Mixed-signal equalizer for disk drive read channel and digital calibration for time-interleaved A/D converter

With the rapidly growing power of digital signal processor, the bottle neck of modern signal processing has been pushed forward to the analog front end. Therefore the implementation of high speed and high performance interface with mixed signal technique has become the focus of IC design. In this dissertation, two representative topics in this area are chosen to demonstrate the philosophy as well as the potential of such design.; Hard disk storage system is the fastest developing technology during the past one decade. As the density keeps increasing, the read out channel must include signal equalization to control the Inter-Symbol-Interference (ISI) problem. In this research, a new way of implementing such equalizer using mixed-signal technique is proposed. The core of this approach is a new mixed-signal multiplier that significantly simplifies the hardware implementation of the algorithm. Two prototype chips were fabricated in 0.5μm CMOS technology and tested. Testing results show the robustness of the proposed mix-signal implementation. The prototype chips boost the working speed from the reported 200MHz to 360MHz without using more advanced fabrication process. The power dissipation is lowered down to 160mW at the same time. Compared with prior art in this area with similar performance, these two designs provide the highest working speed and the best power-per-speed performance.; Another important topic of high performance mixed-signal system is the Analog-to-Digital Converter, which acts as the direct interface between the analog world and the digital world. Among various structures, Time-Interleaved structure has the potential of achieving high speed and high resolution operation to meet the stringent requirement in wireless communication. In this dissertation, two calibration methods are presented to reduce channel mismatch effects in this architecture. As for the first calibration method, a digital background calibration for reducing timing error effects is proposed. It is the first published method to calibrate this kind of effects for real-time conversion application. The second method is to use channel randomization method to solve the distortion effects that come from device mismatches which induce gain and offset mismatches.