What the VRM of the motherboard consists of

The VRM, or Voltage Regulator Module, is a critical functional unit on the motherboard that converts 12V from the power supply to a stabilized low voltage for the processor and RAM. It consists of five basic functional elements, which we will discuss in detail.

You can learn more about how VRM works in dedicated material.

PWN controller
This is the control center of the entire VRM, which coordinates the amount of power transferred from the 12V line and through the power phases to the processor. Pulse-Width Modulation translates to “pulse width modulation”, which is why it is also called a PWM controller.

Six- (left) and eight-channel PWM controllers

The control is carried out by changing the width of the pulses generated by the controller, which are fed to the MOSFETs through the drivers (we will talk about them in the following sections). The pulse width depends on the amount of energy required by the processor at a given time. If the computational load has increased, then the power consumption also increases, and the processor supply voltage decreases. The PWM controller fixes this through the feedback loop and increases the width of the control pulses, thereby increasing the amount of energy entering the processor through the power phases. The voltage is restored to its original value.

Basic VRM block diagram

The block diagram shows that VRM is quite complex. It would seem, why fence a vegetable garden using a PWM controller, drivers, mosfets, and smoothing filters? After all, it is enough to use a linear stabilizer, which perfectly smooths the output voltage and is very simple. Let’s figure it out.

300-420 Practice test Questions

Identity-and-Access-Management-designer PDf

Salesforce-Certified-Administrator Dumps

C_C4H320_02 Practice Dumps

C_ARP2P_2108 Practice Dumps

L4M8 Dumps

EMT PDF Dumps

E20526 Dumps

1Z0-1056-21 dumps

DES-1121 exam

AD0-E406 Dumps

PSAT Dumps

CDPSE Dumps

CLEPHistoryAndSocialSciences Dumps

The linear regulator consists of a divider, the input of which is supplied with voltage. Stabilization occurs by changing the resistance of the regulating element (RE).

The electrical circuit of the simplest linear stabilizer

Let’s imagine that such a regulator is used to convert the voltage of the power supply (12V) to the processor voltage (1.2V). The current consumption of a 120W CPU at 1.2V is 100A (100A x 1.2V = 120W). It passes through the regulating element. In this case, an excess voltage is allocated at the latter, equal to the difference at the input and output (12 – 1.2 = 10.8 V). The power dissipation on the control element will be an impressive 1080 W (100 A × 10.8 V = 1080 W), which corresponds to the power of an average heater! The cooling system of such a power supply module would be a real monster and have a colossal cost. And the efficiency is only 10% (120W / 1200W = 0.1 × 100% = 10%).

That is why switching regulated power supplies, in particular VRM, are used to power processors. Its MOSFETs operate impulsively, periodically opening (saturation mode) and closing (cutoff mode). In the first case, the resistance is very small, on average up to 0.004 ohms. For example, let’s take the same 100 A. Power is the current squared, multiplied by the resistance: (100 A) 2 = 10,000 × 0.004 = 40 W. Now compare this figure with the power dissipated on the linear regulator.

GIGABYTE B550 AORUS ELITE V2 10 999 motherboard *
Phase multiplier (doubler)
The PWM controller has a limited number of control channels and can control the same number of VRM power phases. To get around this, phase multipliers are used, which increase them by a factor of 2-4. Doubling is more often used, therefore such elements are called doublers.

In the VRM circuit, the signal from the outputs of the PWM controller is fed first to the doubler. Then two separate signals go from it to the power phase drivers.

 

The phase multiplier generates control pulses with a time shift, while their output frequency will be half the input frequency.

This functional element is designed to control a pair of field-effect transistor switches (mosfets). It matches the low-voltage signals from the PWM controller or doubler with the required control voltages.

IR2110 driver

Switching keys from open to closed (and vice versa) leads to a short-term transition to the active mode of operation. In this mode, the heat dissipation of any transistor sharply increases, so the driver must minimize this gap. At switching frequencies in the region of 500 kHz, this is not so easy to implement. Powerful mosfets have a sufficiently large gate capacitance (over 100 pF) – for fast switching, the driver must very quickly recharge parasitic capacitances.

In addition, when simultaneously switching the keys of the upper and lower arms, a situation arises when one key has not yet had time to fully close, and the other is already opening. In this case, a through current flows through them along the 12 V circuit – the key of the upper arm – the key of the lower arm – the case. At the same time, the mosfets will get very hot.

To prevent this from happening, the driver’s task is to form a delay between the control signals of the keys. In this case, the occurrence of through currents is minimized. We will not analyze the functional device in detail, this is a topic for a separate article.

IR2110 driver block diagram

GIGABYTE Z590 GAMING X motherboard

MOSFET (Metal Oxide Semiconductor Field Effect Transistor) is a powerful low-voltage electronic switch. These power switches are used in pairs in the VRM power phases. The upper side switch connects between the 12V power supply and the input of the anti-aliasing LC filter, and the lower side switch between the filter and the case.

Scheme for switching on mosfets in the power phases

When the control signals from the driver arrive at the gates, they alternately connect the input of the smoothing filter to a 12 V power supply or to the case, providing the required flow currents in direction and magnitude.

Mosfets are made on the basis of N and P silicon semiconductors.

N-type semiconductors are based on alloyed silicon. This means that atoms of other chemical elements are added to it, which have one extra electron in relation to silicon. Atoms are embedded in the crystal lattice – as a result, surpluses are formed, which are the main charge carriers.

In P-type semiconductors , silicon is also the basis, but atoms of chemical elements are doped into it, in which one electron is missing . As a result, “holes” are formed in the crystal lattice, which are also charge carriers.

Structure of N- (left) and P-type semiconductors

As an example, consider a mosfet with an N-channel. It consists of a P-type substrate, along the edges of which there are semiconductor sections (Drain and Source). Between them is a metal plate called the Shutter. It is insulated from the substrate by a dielectric silicon oxide layer.

In the absence of voltage at the Gate, the energy of the Source electrons is not enough to overcome the energy barrier and form a channel through the substrate to the Stok. No current will flow through the transistor.

The closed state of the MOSFET

If an unlocking positive voltage is applied to the gate, an electric field will appear, which will begin to push the main charge carriers (“holes”) deep into the substrate and will attract electrons to itself, forming a conduction channel between the Source and the Drain. Current will flow through the transistor.

The open state of the MOSFET

With an increase in the voltage at the gate, at one moment the current through the transistor will reach its maximum value and will not grow any more. This mode is called saturation. It is in this mode that VRM mosfets enter when they are opened.

Saturation mode

GIGABYTE H510M H 5 499 motherboard *
Filters
The anti-aliasing filter consists of an inductance L in series with the load and a capacitor C in parallel. Therefore, it is sometimes called LC filter.

It converts short 12 V pulses to a low voltage DC voltage for the processor (1-1.4 V). The process takes place in two stages.

When the key of the upper arm is opened, current begins to flow through the inductance. Energy is stored, charging the capacitor in the process.

After the voltage across the capacitor reaches the set value, the key of the upper arm closes and the lower arm opens. Inductance has the property of maintaining the direction and magnitude of the current unchanged, so the EMF that has arisen in it retains them for some time.

The stored energy is used to power the load, helping the capacitor. After the energy in the inductor is exhausted, the current through it stops and the cycle repeats again.