An Interview with Walter MacWilliams

 Developing the First “Working” Transistor Application


Curator’s Introduction

As an introduction to the Walter MacWilliams Oral History, the following commentary provides useful background and context for this historic transistor work.  See [1] for additional related material. 


Walter H. Mac Williams, Jr. joined Bell Labs in January 1946, after five years of Navy service in research and development of antiaircraft fire control equipment.  His initial work at Bell Labs was a continuation of the so-called Mark 65 program, which was a broad-based study of the defense of a combatant ship, armed with 5”/38 gun mounts, against a coordinated air attack, involving attack speeds expected to exist in 1960, then 15 years into the future.


The integrated defense system evolved represented the developing air attack as track-while-scan data from the ship’s air search radar, assigning gun directors (with their associated computers generating data for firing the defensive 5”/38 gun mounts), and issuing orders for firing the guns, taking into account the relative threats of the incoming aircraft and the firing clearance of the guns as the ship undertakes evasive maneuvers.  The system included a tactical display with controls for making the assignments manually as a keyboard operation, simplifying the making of air-defense decisions.


With the higher speeds of incoming aircraft the already-tight period for making defense assignments would be further compressed.



Curator’s Introduction 

Could the tightened time available for making defense decisions be countered by making a computer-controlled defense?  Hence a defense calculus was developed, whereby the dynamic threats represented by the incoming aircraft and the quantized effectiveness of defensive gunfire from available gun mounts could be used to make assignment decisions automatically.   This is believed to be the first such reduction of an air defense operation to computer control.


Among the questions that arose naturally were, “How effective could a computer-controlled defense be?”, and “How effective would the computer-controlled defense be, compared to the defense made by human operators using the sophisticated display and assignment facilities?”


To address these questions, he proposed the construction of a Gunnery System Simulator, which could run simulated attack and defenses – both manual and automatic - against them.   A Gunnery System Simulator was duly designed, which represented the dynamics of the simulated attack and defense. One of the components of the simulator was the Gun-to-p-computer Switch, which allowed taking into account the statistical effectiveness of individual gun shots, including the effect of firing from more than one gun mount.   This Gun-to-p-computer Switch was implemented with point contact transistors, which had just become available as pre-production samples.

Go To MacWilliams Curator's Introduction, Pg 2


COPYRIGHT © 2005 by Jack Ward.  All Rights Reserved.  http://www.transistormuseum.com/