Search

This letter proposes a loop-free autocalibration technique for self-balancing of flying capacitors in a three-level dc–dc buck converter. The proposed converter utilizes a dual-path power stage with balancing switches that adaptively short flying capacitors at de-energizing phases. Thus, the dual-path three-level converter ensures that flying capacitor voltages are self-balanced at half the input voltage without using additional feedback calibration loops, leading to robust operation and competitive efficiencies. The prototype was fabricated in the 0.13- μ m CMOS BCD process and adopted two flying capacitors of 4.7 μ F each and an inductor of 10 μ H with a dc resistance of 128 mΩ. When an 8-V input was converted to a 3-V output voltage with a load current of 400 mA, the measured efficiency was 88.6% at 800 kHz switching while flying capacitors are automatically balanced.

Although the number of cores is increasing continuously in modern microprocessors for applications such as HPC and AI, the available power is strictly limited by the thermal power budget. To overcome this limitation, recently, each core has been implemented with a dedicated integrated voltage regulator to increase the efficiency of power usage. Distributed digital LDO (DLDO) is a powerful solution for the integrated voltage regulator because it can supply uniform power over the entire core with reduced IR drop and help the thermal management [1– 4]. In the previous distributed DLDOs [1– 3], even though all LDO outputs are connected to drive the power-delivery network, the LDOs operate independently using their own controller, which occupies a large portion of the LDO size. Therefore, the current density in these types of structures is low. In [4], the distributed DLDO uses a dual-loop structure. In this scheme, the high current density can be achieved because the four shared global controllers control the 16 local LDOs (LLDOs) for highly accurate regulation. However, the LLDOs consume large quiescent current since they operate at a switching frequency of several-GHz for a fast transient response. Besides, the load current range is narrow due to the small switching duty-cycle range of the power FETs. Because of these drawbacks, the structure proposed in [4] has limitations in practical applications.

As internet-of-things (IoT) devices continue to be installed everywhere, the concept of disposable IoT is emerging owing to better cost-efficiency and ease of maintenance without battery recharging. Lasting several years, IoTs powered with standard alkaline batteries can be a promising solution due to the long shelf life, low cost, and high reliability of these batteries as compared to the Li-ion type. As shown in the top portion of Fig. 11.8.1, a single alkaline cell has a maximum voltage of 1.5V, but it can decay down to 0.9V [1]. To power an IoT device operating with $\mathrm{V}_{\mathrm{DD}}=2\mathrm{V}$ , a variety of design options can be considered, such as the battery configuration and the power conversion topology, as shown in Fig. 11.8.1 (bottom). The first approach is step-down conversion [2], [3] from multi-cell batteries (3×BATs) connected in series. In this configuration, cell-balancing must be carefully considered; otherwise, the energy of the BATs cannot be fully utilized if any weak cells exist among them. Considering the power conversion stage, the input RMS current $\mathrm{I}_{\mathrm{IN},\mathrm{RMS}}$ becomes much higher than the average current $\mathrm{I}_{\mathrm{IN},\mathrm{AVG}}$ due to the inherently discontinuous $\mathrm{I}_{\mathrm{IN}}$ supplied from batteries, resulting in a significant power loss in the series combination of the direct current resistance $(\mathrm{R}_{\mathrm{DCR}, \mathrm{BAT}})$ of batteries which in the case of 3 batteries in series is $3\times \mathrm{R}_{\mathrm{DCR}, \mathrm{BAT}}$ (-300m $\Omega$ ). Regarding step-up conversion [4], [5] with parallel-connected batteries, most of the energy imbalances are compelled to be uselessly wasted via reverse currents $\mathrm{I}_{\mathrm{B}}$ ‘ The $\mathrm{I}_{\mathrm{B}}$ also can cause explosion or leakage of corrosive substances. Furthermore, the inductor current $\mathrm{I}_{\mathrm{L}}$ is likely to be high, caused by the discontinuous $\mathrm{I}_{\mathrm{D}}$ delivered to the output; this significantly contributes to the power loss with a large $\mathrm{R}_{\mathrm{DCR},\mathrm{IND}}$ of the inductor. In summary, the battery cell imbalance and the discontinuous input/output (I/O) current of converter are perhaps the major issues that prevent the full utilization of alkaline batteries in IoT applications.

A DC-DC converter, which tracks minimum power loss under various conditions, is presented in this paper. With this DC-DC converter, a new approach for optimum efficiency, which implements theoretical equations intactly at the circuit level, was adopted. This approach has high reliability against process variation. The concept of lossless soft-switching was also adopted in discontinuous-conduction-mode (DCM) operation. The proposed DC-DC converter has a maximum power efficiency of 91% at the switching frequency of 1 MHz. The proposed converter is able to maintain a maximum power efficiency of around 90% over the whole range of input voltage from 2.9 to 4.5 V and output voltage from $-$ 3.8 to $-$ 5.9 V. This chip is implemented in a 0.35 $\mu$ m BCD process and occupies an area of 2.1 $\times $ 1.4 mm $ ^{2}$ .

This paper presents a novel on-chip compensation scheme, the Time-Mode Miller Compensation (TMMC), for DC-DC converter in which the compensation components are integrated on-chip. Using this proposed scheme, the DC-DC converter is stably compensated and insensitive to process variations, with significantly small compensation components ( 1 pF and 80 kΩ in this work) consuming very small silicon area owing to the characteristic of the TMMC. The small compensation components make the chip size small, with 0.12 mm2 of core area (w/o power transistors) using 0.18 μm I/O process. This core size is as small as that of the digital DC-DC converters implemented with less than sub-50 nm process. The measurement result shows that the maximum power efficiency of 90.6% is obtained at the load current of 220 mA with the switching frequency of 1.15 MHz when the input and the output voltages are 3.3 V and 2 V, respectively.

