Browse Prior Art Database

Closed Loop Delay Monitor With Correction

IP.com Disclosure Number: IPCOM000083862D
Original Publication Date: 1975-Aug-01
Included in the Prior Art Database: 2005-Mar-01
Document File: 2 page(s) / 43K

Publishing Venue

IBM

Related People

Buckley, F: AUTHOR [+4]

Abstract

Large-scale integration (LSI) poses two problems. Circuit skew creates wide timing tolerances to timing pulses and the density inhibits monitoring these pulses at the critical point. Therefore, the performance of any digital circuit is degraded by the skew in its critical timing pulses.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 2

Closed Loop Delay Monitor With Correction

Large-scale integration (LSI) poses two problems. Circuit skew creates wide timing tolerances to timing pulses and the density inhibits monitoring these pulses at the critical point. Therefore, the performance of any digital circuit is degraded by the skew in its critical timing pulses.

The present system provides a fixed predetermined path delay between two points in a digital logic system, in spite of large delay variations along the path. One application of the system, for example, is in the distribution of clock signals in a LSI processor.

One leg of a clock distribution system is shown in Fig. 1 wherein A represents a variable-delay block, B is the wiring to a clock chip C, and D represents the wiring delay between the chip C output and block E which controls the delay in variable-delay block A. Signal a is provided by the system oscillator and the digital output of block E is a function of the time between the pulse edges a and f, shown in Fig. 2.

Assume that TA/1/ is the minimum delay available from block A. If the delay of block A is adjusted until the rising edge at f is aligned with the falling edge at a, then the delay from a to f or TA + TB + TC (where TA has been adjusted from an initial value TA/1/) must be T1, Since the delay between f and b is just TD, the adjustment from TA/1/ to TA has caused the delay from a to b to be set to T1 - TD. If the resolution of the adjustment and the function of block E is hi...