Has CMOS killed the BJT?

Jeff Dykstra, Ph.D., Motorola Labs, Schaumburg, IL

ABSTRACT:
A common aspiration of chip designers is to make a system on a chip. It is a foregone conclusion that this is a good thing since SOCs will obviously reduce cost and make better system designs. In the rush to make SOCs, the widespread assumption is that CMOS is the only technology that is available. This talk will take a look at what has been lost along the way by banishing the BJT. We will present a quick introduction to the BJT, look at some of the tradeoffs between FETs and BJTs, and then look at some applications appropriate for BJT IC design. Next, we will look at a few circuits that are very difficult to reproduce with only FETs. Then a few examples of production BiCMOS chips from Motorola/Freescale will be presented.

BIOS:
Jeffrey A. Dykstra received the B.S.E.E. degree in December 1982, and the M.S.E.E. degree in April 1984 from The University of Michigan; requirements for the Ph.D. degree were completed in October 1990, also at The University of Michigan. Calvin College granted him the B.S. in Letters and Engineering in May 1981. He is presently employed by Motorola Laboratories in the Wireless Access Research center of excellence. His present responsibilities include working in the area of advanced IC research and design of communication circuits. From 1984 through 1986, he was employed by E-Systems' E.C.I. Division where he was involved in the design and development of a digital UHF radio link. He is a member of Eta Kappa Nu, Tau Beta Pi, and I.E.E.E. He has been granted three patents. He was the chairman of the WIMS ERC IAB from 2002 to 2004.