Kosmos 419

By 1971 the Soviet Mars programme had a plan and a problem. The plan was two spacecraft of the M-71 type, each an orbiter carrying a lander, launched into the most favourable Mars window in fifteen years. The problem was that NASA was sending two orbiters of its own into the same window, and whoever arrived first would own the phrase first spacecraft to orbit another planet. So Lavochkin added a third vehicle. The 3MS, also written M-71S, was a lightened version of the same design with the descent module removed, and its job was to get to Mars orbit ahead of the Mariners. It had a second, quieter job that mattered more to the engineers. Nobody knew exactly where Mars was, to the precision a lander needs. The distance and direction to the planet were known well enough for a flyby and not well enough to choose an atmospheric entry angle, and an entry angle that is a couple of degrees too steep will kill a lander. An orbiter arriving early would fix the planet's position by radio and act as a beacon for the two that followed. It was also the first Soviet interplanetary spacecraft to carry an onboard digital computer, which was supposed to compute those entry settings for the landers.

It launched on 10 May 1971 at 16:58:42 UTC from Pad 23 at Site 81, one day after Mariner 8 fell into the Atlantic. The Proton worked exactly as advertised and put the spacecraft and its Blok D upper stage into a low parking orbit of roughly 145 by 159 kilometres. That is not the end of the journey. A trans-Mars injection from parking orbit has to happen at a particular point on a particular orbit, which for this trajectory meant a burn about an hour after launch, over the Atlantic near the Gulf of Guinea, where the Soviet Union kept a tracking ship for exactly this purpose. The ship reported nothing. A few minutes later the stack came over the Yevpatoria control station in Crimea and the telemetry was unambiguous: the engine had not failed to light, it had never been told to. The command had not been issued. What was left was a four and a half tonne spacecraft in a 150 kilometre orbit, far too low to stay up, with a full upper stage attached and no way to use it. TASS announced Kosmos 419, one more entry in a numbered series that by then had absorbed a decade of failed Soviet planetary launches. It reentered on 12 May.

The cause was human and it was small. The ignition time for the Blok D had been loaded into the onboard computer with an error in one digit, so that instead of firing after an interval of well under two hours the engine was scheduled to fire after roughly 150 hours, by which time the spacecraft had long since burned up. Perminov, who was in the control room, tells it slightly differently: an operator entered an eight-digit code command in the wrong order, and the result was that no start command was ever generated. The two versions describe the same kind of mistake and the same consequence, and the distinction between them has never been resolved in public. What is clear is what it was not. Nothing broke. The rocket flew, the stage was healthy, the spacecraft was alive. The most widely circulated version in English, that the timer was set for 1.5 years instead of 1.5 hours, appears in NASA's NSSDCA entry and in English Wikipedia, and Siddiqi notes it as a single report while giving 150 hours as the figure. The fix took days. Mars 2 flew on 19 May with a corrected program and Mars 3 on 28 May, and both got there.

The loss had a long tail. France had been promised that two copies of its Stereo-1 decametric radio astronomy instrument would go to Mars in 1971, and one of them was riding on Kosmos 419. Since the mission officially had nothing to do with Mars, the Soviet side could not say what had happened to it, and when only one working Stereo appeared aboard the pair that did leave for Mars, the French were allowed to conclude that the other instrument had simply failed. More seriously, Mars 2 and Mars 3 went to Mars without the ephemeris refinement and radio beacon their mission design had assumed, relying instead on an optical angle-measuring system approved as a backup in 1970. On 27 November 1971 the onboard computer aboard Mars 2 generated the wrong pre-separation settings through a software error, the lander entered too steeply for its parachute to help, and it was destroyed. Mars 3 landed on 2 December and transmitted for seconds. Mariner 9, the survivor of the American pair, had already entered Mars orbit on 14 November. The race Kosmos 419 was built to win was decided by a mistyped number in a computer at Baikonur.