Editing LGM-30 Minuteman (section)

===Guidance system===
{{main|Missile guidance}}
[[File:Autonetics D-17.JPG|thumb|Autonetics D-17 guidance computer from a Minuteman I missile.]]

Previous long-range missiles used liquid fuels that could be loaded only just prior to firing. The loading process took from 30 to 60 minutes in typical designs. Although lengthy, this was not considered to be a problem at the time, because it took about the same amount of time to spin up the [[inertial guidance system]], set the initial position, and program in the target coordinates.<ref name="1990_MacKenzie" />{{rp|page=156}}

Minuteman was designed from the outset to be launched in minutes. While solid fuel eliminated the fueling delays, the delays in starting and aligning the guidance system remained. For the desired quick launch, the guidance system would have to be kept running and aligned at all times. This was a serious problem for the mechanical systems, especially the gyroscopes which used [[ball bearing]]s.<ref name="1990_MacKenzie" />{{rp|page=157}}

[[Autonetics]] had an experimental design using [[air bearing]]s that they claimed had been running continually from 1952 to 1957.<ref name="1990_MacKenzie" />{{rp|page=157}} Autonetics further advanced the [[state of the art]] by building the platform in the form of a ball which could rotate in two directions. Conventional solutions used a shaft with ball bearings at either end that allowed it to rotate around a single axis only. Autonetics' design meant that only two gyros would be needed for the inertial platform, instead of the typical three.<ref name="1990_MacKenzie" />{{rp|page=159}}{{NoteTag|A third gyro was later added for other reasons.<ref name="1990_MacKenzie" />{{rp|page=159}}}}

The last major advance was to use a general-purpose digital computer in place of the analog or custom designed digital computers. Previous missile designs normally used two single-purpose and very simple electromechanical computers; one ran the [[autopilot]] that kept the missile flying along a programmed course, and the second compared the information from the inertial platform to the target coordinates and sent any needed corrections to the autopilot. To reduce the total number of parts used in Minuteman, a single faster computer was used, running separate [[Function (computer programming)|subroutines]] for these functions.<ref name="1990_MacKenzie" />{{rp|page=160}}

Since the guidance program would not be running while the missile sat in the silo, the same computer was also used to run a program that monitored the various sensors and test equipment. With older designs this had been handled by external systems, requiring miles of extra wiring and many connectors to locations where test instruments could be connected during servicing. Now these could all be accomplished by communicating with the computer through a single connection. In order to store multiple programs, the computer, the [[D-17B]], was built in the form of a [[magnetic drum|drum machine]] but used a [[hard disk]] in place of the drum.<ref name="1990_MacKenzie" />{{rp|page=160}}

Building a computer with the required performance, size and weight demanded the use of [[transistor]]s, which were at that time very expensive and not very reliable. Earlier efforts to use computers for guidance, [[BINAC]] and the system on the [[SM-64 Navaho]], had failed and were abandoned. The Air Force and Autonetics spent millions on a program to improve transistor and component reliability 100 times, leading to the "Minuteman high-rel parts" specifications. The techniques developed during this program were equally useful for improving all transistor construction, and greatly reduced the failure rate of transistor production lines in general. This improved yield, which had the effect of greatly lowering production costs, had enormous spin-off effects in the electronics industry.<ref name="1990_MacKenzie" />{{rp|pages=160–161}}

Using a general-purpose computer also had long-lasting effects on the Minuteman program and the US's nuclear stance in general. With Minuteman, the targeting could be easily changed by loading new trajectory information into the computer's hard drive, a task that could be completed in a few hours. Earlier ICBMs' custom wired computers, on the other hand, could have attacked only a single target, whose precise trajectory information was hard-coded directly in the system's logic.<ref name="1990_MacKenzie" />{{rp|page=156}}