Define and understand a VRM vs. a MOSFET.
Find out how to calculate MOSFET temperature.
Explore ways to regulate VRM MOSFET temperature.
Heatwaves can be pretty unbearable. With the ambient temperature in the 90s or 100s, it’s not uncommon to feel like you’re getting roasted in the sun. In my desperation during one particular heatwave, I attempted to counteract the oven-like environment by creating the opposite extreme. I poured bags of ice into the bathtub and soaked in it for as long as I could. It’s not for the faint of heart, but the icy-cold bath dissipated all the heat from the burning sun and provided some much-needed relief.
Regulating VRM MOSFET temperature is of utmost importance, and while it may be tempting to throw it in an ice bath as a quick cooling solution, we know that’s out of the question.
When working with VRM MOSFET temperature, you’ll need a smarter approach and that’s what this article is all about. But first, what is a VRM MOSFET?
VRM vs. MOSFET
VRM MOSFET is part of the stepdown regulator on the motherboard.
Let’s get started with the term VRM [link to OrCAD | Implementing VRM Cooling in PCB Power Supply Design when live]. It stands for Voltage Regulator Module, which regulates the voltage supplied to the CPU and GPU on the motherboard. From a designer’s point of view, the VRM is a switching buck DC-DC regulator, which converts the voltage of the power supply unit to the low-level ones needed by the processors. A VRM is typically made up of a switching IC, inductor, capacitor, and the MOSFET.
MOSFET stands for metal-oxide-semiconductor field-effect transistor. As stated above, it’s one of the electrical components of the overall VRM and functions as a switch that turns on and off according to the PWM input signals. You can spot the VRM easily on the motherboard. It’s highlighted by an array of heatsinks near the CPU. Underneath the heatsinks, you’ll find the MOSFETs.
Calculating VRM MOSFET Temperature
Power transistors can get very hot and that’s the case for VRM MOSFET.
When the MOSFET is turned on, the current flows through the component. In an ideal world, all of the power supplied would be transferred to the load, and in this case the CPU and related components. In real-life, some of the power is dissipated as heat on the MOSFET, which can significantly increase its temperature.
The potential increase of the MOSFET temperature is determined by its thermal resistance. Thermal resistance is expressed in °C/W. It indicates the increase in temperature compared to the power dissipated on the MOSFET. In the datasheet, you’ll often find the parameters, RthJC and RthJA.
RthJC refers to the difference between the junction temperature and the case while RthJA measures the difference between the junction and the ambient. To get an idea of how hot a VRM MOSFET can be, use the RthJA value.
Calculate the MOSFET Power Dissipation
Start by calculating the power dissipated by the switching MOSFET. You can do so by totaling the resistive and switching losses as follow:
PDMOSFET TOTAL = PDRESISTIVE + PDSWITCHING
Find Resistive Losses
Resistive losses, PDRESISTIVE, can be calculated from the following equation:
PDRESISTIVE = [ILOAD² × RDS(ON)HOT] × (VOUT/VIN)
Find Switching Losses
Getting the switching losses is more complex but can be done with the following formula:
PDSWITCHING = (CRSS × VIN² × fSW × ILOAD)/IGATE
CRSS is the MOSFET’s reverse-transfer capacitance, which can be found in the datasheet.
Once you’ve got the total power dissipated, you’ll know the temperature of the MOSFET as follows:
TJ = PDMOSFET TOTAL x RthJA + TA
For a CPU, it is not unusual to get a value of 100°C or more. That’s how hot a MOSFET could get when powering high-end CPUs and GPUs. In the next section, we’ll explore how to keep that VRM MOSFET temperature in check.
How to Regulate VRM MOSFET Temperature
A heatsink helps to disperse heat off of the VRM MOSFET
The VRM MOSFET inevitably releases an immense amount of heat when driving power-hungry chips. Since ice baths aren’t an option, you’ll need to take a realistic approach to regulate the temperature instead so that its efficiency and longevity aren’t compromised.
The motherboard itself is a giveaway of what works for keeping the heat in control. Heatsinks, obviously, are handy in dispersing the heat from the MOSFET since their purpose is to transfer heat away from high-temperature devices. A properly ventilated casing helps to displace the heated air to the exterior and this means you’ll need cooling fans installed. In some cases, liquid coolant is used to keep the MOSFET temperature under control. The heat is then transferred to the radiator by the circulating coolant.
For PCB designers, you can do your part in keeping the VRM MOSFET temperature in check through passive techniques like allocating a good amount of thermal vias to help with heat dispersion. Also, ensure that temperature-sensitive components aren’t placed near the MOSFET.
If you’re looking to learn more about how Cadence has the solution for you, talk to us and our team of experts.