No sooner were two faulty heat-sensors blamed for Monday's shutdown of the space shuttle Challenger's center engine than a $60 million telescope-pointing device capable of tracking a dime two miles away failed even to locate the sun properly yesterday.

"I am convinced that the problem we're having now is solvable," Mission Manager Leon Allen said at Marshall Space Flight Center in Alabama. "If the Instrument Pointing System doesn't work, the whole mission is down the drain and our mission to carry three ultraviolet telescopes to observe Halley's comet next March could be in peril."

The pointing system and its four solar telescopes have been described as the most important of the 13 experiments carried on the European-built Spacelab in Challenger's cargo bay. Said one scientist: "The IPS and its solar telescopes are anywhere from 30 to 40 percent of the science carried on this mission."

The shuttle, meanwhile, continues to be plagued by a sequence of technical troubles: The launch emergency and the experiment problems came about two weeks after trouble with Challenger's No. 2 main engine forced a safe, automatic engine shutdown on the launch pad seconds before liftoff July 12.

Because 21 percent of the ship's fuel was dumped to reduce weight during the climb to space in Monday's crisis, some experiments planned for this mission are being canceled and others curtailed. And the engine trouble put the shuttle in an orbit 195 miles high -- far short of the 380 miles needed to carry out all the experiments.

An experiment to measure sun-driven helium abundance has already been abandoned because there is no solar helium at the unexpectedly low altitude at which Challenger is flying. A second experiment to photograph the sun's magnetic field is not working because of a power failure in the instrument.

The German-built Instrument Pointing System was stowed in Challenger's cargo bay yesterday after it failed repeatedly to aim four telescopes at regions of the sun they were designed to observe. Flight directors blamed a filter on the IPS for letting in too much light and its computer software instructions for causing it to drift.

Flight director Granville Pennington at Houston's Johnson Space Center said the crew will be told to bypass the faulty filter on the IPS and use a new set of software instructions being written at the flight center in Alabama. He said the software used yesterday had forced the IPS to drift so badly that the crew was told to lock up the pointing system and its solar telescopes until today.

While computer engineers struggled to revive the instrument, flight engineers said they were "99 percent sure" that Challenger lost one of its three main engines after liftoff on Monday because two heat-sensing monitors failed during flight, forcing onboard computers to shut down the engine prematurely.

A spokesman at Marshall flight center said one sensor began reading fuel-pump temperatures of 1,950 degrees Fahrenheit, then suddenly failed. A backup sensor read temperatures of 1,850 degrees -- 130 degrees above the pump's "red-line" temperature and enough to have Challenger's onboard computer shut down the spaceliner's center engine.

"The data we saw later on gave us . . . assurance the engine performed the way it was supposed to," he said. "We'd seen bad sensor behavior in ground tests and on one previous shuttle flight, but this was the first time we had both sensors fail in flight."

The sensors are designed to protect the engine in case the fuel pump overheats -- a potentially catastrophic situation in an engine that is burning liquid hydrogen and oxygen and whose fuel-pump turbine blades are turning 36,000 times per minute.

"If these engines are going to malfunction, it's usually catastrophic because they're at such a high temperature and pressure that they'll consume themselves," astronaut Joseph P. Allen, a veteran of two shuttle flights, told The Washington Post.

After the center engine was shut down almost six minutes into the flight Monday, a sensor on a second engine began reading "red-line" temperatures. Flight directors quickly told the crew to disconnect the sensor before computers shut down the second engine, which could have forced an emergency landing