BUCK step-down regulator efficiency and size trade-off

qwqwqw2088

BUCK step-down regulator efficiency and size trade-off [Copy link]

The implementation of a buck regulator inevitably involves a trade-off between efficiency and size. While this principle applies to many switch-mode DC/DC topologies, it does not necessarily apply when the application requires low output voltage and high output current (e.g., 1V and 30A), which requires a small power solution that balances efficiency and size.

High efficiency is an important performance benchmark, not only reducing power losses and component temperature rise, but also bringing more useful power under given airflow and ambient temperature conditions. From this point of view, low switching frequency is very attractive, but at the same time requires large filter components to meet target specifications such as output ripple and transient response, which increases cost and size.

A huge constraint facing system designers is the PCB area dedicated to power management. To address this, let’s review the advantages of high switching frequencies. First, inductance and capacitance requirements decrease with increasing frequency, allowing for a more compact PCB layout and smaller form factor. Lower inductance not only allows for faster large signal current changes and higher control loop bandwidth, but also faster load transient response. As a rule of thumb, the maximum loop bandwidth is 20% of the switching frequency. Finally, some interesting options emerge in terms of component selection at higher frequencies.

For example, we can look at this regulator design, which achieves the best efficiency/size/cost through careful component selection.

(1) Inductors – While powdered iron or composite core inductors offer good performance at low frequencies, higher core losses negate their value proposition above about 500kHz. At this point, ultra-low DCR ferrite magnetics are more easily achieved with low copper and core losses. Note that core losses are relatively easy to measure by simply focusing on the converter’s no-load input current. Ferrite inductors with single-turn staple windings are widely available off the shelf, and with only one winding turn required, DCRs below 1mΩ are easily achievable!

(2) PWM Controller – Now, if the design specifically requires the hard saturation characteristics of a ferrite core inductor, then the saturation current of the inductor must not be exceeded. This requires a PWM controller that can take advantage of parasitic circuit resistance to achieve accurate and lossless current sensing (read my previous blog “Accurate and Lossless Current Sensing in High Current Converters” for more details on this topic). Other key features include high efficiency gate drivers, remote BJT temperature sensing, and a fast error amplifier.

(3) MOSFET - Power semiconductor devices are the foundation for improving efficiency and size. For example, the power block NexFET series is widely praised for its innovative co-packaging of high-side and low-side MOSFETs in a vertical stacking manner. When frequency proportional losses are of concern, low QG, QRR and QOSS charges are required. In addition, low RDS(ON), high current copper clips, Kelvin gate connections and ground tabs are also important.

(4) Capacitors - At higher frequencies, ceramic capacitors are preferred over electrolytic capacitors. Large amounts of output energy storage are now unnecessary because the control loop responds quickly to transient demands. Ceramic products not only offer lower ESR, but also lower ESL, which mitigates output ripple caused by inductor splitting effects and low filter inductance.

What other factors affect the efficiency and size of a regulator? Popular topics lately include GaN MOSFETs, Power System in Package (PSIP), and Power System on Chip (PSOC). Let me know what you think?

Timothy Hegarty