From: Robert P. Johnson [email@example.com]
Sent: Monday, March 01, 2004 14:44
To: Steve Ritz; William Atwood
Cc: Peter Michelson; Jerry Clinton; Lowell Klaisner
Subject: copper numerology
Regarding the impact of 1/2 oz copper in bias circuits instead of 1/4 oz, here are the numbers that I come up with:
1/4 oz/square foot = 7.62E-3 g/cm2
Cu density = 8.96 g/cm3
So 1/4 oz thickness = 8.5 microns
Cu radiation length = 12.86 g/cm2
So 1/4 oz radiation lengths = 0.059%
The bias circuit has one solid ground plane. The top layer of copper has only a very small percentage coverage (narrow traces and small pads), so the impact on it can be neglected to first order.
Hence we can expect that the increase in radiation lengths per tray will be about 0.12% if we go from 1/4 oz copper to 1/2 oz copper. This compares with 0.86% rad len per tray for the silicon itself.