A Continuously Scalable-Conversion-Ratio SC Converter with Reconfigurable VCF Step for High Efficiency over an Extended VCR Range
Yuanfei Wang, Mo Huang, Yan Lu, Rui P. Martins
Abstract
DC-DC converters are widely used in energy harvesting systems for maximum power point tracking (MPPT) from energy sources. It is desirable to accommodate different energy sources (a wide input voltage <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{IN}}$</tex> range), together with a wide range of output voltage <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{OUT}}$</tex> range, from down-to-0.5V to up-to-2V (for button battery). Therefore, a buck-boost converter with high efficiency over a wide voltage conversion ratio (VCR) range is favorable. To reduce the volume and cost, switched-capacitor (SC) converters are in demand. However, a conventional SC only obtains good efficiency at discontinuous VCRs [1] (Fig. 30.7.1). This stems from the large voltage swing <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\Delta V_{\text{CF}})$</tex> on the flying capacitors <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{\mathrm{F}}$</tex> at non-optimum VCRs, leading to large charge sharing loss <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathrm{P}_{\text{CSL}})$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{\mathrm{F}}$</tex> bottom plate (BP) parasitic capacitance loss (proportional to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\text{CF}^{2}})$</tex> . Multi-phase continuously scalable-conversion-ratio SC (CSC) converters [2–5] split the large <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\text{CF}}$</tex> into small steps, with the help of multiple internal voltage rails from the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{\mathrm{F}}$</tex> of the adjacent phases, and the out-phasing technique [2]. For example, in a CSC step-down converter (Fig. 30.7.1), the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{\mathrm{F}}$</tex> top plate (TP) in each phase connects to either <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{IN}}, V_{\text{OUT}}$</tex> , or the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$M$</tex> internal TP rails <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(V_{\mathrm{T}1}$</tex> to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{TM}})$</tex> , while the bottom plate connects to either <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{OUT}}$</tex> , ground <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{SS}}$</tex> or <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$N$</tex> internal BP rails <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(V_{\mathrm{B}1}\ \text{to}\ V_{\text{BN}})$</tex> ‘. Then the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{CF}}$</tex> step is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{T}}$</tex> when TP connects to an internal rail: <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{T}}=(V_{\text{IN}}-V_{\text{OUT}})/(2M+1)$</tex> . Likewise, the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{CF}}$</tex> step becomes <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{B}}$</tex> when BP connects to an internal rail: <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{B}}=V_{\text{OUT}}/(2N+1)$</tex> . This reduces the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{P}_{\text{CSL}}$</tex> and BP parasitic losses at non-optimum VCRs, and hence allows a high efficiency over a continuous VCR range. The CSC step-up converter [3] shares the same benefit. Yet, with a fixed <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$M+N$</tex> value (amount of resources consumed), once at a non-optimum VCR, e.g. <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$V_{\text{OUT}}$</tex> becomes small, the reduction in <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{B}}$</tex> is too small such that the efficiency improvement is negligible, while the increased <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta V_{\mathrm{T}}$</tex> degrades the efficiency greatly, as shown in Fig. 30.7.1.