. Its temperature, $T_j$, is the hottest spot. If it gets too high, the chip can fail instantly or degrade over time. In practice, every component has a maximum $T_j$ in its datasheet, often 125°C or 150°C. Try moving the "Power Dissipation P" slider in the simulator above—you'll see how directly it drives up $T_j$.
🙋
Wait, really? So the heat has to travel through the chip package and the board? How do we model that path?
🎓
Exactly! We use a concept called thermal resistance, measured in °C/W. It's like electrical resistance but for heat flow. The path from junction-to-case ($\theta_{jc}$), case-to-heatsink ($\theta_{cs}$), and heatsink-to-ambient ($\theta_{sa}$) are like resistors in series. The simulator lets you adjust each of these $\theta$ values to see their individual impact on the final temperature.
🙋
That makes sense. But what about the PCB itself? I see parameters for "Thermal Via Count" and "Board Thickness."
🎓
Great observation! The PCB isn't just a mount; it's a heat spreader. Thermal vias are small copper-plated holes that conduct heat from the top layer down into the ground plane. In the simulator, when you increase the "Thermal Via Count N," you're adding more parallel paths for heat, which dramatically lowers the thermal resistance from the component into the board. A common case is using an array of vias under a BGA package.

Physical Model & Key Equations

The core model treats the heat flow path as a series of thermal resistances. The total temperature rise from the ambient air to the silicon junction is the product of the power dissipated and the sum of all resistances.

$$T_j = T_a + P \times \theta_{ja}$$

Where $T_j$ is the junction temperature (°C), $T_a$ is the ambient temperature (°C), $P$ is the power dissipation (W), and $\theta_{ja}$ is the total junction-to-ambient thermal resistance (°C/W).

The total resistance $\theta_{ja}$ is broken down into the resistances of each material and interface in the path. For a component with a heatsink, the series sum is:

$$\theta_{ja}= \theta_{jc}+ \theta_{cs}+ \theta_{sa}$$

$\theta_{jc}$: Junction-to-Case (from silicon to package surface). $\theta_{cs}$: Case-to-Spreader/Heatsink (includes thermal paste/interface). $\theta_{sa}$: Heatsink-to-Ambient (depends on heatsink size & airflow).

The cooling effect of thermal vias in the PCB is modeled as parallel thermal resistances. The resistance of a single via and an array of N vias are:

$$R_{via}= \frac{L}{k_{Cu} \cdot \pi r^2}, \quad R_{array}= \frac{R_{via}}{N}$$

$L$ is the board thickness (m), $k_{Cu}= 385 \text{ W/m·K}$ is copper's conductivity, $r$ is the via radius (m). Adding more vias ($N$) directly reduces the resistance, pulling heat away from the component more effectively.

Frequently Asked Questions

θjc is listed in the semiconductor manufacturer's datasheet. θcs is based on the specifications of the thermal sheet or grease, and θsa is estimated from the heat sink manufacturer's catalog values or conditions of natural convection and forced airflow. You can also use the reference value list provided in the tool.
Thermal vias are indirectly reflected as a reduction in θjc or θsa. For example, calculate the thermal resistance based on the number of vias and copper foil area, and incorporate it into the θsa value. For detailed via design values, referring to a separate calculation sheet will improve accuracy.
The basic formula Tj = Ta + P × (θjc + θcs + θsa) is common. However, for BGA and QFP, the definition of θjc may differ depending on whether it is the top case or the ball side, so be sure to correctly select the thermal resistance value from the datasheet. The tool presets representative values for major packages.
Main countermeasures include: (1) lowering θsa by enlarging the heat sink or enhancing airflow, (2) reducing θcs by changing the thermal sheet, (3) reducing the power consumption P, and (4) limiting the ambient temperature Ta. The tool allows you to change each parameter and compare the effects in real time.

Real-World Applications

Power Electronics (EV/Server PSUs): High-current MOSFETs in TO-220 or D²PAK packages dissipate significant heat. Engineers use this exact model to select heatsinks and design PCB copper pours with via arrays to keep $T_j$ below 150°C, preventing thermal runaway.

Processor & GPU Cooling: Modern BGA-packaged CPUs have a $\theta_{jc}$ provided by the manufacturer. System designers combine this with the performance of a cold plate, thermal paste ($\theta_{cs}$), and a liquid cooling radiator ($\theta_{sa}$) to ensure stable boost clocks under load.

LED Lighting Design: High-power LEDs have a very low maximum junction temperature (often ~105°C). The thermal path includes the MCPCB (Metal Core PCB), thermal interface material, and an aluminum heat sink. Accurate $T_j$ calculation is critical for lumen maintenance and lifetime.

Automotive ECU Design: Electronic Control Units under the hood face high ambient temperatures ($T_a$ ~ 85°C). Thermal analysis ensures components like motor drivers can dissipate power through the PCB and enclosure without exceeding their rated $T_j$, meeting stringent automotive reliability standards.

Common Misconceptions and Points to Note

When you start using this type of calculation tool, there are a few common pitfalls you might encounter. First is the misconception that thermal resistance θja is a fixed component-specific constant. The θja listed on a datasheet is a value measured under specific conditions (like a JEDEC standard board). If your actual PCB's layer stack or copper area is different, the effective θja can change significantly. The value of this simulator lies in its ability to derive the "effective θja under your specific design conditions" by changing parameters for the PCB and heat sink.

Next is a blind spot in parameter input. For example, you might be tempted to enter the catalog value for "Heat Sink-to-Ambient Thermal Resistance θsa" directly, which is typically for "still air." However, if there's a fan inside the enclosure, you have forced air cooling. For instance, a heat sink with a θsa of 10 K/W in still air could easily drop to around 4 K/W with an airflow of 2 m/s. The tool's "Air Velocity v" parameter exists precisely to account for this effect, so set it by imagining your actual usage environment.

Finally, remember the fundamental principle that "calculation results are not absolute." This tool is a convenient first-order approximation tool using a simple model: the thermal resistance network (lumped model). Actual heat flow is three-dimensional, and thermal interference (coupling) between components also occurs. For example, if two heat-generating ICs are placed close together on a board, they will "heat" each other, resulting in higher temperatures than calculated. Use the simulation results as a compass to confirm your design direction and prioritize countermeasures.