Browse Prior Art Database

BJT Diode for Full-Swing Complementary MOS/Bipolar Logic Circuits

IP.com Disclosure Number: IPCOM000099920D
Original Publication Date: 1990-Mar-01
Included in the Prior Art Database: 2005-Mar-15
Document File: 2 page(s) / 55K

Publishing Venue

IBM

Related People

Shin, HJ: AUTHOR

Abstract

A new parallel, complementary, BJT (Bipolar Junction Transistor) clamp diode that improves the speed of full-swing complementary MOS/bipolar logic circuits is disclosed. The new diode is merged with the driving transistors in a double-emitter structure so that its area and parasitic capacitances are minimized. The diode is additionally advantageous because it has hard clamping characteristics that track VBE variation, in contrast to CMOS diodes.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 66% of the total text.

BJT Diode for Full-Swing Complementary MOS/Bipolar Logic Circuits

       A new parallel, complementary, BJT (Bipolar Junction
Transistor) clamp diode that improves the speed of full-swing
complementary MOS/bipolar logic circuits is disclosed.  The new diode
is merged with the driving transistors in a double-emitter structure
so that its area and parasitic capacitances are minimized.  The diode
is additionally advantageous because it has hard clamping
characteristics that track VBE variation, in contrast to CMOS diodes.

      The new clamp diode applied to the full-swing MOS/bipolar logic
circuit is shown in the figure.  The diode (combination of QN2 and
QP2) is configured using the base-emitter junctions of double-emitter
BJTs such that the emitter of QN2 is connected to the node OD and
that of QP2 is tied to the node OU.  The parasitic capacitances at OU
and OD to AC ground added by the clamp diode are the portions of
base-collector junction capacitances which are increased from those
of single- emitter BJTs.  Therefore, the new BJT diode adds minimal
parasitics.

      Initially when both inputs (A and B) are '1', since MN1, MN2,
MP5, and MN3 are ON and MP1, MP2, MN5, and MP3 are OFF, OD and
the node O are shorted to VSS (= GND).  So QP2 discharges OU down to
VEBp (Xpnp turn-off voltage), clamping QN1 and QN2 in the OFF state.

      If A is changed to '0', a current flows from MP2 and most of
it flows into the base of QN1-QN2 because MP3 remains O...