High efficiency, low ripple DCS-Control achieves seamless PWM energy-saving conversion

The synchronous buck converter with DCS-Control technology is a direct control regulation topology that can seamlessly switch to energy-saving mode. This topology combines many advantages of voltage mode, current mode and hysteresis control topologies, and achieves smooth transition to energy-saving mode. This article introduces the working principle of the DCS-Control topology and demonstrates its low output voltage ripple, excellent transient response and seamless mode conversion performance in energy-saving mode .

Basic working principle

The DC-Control topology is basically a hysteresis topology. However, it combines several circuits to have the advantages of both voltage-mode and current-mode topologies. Figure 1 shows the basic block diagram of the DC-Control topology (taken from TI's TPS62130 buck converter data sheet).

Figure 1 DCS-ControlTM topology diagram

There are two inputs to the DC-Control topology: the feedback (FB) pin and the output voltage sense (VOS) pin. The FB pin input of most DC/DC converters behaves the same. It is a high-impedance input of the error amplifier or operational amplifier, and its purpose is to output the error signal of the FB pin to a certain internal reference voltage VREF. As in other DC/DC converters, the error amplifier provides precise output voltage regulation. The voltage divider between the output voltage (FB pin) and ground sets the set point of the output voltage. For some devices, such as TI's TPS62131, the FB pin is internally connected through a VOS pin voltage divider. In this way, the output voltage can be set, reducing 2 external components and reducing the sensitivity of the FB pin at the same time. Appropriate compensation is included around the error amplifier to ensure its stability.

The VOS pin is connected directly to the output voltage of the converter at the output capacitor. Like the FB pin, it is a high impedance input to the control loop. Unlike the FB pin, the VOS pin enters a proprietary circuit that forms a voltage ramp. This voltage ramp is then compared to the error signal of the error amplifier, as is done with voltage-mode and current-mode control. The path from the VOS pin to the comparator gives the DCS-Control topology a fast hysteresis response. Output voltage changes at VOS are fed directly to the comparator and immediately affect the operation of the device. Because of this, the VOS pin is sensitive to noise; therefore, the path for the output voltage to return from the output capacitor to the VOS pin of the device should be as short and direct as possible. Appropriate compensation around the VOS pin circuit is intended to ensure stability.

The comparator then outputs a signal to the control circuit, telling it whether to output a switching pulse to the gate driver to control the high-side MOSFET. The comparator works in conjunction with the timer circuit to provide the fastest load transient response and regulated switching frequency at the same time.

Based on the ratio of VOUT to VIN, the timer sets a minimum "on" time that extends the comparator's "on" time control. The device data sheet usually uses an equation to describe the minimum "on" time set by the timer, such as:

In this TPS62130-based example, the target switching time is 400ns; therefore, the switching frequency is its inverse, 2.5MHz. Due to the VOUT/VIN factor, the switching frequency is adjusted to maintain the input and output voltage range, which adjusts the minimum "on" time according to the ideal duty cycle of a certain buck converter. Therefore, the "on" time equation can also be written as

The low-side MOSFET control is relatively simple. After the high-side MOSFET turns off, the low-side MOSFET turns on and effectively ramps down the inductor current. When the inductor current decays to zero or the comparator turns the high-side MOSFET on again, the low-side MOSFET turns off. An appropriate dead time is applied to avoid shoot-through current in the MOSFET. Power-saving Mode

A key component of the DCS-Control topology is its energy-saving mode. Generally speaking, most energy-saving modes are enabled at low load currents, which improve conversion efficiency by skipping switching pulses and reducing the current consumption (quiescent current) of the device. Skipping switching pulses allows the device to operate in discontinuous conduction mode (DCM), eliminating negative inductor current (flowing from the output to the input), which would otherwise appear under light load conditions. This type of current only disrupts the work of the previous switching cycle and brings more losses, thereby reducing efficiency. Reducing quiescent current can improve efficiency at ultra-light loads.

The energy-saving mode of the DCS-Control topology is very simple. Its implementation circuit is the same as described above: during the transition from energy-saving mode to PWM mode, there is no switching between the two different control modes. Some other control topologies switch between one energy-saving mode control method and another PWM mode method. In doing so, electronic pulse interference and random noise may appear during the transition. This phenomenon will be explained in detail in "Seamless Transition" later in this article.

The DCS-Control topology uses a simple method to implement its energy-saving mode: if the comparator does not need a switching pulse, no pulse is generated. Therefore, if the output voltage is above its set point (measured by the error amplifier) ​​when the inductor current decays to zero, the device does not output a new switching pulse; instead, it reduces its quiescent current and enters energy-saving mode. Unless the error amplifier tells the comparator that the output voltage has dropped to its set point and should now be boosted, it will wait. After that, the device outputs a switching pulse that lasts for the minimum "on" time, raising the output voltage enough to stay within the regulation range. The minimal propagation delay of these circuits in energy-saving mode results in high efficiency and well-regulated output voltage.

