Analog & Power Business Intelligence Product News Technology

Honey, TI Shrunk The Buck Converter!

Size does matter! In contradiction to other areas for electronics the rule usually means the smaller the better. However laws of physics often restrict us from saving space and shrinking parts. This has been the case for power conversion for several years. Texas Instruments now developed a new buck converter based on the innovative capacitive conversion topology which enables engineers to shrink the size of on-board electronic power supplies by up to 80%.

As power conversion often takes up a significant area of the board space this is a remarkable achievement. The size reduction will benefit a huge variety of applications and segments especially if we think of the increasing number of electronics which penetrate automotive, industrial, smart home and consumer applications these days.

The new conversion topology from TI is targeted at step-down applications that convert power inputs from higher to lower voltages and it enables much higher-frequency operation than was previously possible in similar chips. The first products which are based on the new technology are designed for communications infrastructure, mass storage and test and measurement applications.

In order to understand the benefits of the new solution let’s take a look at conventional buck converters first. In order to step down the input voltage (VIN) it is switched on and off rapidly and an inductor along with other components are used in order to smooth the pulses into continuos output voltage (VOUT). VOUT is proportional to the time VIN is switched on e.g. if VIN is switched on 20% of the time VOUT should be 20% of VIN (note: the actual VOUT usually varies due to switching losses).

As shown below a buck converter usually consist out of an inductor (L), output capacitor (C) and two FETs (Q1, Q2) that alternately turn on and off the voltage input, creating two phases of operation.

Block Diagram Basic Buck Converter
Basic Buck Converter (image: Texas Instruments)

Looking at the size of such a converter the inductor will stand out as bulky. The dimensions of the inductor could be reduced with higher switching frequencies, however this leads to several challenges. The first problem are the increasing switching losses with higher frequencies. The second issue is the signal on-time which becomes difficult to control at high frequencies. This is especially problematic if there is a high VIN-to-VOUT ratio, since the signal is in its on-state during only a fraction of the pulse which isn’t enough time to accurately control converter operation within many common converter designs.

Size reduction of buck converters is highly desirable especially given that many electronic systems use more than one converter to operate the electronic components on the board and that in some cases multiple stages are used to break down the VIN to the necessary VOUT (see graphic below for a line powered system). As described above shrinking the converters isn’t easy per se, in addition the power supply usually needs to provide a high output current, so that the converter can drive a high-current device such as a microcontroller or several lower-current devices.

Power delivery system
Power delivery system (image: Texas Instruments)

The challenges above are the reason that buck converters have slimmed down only slightly due to new materials and manufacturing advances. The innovative TI approach however uses a new design layout with a capacitor in series, together with other circuit modifications in order to achieve high efficiency, smaller size and to resolve frequency issues.

The key to the new design are two parallel sections (or phases) of the converter, each of which has its own inductor (La and Lb, see graphic below). There are also additional switches that control the energy flow through the second phase.
The capacitor alternately charges and discharges and the FET power switches (Q2a and Q2b) are opened and closed which lets the current flow alternately through the two inductors in four-time intervals to establish a steady-state output at the appropriate stepped-down level.

Block Diagram Series capacitor buck topology
Series capacitor buck topology (image: Texas Instruments)

Power losses are minimised as the voltage across the capacitor is nominally 50% of VIN which allows the use of higher frequency switching. The biggest advantage of this design is that inductors and capacitors can be scaled down significantly but there are additional benefits with reduced inductor current ripple, automatic current balancing between inductors, a soft charge and discharge through the capacitor, a doubled on-time and excellent load transient response.

TI’s new technology will debut in the TPS54A20 SWIFT step-down buck converter which provides point-of-load voltage regulation in high-density systems such as communications infrastructure and massive data storage, as well as compact systems with high-performance needs, such as test and measurement equipment. The graphic below compares the board space required by a conventional buck converter operating at 500 kHz and one implemented using the new converter at 2 MHz. As the inductors in the new design are 12 times smaller than in the conventional one a big amount of cost and space can be saved.
The new design is small enough to be placed on the backs of boards, where there is limited clearance, freeing up valuable space on the top of the board for other circuitry.

Rescaling buck converter designs
Rescaling buck converter designs (image: Texas Instruments)

TI will support the implementation of the new capacitive buck conversion technology with additional products based around this technology. In order to learn more about the technology read TI’s White Paper “Breakthrough power delivery for space-constrained applications”.
For support with the implementation and advise regarding your power supply designs contact EBV here.