Power Supply Design Tips: Choose the Right Operating Frequency for Your Power Supply

Choosing the best operating frequency for your power supply is a complex trade-off involving size, efficiency, and cost. Generally speaking, low-frequency designs tend to be the most efficient, but they are also the largest and most expensive. While increasing the frequency can reduce size and cost, it increases circuit losses. Next, we use a simple buck power supply to illustrate these trade-offs.

Let's start with the filter components. These components take up a large portion of the power supply volume, and the size of the filter is inversely proportional to the operating frequency. On the other hand, every switching transition is accompanied by energy loss; the higher the operating frequency, the higher the switching losses and the lower the efficiency. Secondly, higher frequency operation usually means that smaller component values ​​can be used. Therefore, higher frequency operation can bring great cost savings.

Figure 1 shows the relationship between frequency and volume for a buck power supply. At 100 kHz, the inductor occupies the majority of the power supply volume (dark blue area). If we assume that the volume of the inductor is related to its energy, then its volume will shrink in direct proportion to the frequency. This assumption is not optimistic in this case because the core losses of the inductor will become very high at a certain frequency and limit further size reduction. If the design uses ceramic capacitors, then the output capacitor volume (brown area) will shrink with frequency, that is, the required capacitance will decrease. On the other hand, the input capacitor is usually selected because of its ripple current rating. This rating does not change significantly with frequency, so its volume ($ area) can often remain constant. In addition, the semiconductor part of the power supply does not change with frequency. Thus, due to low-frequency switching, the passive components occupy a large proportion of the power supply volume. As we move to high operating frequencies, the semiconductors (i.e. semiconductor volume, light blue area) begin to occupy a larger proportion of the space.

Figure 1: The volume of power components is mainly occupied by semiconductors.

This graph shows that semiconductor volume does not essentially scale with frequency, which is probably an oversimplification. There are two main types of losses associated with semiconductors: conduction losses and switching losses. Conduction losses in a synchronous buck converter are inversely proportional to the die area of ​​the MOSFET. The larger the MOSFET area, the lower its resistance and conduction losses.

Switching losses are related to how fast the MOSFET switches and how much input and output capacitance the MOSFET has. These are related to the size of the device. Larger devices have slower switching speeds and more capacitance. Figure 2 shows the relationship for two different operating frequencies (F). Conduction losses (Pcon) are independent of the operating frequency, while switching losses (Psw F1 and Psw F2) are directly proportional to the operating frequency. Therefore, a higher operating frequency (Psw F2) will produce higher switching losses. When the switching losses and conduction losses are equal, the total losses are the lowest for each operating frequency. In addition, as the operating frequency increases, the total losses will be higher.

However, at higher operating frequencies, the optimal die area is smaller, resulting in cost savings. In fact, at low frequencies, minimizing losses by adjusting the die area results in a very expensive design. However, moving to higher operating frequencies allows us to optimize the die area to reduce losses, thereby shrinking the semiconductor size of the power supply. The downside of this is that if we do not improve semiconductor technology, the power supply efficiency will decrease.

Figure 2: Increasing the operating frequency results in higher overall losses.

As mentioned earlier, higher operating frequencies reduce the size of the inductor; fewer inner cores are needed. Higher frequencies also reduce the output capacitor requirements. With ceramic capacitors, we can use lower capacitance values ​​or fewer capacitors. This helps reduce the semiconductor die area, which in turn reduces cost.