During the first minutes of flight, there seemed to be two difficulties with the AACS. The first was a problem with one of the three stabilizing gyroscopes, but fortunately, the gyroscope began operating normally without intervention from the ground. The other problem appeared to be with one of the AACS computers; the spacecraft switched to a backup computer during the Titan burn, and initial data transmissions were incomplete. Early analysis seemed to indicate that an event during the launch itself, rather than a faulty spacecraft computer system, was the cause of the data loss. At first, on August 23, officials suspected that perhaps the spacecraft had been bumped by the rocket motor one hour after liftoff and again about seventeen hours later, when telemetry signals indicated that the spacecraft had been jolted. However, by the next day, flight engineers determined that electronic gyrations in the AACS seemed to have caused the difficulty. [Illustration: The Titan-Centaur rocket used to launch Voyager stood as tall as a 15-story building and weighed nearly 700 tons. Here the rocket waits for launch at Kennedy Space Center, with the Voyager spacecraft enclosed in the white protective shroud at the top. [P-19471A]] Within an hour after launch, Voyager 2’s science scan platform boom was to have been fully extended and locked. Instructions to deploy were given, and the boom moved outward; however, there was no signal to indicate that the boom was actually locked in place. Efforts to command the boom to move into the locked position were thwarted by the spacecraft. The first maneuver designed to try to lock the boom was aborted by the computer command subsystem (CCS) when the AACS erroneously indicated that it was in trouble. Three days later another maneuver was scheduled to reprogram the faulty computer in the AACS, to align the Sun sensors, and to try to lock the science boom. To provide a direct check of the boom position, the scan platform was turned so that the TV cameras could see the spacecraft. Careful measurement of these pictures verified that the boom was within ½ degree of full deployment, but still there was no indication that it was locked into place. Ultimately, it was decided that the sensor to signal actuation of the lock was at fault, and that the boom itself was almost certainly fully extended and operational.