The Soviet 1973 armada fails four different ways
Four spacecraft launched in 19 days. Mars 4 flew past at 1,844 km, Mars 5 leaked after nine days, Mars 6 died at touchdown, and Mars 7 missed the planet by 1,300 km.
Between 21 July and 9 August 1973 the Soviet Union threw four spacecraft at Mars in nineteen days. It remains the largest interplanetary launch campaign anyone has ever attempted. The 1973 window was energetically worse than 1971, and a Proton could not lift the combined orbiter-plus-lander design that had flown as Mars 2 and Mars 3, so the mission was split in half: Mars 4 and Mars 5 were orbiters carrying no landers, Mars 6 and Mars 7 were flyby buses carrying descent modules and no way of stopping themselves. Seven months later all four arrived. All four failed, and each one failed differently.
Mars 4 got there first, on 10 February 1974. Its braking engine never fired. It swept past at 1,844 kilometres — close enough to return one swath of imagery and a radio occultation that caught the nightside ionosphere for the first time — and then kept going into solar orbit. Mars 5 did on 12 February what Mars 4 could not, settling into a 1,755 by 32,555 kilometre ellipse with a period just under 25 hours. It worked for roughly nine days and 22 orbits, returned about 60 images, found an ozone layer near 30 km altitude, and then the pressurised instrument compartment lost its air and took the transmitter housing with it. Controllers blamed a micrometeoroid struck during insertion. Nobody can check.
Mars 6 came closest to the hard thing. Its descent module separated 48,000 km out and entered the atmosphere at 09:05:53 UT on 12 March 1974, shedding speed from 5,600 to 600 metres per second before the parachute opened. For 224 seconds it did what no machine had done: measured the Martian atmosphere from inside it. The profile ran from about 25 km altitude at 150 K down to a surface at 230 K, with a pressure near 6 millibars and a hint of more water vapour than anyone expected. At 09:11:05 UT, with the retrorockets about to light, it stopped — probably hitting the ground near Margaritifer Sinus at something like 61 metres per second. Much of what it had already sent turned out to be unreadable.
Mars 7 was the cheap failure. Its descent module let go four hours early on 9 March, and bus and probe went by the planet together, 1,300 km wide.
The common cause was a transistor. The 2T-312s in the onboard computers came from a batch built with aluminium contacts instead of gold-plated ones, and over seven months in vacuum they degraded — taking out Mars 4's insertion burn, scrambling Mars 6's telemetry, and very likely triggering Mars 7's early separation. A parts substitution invisible on the ground destroyed a fleet in flight.
Then the Soviet Union stopped. Having launched at very nearly every window since 1960, it sent nothing to Mars for fifteen years, until Phobos 1 and 2 in 1988. Mars 6's 224 seconds stayed the only direct measurement of the Martian atmosphere until Viking 1 came down in July 1976.