A single switching pulse with a minimum “on” time delivers minimal energy to the output, resulting in minimal output voltage ripple. As light load current increases, the single pulses are moved closer together, increasing the switching frequency above the audio band at a faster rate than other energy-saving topologies. Other topologies use arrays or continuous pulses in energy-saving mode, resulting in more energy at the output during the pulse. Since it takes longer for the output voltage to fall back to its set point, the pulses are spaced further apart, keeping the effective frequency in the audio range longer. DCS-Control’s single pulse architecture allows it to operate above the audio band and at lower load currents than other topologies.

When the load increases to a certain level and there is no time interval between the individual pulses, the inductor current will not return to zero before the comparator tells the high-side MOSFET to turn on again. This load condition occurs at the boundary of DCM, at which point the converter exits power-save mode and enters PWM mode.

Output voltage ripple in power saving mode

The combination of energy-saving mode (single pulse with minimum "on" time) and entering PWM mode when zero inductor current is reached makes the DCS-Control topology more flexible than other topologies, allowing for simpler configuration to ultimately meet system requirements. For example, consider the output voltage ripple of a system with 12V input and 3.3V output in energy-saving mode. TI's TPS62130 evaluation module (EVM) operating at a 2.5MHz setting is used in Figure 2 to demonstrate how this ripple can be reduced by adding external inductors and output capacitors. The no-load state is used to show the extreme output voltage ripple in energy-saving mode.

Figure 2 Output voltage ripple of TPS62130

Figure 2a shows the already low 26mV peak-to-peak output voltage ripple, or 0.8% of the 3.3V output voltage, obtained using the default circuit. Since the same amount of energy is transferred during each switching pulse, increasing the output capacitance reduces the output voltage ripple. Higher output capacitance results in less voltage ripple for a fixed amount of energy (Figure 2b). Since the “on” time is constant, increasing the inductor reduces the peak current reached within the switching pulse. The lower peak current also stores less energy (E = ½ × L × I2), so less energy is transferred to the output, again reducing the voltage ripple (Figure 2c). Note that the “on” time is the same for each circuit because it is fixed internally to the device and cannot be changed by external components.

Engineers can also set the load current at which the power save mode is entered by adjusting the inductor, which changes the boundary to DCM. Larger inductance leads to smaller inductor current ripple, which means that the inductor current is kept above zero, resulting in lower output current levels. It allows the power save mode entry point and output voltage ripple to meet various special needs, allowing this topology to be used in a variety of applications, including those that are highly sensitive to noise, such as: low-power wireless transmitters and receivers in medical or industrial applications, portable power supplies for consumer devices, and solid-state drive power supplies. Transient response

Since the DCS-Control topology senses the actual output voltage via the VOS pin, it is well suited to responding to load transients. This signal is fed directly to the comparator and does not pass through a bandwidth-limited error amplifier, which does not affect the "on" time. The DCS-Control topology responds faster to load transients due to its hysteresis characteristics, and the device's 100% duty cycle further enhances this capability.

In this mode, the device keeps the high-side MOSFET on as long as the output voltage recovery requires. In other words, the comparator’s “on” time requirement is fully met. Figure 3 shows the TPS62130 EVM responding to a no-load to 1A load transient with its 100% duty cycle. The 300ns time delay between the start of the transient and when the high-side MOSFET turns on means that the transient response is almost entirely limited by large-signal issues (the inductor) rather than small-signal issues (the control topology). Therefore, the DCS-Control topology is not the primary limitation on the device’s transient response capability; it achieves excellent transient response when using specific output filter components.

Figure 3. 100% duty cycle mode of the TPS62130 EVM during transient response.

Seamless transition

Earlier, we noted that in the DCS-Control topology, only one circuit controls both PWM and energy-saving modes. It achieves rapid and seamless transitions between the two control modes. In addition, when the circuit's operating state approaches the boundary between the two modes, it still has higher performance. Since there is no mode switch, there is no output pulse interference.

Figure 4 compares the mode-switching performance of the TPS62130 with a device using an alternative control topology. In the quasi-triangular mode, the load current (bottom line in green) ranges from 10 mA to 1 A. We observe both perturbations or disturbances in the inductor current and the output voltage ripple.

Figure 4: Transition from PWM mode to energy-saving mode

For the TPS62130 using the DCS-Control topology, Figure 4 shows that both the output voltage and inductor current waveforms are smoother than those of the device using the alternative control topology. The TPS62130 outputs a smaller voltage ripple at all load currents. The ripple increases slightly at higher loads; however, this ripple increase is much less than that of the device using the alternative topology because the device enters power-save mode. Last but not least, the output voltage drops significantly as the load increases (under some limited operating conditions, such as load ramping), while the device using the alternative topology exits power-save mode and enters PWM mode. Obviously, this is an undesirable load or system situation, and the DCS-Control topology can avoid this situation.

in conclusion

The DCS-Control topology is a huge improvement over other control topologies, with excellent transient response and seamless transition to energy-saving mode. Its single-pulse energy-saving mode has lower output voltage ripple and improves the performance of various end equipment and systems, including noise-sensitive applications.