Phobos 1 and 2

The Phobos project began as something much larger than the two spacecraft that eventually flew. A Central Committee and Council of Ministers decree of 8 May 1980 authorised an automatic complex designated 8K82K-F, and the technical specification finalised in early 1981, held today by the Russian state archive RGANTD and secret until recently, describes two 1F craft launching in 1984 to study Phobos from Mars orbit and from close range, followed in 1986 by two 2F craft that would bring pieces of the moon back to Earth on an 850-day return. The 2F half was abandoned in a 1985 revision and everything was concentrated on 1F. What survived was still extraordinary. The 1F bus was a new unified design nicknamed UMVL, for unified Mars, Venus, Luna, built around a pressurised toroidal instrument ring with a jettisonable propulsion unit slung underneath and a science bay on top. Its main engine, the S5.92, was later developed into the Fregat upper stage. The two spacecraft, 1F No. 101 and No. 102, launched five days apart in July 1988 carrying instruments from a dozen countries plus the European Space Agency: a French infrared spectrometer, a Swedish plasma analyser, an Irish particle detector, Bulgarian and East German contributions to the camera system. NASA supplied Deep Space Network tracking. The Soviet share of the bill came to 272 million roubles.

The endgame was the part nobody had tried. After months in Mars orbit the spacecraft was to be walked inward until it was flying over Phobos at a relative speed of two to five metres per second, roughly 50 metres up: not an orbit but a matched drift alongside, the spacecraft and the moon travelling together while Mars did the real gravitational work. From there two active experiments would attack the surface without touching it. LIMA-D would fire a half-joule laser pulse into a spot one millimetre across, vaporising the regolith, and catch the ions in a time-of-flight analyser. DION would inject a beam of krypton ions at two to three kiloelectronvolts and read the secondary ions knocked loose with a quadrupole. Then the landers would go down. DAS, the long-lived autonomous station, 67 kg with about 18 kg of instruments, rode on both spacecraft. On contact it would fire a harpoon penetrator into the ground to pin itself in place, because on a moon this small anything that lands can simply bounce back off, and then raise a platform on solar arrays for roughly a year of work: an alpha-backscatter and X-ray fluorescence spectrometer for surface chemistry, a television camera, a seismometer, the Razrez anchor penetrometer, and a Sun sensor named Stenopee to measure Phobos librating on its axis and finally settle when it will fall into Mars.

The other lander is the one people remember. PrOP-FP, carried only by Phobos 2, weighed 43 kg and is credited to VNIITransMash in Leningrad, the tank-research institute that had built the running gear for Lunokhod and the little skiing rover on Mars 3. It was a squat sphere that moved by hopping. Released from low altitude behind a stabiliser shaped like a shuttlecock, it would shed the stabiliser once it settled, unfold sprung whiskers to right itself support-side down, and then shove that support against the ground hard enough to throw itself perhaps 20 metres. Between hops it would measure: X-ray fluorescence for composition, a penetrometer and accelerometers for how hard the ground was, a magnetometer and a kappameter for magnetic susceptibility, a gravimeter, temperature sensors, a conductometer and a tilt sensor. It ran on batteries and was designed for up to ten hops over a few hours. It was designed thirty years before Japan's MINERVA-II rovers and the German and French MASCOT actually hopped across the asteroid Ryugu, and it never got to jump. Its engineering model sits in the museum at NPO Lavochkin. The name is transliterated inconsistently even in Russian: Lavochkin and Russian Wikipedia write PrOP-FP, other sources PrOP-F, and English catalogues usually render it PROP-F or simply 'the hopper'.

Phobos 1 died of one missing character. Yuri Markov, an engineer on the control team, wrote that on 25 August 1988 the gamma spectrometer team asked for their instrument to run about ten hours on 30 August, an entirely routine request. The operator writing the session program entered the address of a cell in the onboard timing system and left out a single letter. His supervisor checked the program and missed it. Nobody ran it against the mathematical model at the Yevpatoriya deep space centre or against Phobos-T, the engineering twin sitting powered up at the factory, because the task looked trivial. Uploaded on 29 August, the corrupted command did not turn into gibberish. It turned into a valid instruction to shut off the pneumatic attitude and stabilisation system. Roald Kremnev, who ran the Lavochkin plant, later explained to Science magazine why such a command existed at all: a steering test routine burned into read-only memory for ground checkout could not be erased without removing the computer from the spacecraft, and with launch close it had instead been sealed off in software, 'locked in a safe'. The typo found the key. The arrays drifted off the Sun, the batteries flattened, and the state commission later dated the moment the spacecraft stopped functioning to 31 August. Controllers called on the night of 1 to 2 September and heard nothing. The operator responsible found the bug himself, was not punished beyond being kept off the Phobos 2 shift teams, and the mission was formally abandoned on 3 November 1988.

Phobos 2 was already ailing when it arrived. On 1 November 1988 one of its two centimetre-band transmitters failed, and its computing complex kept shutting itself down, with one channel hanging repeatedly; Brian Harvey's history records that by Mars arrival two of the three computers were sick, which mattered because the design required two of them to agree before anything happened. It nevertheless entered orbit on 29 January 1989 at 12:55 UTC and spent 57 days working. It returned 37 images of Phobos from three passes, down to about 40 m per pixel and covering roughly 80 per cent of the moon; the Termoskan radiometer made four thermal and visible panoramas of the Martian equator at 1.8 km per pixel, the highest-resolution thermal data anyone had for Mars at the time; the French ISM spectrometer took 45,000 near-infrared spectra of the planet and 400 of Phobos; and ASPERA made the first direct measurement of ionospheric plasma escaping Mars into the solar wind, published in Nature that October. Then on 27 March, imaging on automatic with the transmitter off to save power, it failed to call home at 15:58 UTC. A weak tumbling carrier came and went between 17:51 and 18:03 UTC. The landings had been set for 4 to 5 April. Lavochkin engineers afterwards argued the approach had been too cautious and could have been made in mid-March, while the spacecraft still worked. The 1F bus flew again as Mars 96, which fell back to Earth; nothing has landed on Phobos to this day, and Fobos-Grunt, the next Russian attempt, never left Earth orbit in 2011.