While the demand for micro-energy harvesters (μEHs) is increasing (for seamless energy source in applications such as wireless sensor node), two major problems still obstruct versatile use of them. The first problem is the self-startup capability. Because many wireless sensor nodes are likely to be located where human-maintenance is difficult, starting them up manually can be as difficult as replacing the battery. Its realization has been difficult because μEHs must be able to turn itself on without any stored energy. Some previous works have reported such a function: some needed high voltage [1] or vibration [2] and one used a transformer as a starter [3]. The other problem is the maximum power point tracking (MPPT) capability. Because it is known that MPPT algorithms usually require considerable power consumption [4], using them in μEHs is impractical. This paper suggests a new boost converter architecture and MPPT control method which can bring μEH into practical use.

A modern mobile application processor (AP) requires a variety of power voltage levels, which increases the number of external capacitors around the mobile AP. This is because the supply PCB routes from the power management IC (PMIC) to the AP have parasitic inductors. The parasitic inductors introduce ripples on power voltage lines. Therefore, external capacitors are required on the PCB routes between the PMIC and the mobile AP. To reduce the number of external capacitors, one power voltage level from the PMIC is converted into a variety of power voltage levels inside the mobile AP. This both reduces external capacitors as well as the number of power pins on the mobile AP and simplifies PCB routes. For such reasons, integrated low-dropout regulators (LDOs) are preferred in mobile APs.

This paper presents a new control scheme dubbed Bang-Bang Zeroth-Order Control (BBZOC) for Single Inductor Multiple Output (SIMO) buck converter. The main loop control utilizes a phase detector, charge pump, filter, and comparator. The SIMO buck converter with BBZOC simplifies the compensation design compared to conventional voltage mode control. This work is fabricated in 1P4M 0.35um BCD process and achieves 83% maximum efficiency with the rated output power of 1.04W. The maximum output power is 2.7W and the maximum power density is 1.07 W/mm2. Considering the difference in the process, this work represents the state of the art in the power density.

Energy harvesting is one of the key technologies used to realize self-sustaining systems such as wireless sensor networks and health-care devices. Much research on circuit design has been conducted to extract as much energy as possible from transducers, such as the thermoelectric generator (TEG) and the piezoelectric transducer (PZT). Specifically, the energy in a PZT could be extracted more efficiently by utilizing resonance as [1] and [2] demonstrated. However, the maximum output voltage swing in those techniques are limited to twice of the original swing of the PZT, and thus, had a limited energy extraction capability in spite of more energy being available from the PZT. In [3], on the other hand, the large energy is obtained with higher voltage swing, but is limited up to 247% because the load energy is used to increase the output voltage swing of PZT. To obtain far more power from PZT, we propose an alternative resonance technique through which the PZT output swing can be boosted as high as CMOS devices can sustain. This technique is applied to a dual-energy-sourced (PZT and TEG) energy-harvesting interface (EHI) as a battery charger.

This paper introduces an on-chip controller without off-chip components to reduce controller size for use as a Zero-Order-Control converter (ZOC). DC offset error by adding sawtooth signal in ZOC is self-corrected using a new control scheme, so as to overcome weak point of ZOC. This controller is applicable to both buck and boost, but a boost converter is chosen for verification here. A 0.35μm BCD process is used for chip with controller area of 0.791mm2. An efficiency of 90% is obtained for an 8 V output from 3.7 V at 480mW with 926 KHz.

For the controller design of a DC-DC converter, a Time-Mode Miller Compensation (TMMC) is introduced in this paper. Using this concept, the consuming area of the DC-DC converter can be significantly reduced without any off-chip compensation components. The chip is implemented in 0.18µm I/O CMOS whose size is similar to 0.35µm CMOS, and the core size of this work is only 0.12mm2. Peak efficiency is 90.6%, with switching frequency of 1.15MHz.

A new Ripple Current Controlled Boost Converter with capacitor-free LDO as a power IC for digital driving AMOLED display is presented to achieve high stability for load variation. And a flying capacitor as well as an inductor is adopted as another energy transfer medium to increase the converter efficiency. The proposed chip is implemented in a 0.35-μm power BCD process. Transient dip voltage of the proposed boost converter is 30mV with 70mA load change. A peak efficiency of 90% at an output power of 480mW is measured without LDOs. The maximum power is 1.5W.

A variety of controllers are used in DC-DC converters. Among them, voltage-mode control and current-programmed-mode control are widely used in industrial applications. In such controllers, however, values of inductor, output capacitor, and/or load condition usually affect loop stability and limit the performance of switching converter. Recently, load-independent-control (LIC) method is reported, where freewheeling current is fed back to overcome such a limitation [1]. While this is a viable solution in principle, it still has a vulnerable aspect that must be addressed: the feedback control is affected by the level of freewheeling current and an extra power switch is needed for freewheeling current flow which lowers power efficiency. Another LIC method using vestigial current control is reported in [2]. The weak points of vestigial control are that it needs an auxiliary output and power is consumed in steady state to regulate vestigial current. In this paper, we present a zero-order-controlled (ZOC) boost DC-DC converter that has a robust control loop and does not consume any extra power in the steady state.

The luminance of AMOLED displays varies depending on the temperature owing to panel degradation. In this paper, a luminance compensation method using a monitor pixel with a multiple output boost converter and DAC for digital driving scheme is introduced. This method makes the panel luminance constant irrespective of the temperature. Furthermore, it has lower power consumption compared with the previous compensation method using monitor pixel.