The International Technology Roadmap for Semiconductors predicts that the nominal power supply voltage, VDD, will fall to 0.7 V by the end of the bulk CMOS era. At that time, it is expected that the long-channel threshold voltage of a MOSFET, VT0, will rise to 35.5% of VDD in order to maintain acceptable off-state leakage characteristics in digital systems. Given the recent push for system-on-a-chip integration, this increasing trend in VT0/VDD poses a serious threat to the future of analog design because it causes traditional analog circuit topologies to experience progressively problematic signal swing limitations in each new process generation. To combat the process-scaling-induced signal swing limitations of analog circuitry, researchers have proposed the use of bulk-driven MOSFETs. By using the bulk terminal as an input rather than the gate, the bulk-driven MOSFET makes it possible to extend the applicability of any analog cell to extremely low power supply voltages because VT0 does not appear in the device's input signal path. Since the viability of the bulk-driven technique was first investigated in a 2 um p-well process, there have been numerous reports of low-voltage analog designs incorporating bulk-driven MOSFETs in the literature - most of which appear in technologies with feature sizes larger than 0.18 um. However, as of yet, no effort has been undertaken to understand how sub-micron process scaling trends have influenced the performance of a bulk-driven MOSFET, let alone make the device more adaptable to the deca-nanometer technologies widely used in the analog realm today. Thus, to further the field's understanding of the bulk-driven MOSFET, this dissertation aims to examine the implications of scaling the device into a standard 90 nm bulk CMOS process. This dissertation also describes how the major disadvantages of a bulk-driven MOSFET - i.e., its reduced intrinsic gain, its limited frequency response and its large layout area requirement - can be mitigated through modifications to the device's vertical doping profile and well structure. To gauge the potency of the proposed process changes, an optimized n-type bulk-driven MOSFET has been designed in a standard 90 nm bulk CMOS process via the 2-D device simulator, ATLAS.
Library of Congress Subject Headings
Metal oxide semiconductor field-effect transistors--Design and construction; Metal oxide semiconductors, Complementary--Design and construction; Nanoelectronics
Microsystems Engineering (Ph.D.)
Department, Program, or Center
Microsystems Engineering (KGCOE)
Urban, Christopher, "Scaling the bulk-driven MOSFET into deca-nanometer bulk CMOS technologies" (2011). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus