Critical conduction mode (CrCM) active clamp flyback (ACF) converter is regarded as a good candidate for low power adaptor applications. However, the most adopted control strategy of variable-frequency peak current mode (VF-PCM) and discrete zero voltage switch (ZVS) control makes it difficult to get the accurate small signal model. Therefore, by using sampled-data modeling method, this paper proposes an exact small signal model of CrCM ACF converter for the first time. In the modeling process, the influence of discrete-ZVS control is firstly considered. The conduction time of power switch is regarded as the control signal, which is different from constant-frequency converter. The total current close-loop transfer function is simplified by the linear combination of single VF-PCM and discrete-ZVS control model. After considering the sampling effect and the modulator gain from the derived sampled-data model respectively, the control-to-output transfer function of CrCM ACF converter is developed. The analytical results are analyzed by comparing with single VF-PCM and the widely used average model. Finally, to validate the accuracy of the proposed model, a 65W 20V output CrCM ACF converter is simulated and measured. The verification shown that the proposed sampled-data model is basically agreement with the simulation and experimental results.
26 July 2023 in IEEE Transactions on Nanotechnology
An accuracy-configurable adder (ACA) meets various requirements in energy efficiency and accuracy of many applications. However, it suffers from problems such as a large hardware cost for performing configurations. In this article, we propose two energy-efficient transistor-level accuracy-configurable full adders, referred to as a positive ACA (PACA) and a negative ACA (NACA). Each design uses only two transistors as switches driven by an enable signal to configure the accurate and approximate modes. Dynamic accuracy-configurable adders (DACAs) are further developed by cascading the PACAs or NACAs. The DACA design realizes a remarkable trade-off between accuracy and performance by switching between different degrees of approximations. Simulation results show that compared with an accurate adder, a 16-bit DACA provides a saving of up to 43.61% in power and a reduction of 83.31% in delay, with only a loss of 11.96% in accuracy. The advantage of the DACA is illustrated by applications of image processing and classification using a neural network model.
The resonant gate driver is the most efficient method to reduce the gate drive loss. However, the conventional resonant gate driver cannot be implemented for SiC power MOSFET because of the asymmetrical breakdown voltage of the gate-source. To improve both the efficiency and the reliability of the power converter, a resonant gate driver with a level shift circuit to achieve adjustable asymmetrical on/off voltage for SiC power MOSFET is proposed in this paper. The operating principle of the level shift circuit and the transition of the power MOSFET are described. A prototype is built to verify the effectiveness of the proposed gate driver. Test result shows that the proposed gate drive circuit achieves +13V/-4V gate-source voltages. Besides, the gate drive loss is reduced for over 50% compared to conventional voltage source driver.
In recent years, convolutional neural network (CNN)-based artificial intelligence algorithms have been widely applied to object recognition and image classification tasks. However, the high performance of convolutional neural networks comes at the cost of high-intensity computing and enormous numbers of parameters, which pose substantial challenges to terminal implementations. An end-to-end FPGA-based accelerator is proposed in this work that efficiently processes fine-grained pruned CNNs. A group pruning algorithm with group sparse regularization (GSR) is introduced to solve internal buffer misalignments and load imbalances of the accelerator after fine-grained pruning. A mathematical model of accelerator access and transmission is established to explore the optimal design scale and calculation mode. The accelerator is optimized by designing sparse processing elements and by scheduling the on- and off-chip buffers. The proposed approach reduces the computation of a state-of-the-art large-scale CNN, VGG16, by 86.9% with an accuracy loss on CIFAR-10 of only 0.48%. The accelerator achieves 188.41 GOPS at 100 MHz and consumes 8.15 W when implemented on a Xilinx VC707, making it more energy-efficient than previous approaches.
9 June 2022 in IEEE Journal of Emerging and Selected Topics in Power Electronics
Active clamp flyback (ACF) converter has been demonstrated that which is better than traditional flyback in high-frequency power-adaptor applications. According to the circuit structure, ACF has two modes, primary resonant ACF (PR-ACF) and secondary resonant ACF (SR-ACF). In order to improve the power density, a high-precision primary side regulation (PSR) constant voltage control method for ACF converter is proposed in this paper. The proposed method not only can be effectively adopted in PR-ACF but SR-ACF converter. The high-precision output voltage is predicted by highly accurate sampling points on primary auxiliary voltage. The applied micro controller environment improves the practicability and increases the power density of the ACF system. To verify the proposed PSR method, a prototype which can be changed to PR-ACF and SR-ACF is designed. The output voltage offset, operation efficiency and dynamic performance are analyzed. The tested results show highprecision output is obtained in ACF converter by using the proposed method, the maximum output voltage deviation which are 4% and 3.5% in PR-ACF and SR-ACF converter, respectively. Moreover, SR-ACF has a higher efficiency compared to PR-ACF converter, and a smaller voltage shoot during dynamic process, but which need a larger regulation time since the additional output inductor.