Editor's Note...I originally reported suspect data from two of four ECO sensors. NASA now says data from just one sensor was suspect.

Launch director Mike Leinbach scrubbed today's planned launch of the shuttle Discovery two-and-a-half hours before takeoff time because of data indicating one of four hydrogen fuel flow sensors in the ship's external tank was not operating properly in a test.

NASA's flight rules require all four to be working normally for a countdown to proceed because the sensors control how the ship's main engines shut down when the shuttle reaches space. In certain failure scenarios, the engines could run the tank dry - or shut down early - leading to potential catastrophic failures.

It was not clear what caused today's problem, but the sensor system has encountered unusual trouble in recent weeks, possibly due to suspect transisters in an electronics "black box" in Discovery's aft compartment. The so-called "point sensor box" aboard Discovery included eight transisters from a suspect lot, sources said.

Discovery commander Eileen Collins and her six crewmates - pilot James Kelly, flight engineer Stephen Robinson, Andrew Thomas, Wendy Lawrence, Charles Camarda and Japanese astronaut Soichi Noguchi - were in the process of strapping in when Leinbach radioed the bad news at 1:32 p.m.

"The vehicle, the ECO sensors, for some reason did not behave today, so we are going to have to scrub this launch attempt," he said. "So once we develop our scrub-turnaround plan we'll get that back to you. I appreciate all we have been through together, but this one is not going to result in a launch attempt today."

A new launch target was not immediately announced.

It was a frustrating disappointment for Discovery's crew and for thousands of tourists, area residents and the shuttle launch team, which labored virtually around the clock for two-and-a-half years to ready the shuttle for NASA's first flight since the Columbia disaster. Throughout today's smooth running countdown, the primary concern had been the weather, with rain showers popping up across the Kennedy Space Center as launch time approached.

It was doubly frustrating to mission managers who thought they had resolved earlier problems with the engine cutoff - ECO - sensors that played a role in a decision to equip Discovery with a fresh fuel tank.

NASA originally hoped to launch Discovery in mid May, but the flight was put on hold in the wake of an April 14 fueling test. During that exercise, two of the four sensors in the shuttle's original tank failed to operate properly and the flight rules require four-of-four for launch.

The cutoff sensors are used to make sure the shuttle's three main engines do not run out of hydrogen while running, which would cause a potentially catastrophic oxygen-rich shut down.

"We really don't want the engine to be running at high speed and suddenly run out of fuel," astronaut David Wolf explained from the launch control center. "That can lead to a devastating breakdown of the engine, even uncontained failure of the engine if that should happen.

"Therefore, we have four sensors on the vehicle and it requires two of those to detect a low-level fuel in order to cut off the engines early before they would run out. This only comes into play in special conditions on the ascent, not a nominal ascent, so it's essentially a backup for a backup situation."

While the odds of needing this sensor in flight are low, "in the proper spirit of safety to any really feasible or projectable failure we want a full system, certainly at the point of launch."

In the wake of the April 14 tanking test and a subsequent test in May, engineers decided to replace Discovery's external tank with one being prepared for use by the shuttle Atlantis in September. The decision was made because of the sensor issue, because of concerns about ice buildups on a liquid oxygen line and because of a suspect pressure relief valve.

Troubleshooting the ECO sensor problem, engineers replaced all of the electrical cabling in Discovery that routes data from the sensors to the shuttle's computers and tore down the point sensor box in the shuttle's engine compartment that routes the data to the computers.

The original box was taken out after the April tanking test and disassembled. A box from the shuttle Atlantis was installed for the second tanking test and while it worked normally, it experienced problems later. Engineers tested a sensor box removed from the shuttle Endeavour, which ultimately was installed aboard Discovery. One of the other units will be reassembled, tested and re-installed in Atlantis.

The original failure was categorized as an "unexplained anomaly," meaning engineers never fully resolved what triggered the failures. But they were confident the hardware aboard Discovery was ready to go.

During a news conference Tuesday, NASA Administrator Michael Griffin said "ECO sensor issues are not so much a matter of the box on Discovery, we think that's good because that box has been put through an entire (testing) cycle and it's good to go. We have had more failures on some other boxes than we are comfortable with, which does lead to questions involving aging orbiter avionics."

Additional coverage for subscribers:
VIDEO: TODAY'S LAUNCH IS SCRUBBED PLAY
VIDEO: ASTRONAUTS DEPART QUARTERS FOR LAUNCH PAD PLAY
VIDEO: CREW DONS LAUNCH SPACESUITS PART 1 | PART 2
VIDEO: ASTRONAUTS GATHER FOR PRE-LAUNCH SNACK PLAY
VIDEO: PAD'S ROTATING SERVICE STRUCTURE ROLLED BACK PLAY
MORE: SHUTTLE RETURN TO FLIGHT VIDEO LISTING
SUBSCRIBE NOW