Although nearly all grid power is generated directly as AC power, the electronics we use everyday rely on DC power. Newer technologies in alternative energy, electric vehicles, infrastructure, motor control, and robotics rely on power converters that can take an AC input, rectify it to DC, and convert this to required standard voltage with high efficiency and high power density. In these advanced technologies, the power density and efficiency is very important as these systems all rely on high power delivery with fast response time at a standard voltage.
The selection of DC-DC converter topology is very important as it will determine the power density and level of regulation required, as well as the absolute amount of power that can be extracted from the system. The modern applications I mention above rely on half-bridge isolated DC-DC converters to balance the need for high-efficiency power delivery at standard voltages with cost and footprint.
Why use a half-bridge topology over another topology like full-bridge? As we look through the design points surrounding the half-bridge topology, some of its advantages over full-bridge topologies should become apparent. We’ll examine these points and show some example components you can use to implement your own half-bridge isolated DC-DC converter design.
Half-bridge DC-DC converter circuits have a few common elements that appear in the system that implement driving, rectification, filtering, EMI reduction, and control:
Coupling element: This refers broadly to a pulse/power transformer, or an optocoupler. The choice of switching element depends on power output and where the coupling element is placed and which signal is being coupled to the output. Note that the coupling mechanism could be built into the gate driver.
Switching element: In the past this would have been bipolar Si transistors, but newer systems are using GaN or SiC FETs as switching elements to ensure highly efficient power conversion.
Gate driver circuit: The driver is normally a highly integrated PWM generator, which could operate at logic levels depending on the gate drive characteristics of the switching FETs. These components will sometimes have an integrated current sense pin, and the driver will adjust the output should there be any fluctuations.
Diodes: These rectifying elements are usually found on the output side when the drive circuit/FETs are placed on the input side, or vice versa. These will force the output current to always flow in the same direction regardless of the input current direction.
Some example half-bridge isolated DC-DC converter designs are shown below. You’ll notice that the coupling element could couple either the gate drive signal or power to the output.
I haven’t included feedback to the driver circuitry just to prevent the above images from getting cluttered. However, you could certainly include a feedback loop through a current-sense resistor and a current-sense amplifier. Some highly integrated driver components will have this built in as they are not highly configurable, meaning they are driving particular FETs to deliver specific voltage levels to the load. In any of the circuits where the drive stage is on the input side, you would need optocouplers to span the gap.
These examples implement two types of coupling: power coupling and gate drive coupling. Only the upper-left implementation couples power directly to the output, thus it would be used for high power delivery to a load. The diodes would need to have high reverse breakdown voltage. The output inductor (L1) would need to also have very low DC resistance/high current rating to handle the delivered power. This type of system could be used in step-up or step-down mode, depending on the load requirements. This type of system would typically be implemented as a standalone board, such as in a dedicated PSU.
All of the other methods shown above couple a gate drive signal or gate enable signal. In the case where the gate drive signal is being coupled, the coupling element will normally be an optocoupler, assuming the gate drive is at logic levels, although transformers could be used. All of this could be baked into the gate driver circuit; it does not have to be placed in separate components unless exceeding power ratings is of any concern.
Isolation will be needed when the output current is very high. These circuits all vary in terms of how isolation is implemented in terms of components. In terms of the PCB layout, implementing isolation is simple; you simply split the ground plane into two nets between the input and output sides of the system. This creates two different sets of current loops on each side of the system (input/output or primary/secondary).
In gate drive coupling circuits, isolation is implemented in three possible ways: with an optocoupler, a transformer, or inside the gate drive circuit using junction isolation. The exact method that should be used depends on how power is being coupled across the gap. If it is just the gate drive signal, which will be low voltage/current, then an optocoupler or isolated junction will be appropriate. An optocoupler could be used to couple just the enable signal that turns on a driver component (lower left in the examples above), or the PWM signal could be coupled across the isolation gap using either optocouplers or small transformers (both circuits on the right).
In all three cases with gate drive coupling above, the function of the DC-DC converter is effectively to modulate power delivery from a source on the output side. The “Load” in the above cases would then be the standard rectification stage in a buck/boost converter, or possibly a flyback converter with an additional transformer. This will then take the pulsating output and convert it to stable DC voltage.
If power is being coupled directly to the output, it will be switched on the primary side of the transformer, then it will be coupled to the output side. The ground plane gap would need to be placed directly between the transformer coils in the typical manner. An example of this implementation is shown below, where power is supplied entirely by the V_IN net. This should explain why the primary side half-bridge driver implementation would normally be found in an isolated power supply design.
The GND nets would also need to be tied together with a safety capacitor, such as a Y-type capacitor. The capacitor should be placed across the ground gap between both sections of the converter, and it should have a capacitance exceeding the interwinding capacitance of any transformer used for coupling. Make sure the capacitor’s leakage current is not too large as this can create a mild shock for the user should they interact with the ground terminal on the output side.
The choice of FET is also important in the above topologies as this will effectively determine the operating limits of the converter as a power delivery device. In the above implementations, the type of FET being used does not affect the overall design of the circuit, at least in terms of its topology. Instead, the FET (GaN, GaAs, SiC, etc.) should be chosen based on the required voltage or switching speed, particularly if power will be drawn directly through the output. Total power that needs to be supplied to the device will also matter as these components can quickly overheat when used to provide excessive voltages.
A full-bridge topology follows the same conceptual ideas regarding switching and rectification in a half-bridge converter, but it uses an H-bridge arrangement in the MOSFET bank to implement switching. In other words, this requires at least 4 transistors with a coordinated driving circuit. Just like in a half-bridge DC-DC converter, control might be implemented through a feedback loop by bringing the output back to the driver circuit through a current-sense resistor. The driver’s PWM signal can then be adjusted to compensate for output phase variance or droop in the output voltage.
Normally, full-bridge DC-DC converters would be used in a resonant LLC topology with the MOSFET bank on the primary side of a pulse transformer. However, you could certainly place the MOSFET bank on the output side of the system. The table provides a comparison of full-bridge and half-bridge DC-DC converter options with some of their advantages and disadvantages.
From the above table, we can briefly summarize the tradeoffs in using a full-bridge vs. half-bridge DC-DC converter topology. The full-bridge topology can access higher output power than a half-bridge topology; for the two to be equal you might need to use parallel FETs in the half-bridge arrangement with the banks placed on the output side. However, the tradeoff for higher power and lower ripple is potentially double the switching noise and the need for physically larger inductive components whenever they are implemented on the power output node in the system. The footprint is also larger for the full-bridge system, partially due to the larger number of MOSFETs but also due to the gate driver stage.
Obviously, there are many components to consider when selecting parts for a half-bridge isolated DC-DC converter design. Many semiconductor companies have developed product portfolios that specifically target half-bridge and full-bridge topologies with high-efficiency FETs and integrated drive/control circuits to ensure very precise regulation, even if there is dropout in the input voltage. Some of the components you’ll need for your DC-DC power converter can be found in our other guides:
When you need to find components for your half-bridge isolated DC-DC converter design, use the advanced search and filtration features in Octopart. When you use Octopart’s electronics search engine, you’ll have access to up-to-date distributor pricing data, parts inventory, and parts specifications, and it’s all freely accessible in a user-friendly interface. Take a look at our integrated circuits page to find the components you need.
Stay up-to-date with our latest articles by signing up for our newsletter.