BRIGHAM CITY, UTAH, DEC. 19 -- A test firing of a redesigned space shuttle booster rocket identical to those that will be used in the first post-Challenger flight was canceled today after a frustrating last-second technical glitch.

In freezing weather and with the countdown running 2 1/2 hours behind schedule because of unrelated problems, countdown commentator Roger Williams said the test had been terminated at one second before 3:30 p.m. local time.

"We are aborting the motor," he said. "We had a failure to arm our primary circuit. We had no fire pulse issued. We failed to arm the primary circuit."

Allan McDonald, a top engineer with booster-maker Morton Thiokol Inc., said the test firing, which was canceled for the day, could not be rescheduled before 1 p.m. Monday at the earliest, depending on an analysis of the problem.

The test originally was scheduled for 1 p.m. today, but officials said repeated practice countdowns overloaded a computer that records test data and forced a two-hour delay.

A minor mechanical problem, possibly the result of bitterly cold weather similar to that in which the shuttle Challenger was launched, caused an additional 30-minute postponement.

The rocket, anchored on its side in a massive test stand, was the second of at least five full-scale boosters scheduled for test firings to check the operation of redesigned O-ring joints before Discovery blasts off in June on the first post-Challenger shuttle flight. The previous rocket fired in August was delayed three days by ground equipment problems.

Each rocket is equipped with a "safe-and-arm device" that prevents stray electrical signals from igniting the boosters' propellant prematurely. McDonald said when a computer checked the device one second before scheduled ignition today to make sure the rocket was ready for firing, it found a problem.

"The indication it got was that it was not {armed}. And it shut the test down at that time," he said. "We don't know if the problem is in the safe-and-arm device itself or in the control circuit. It's unfortunate. We're very disappointed."

Royce Mitchell, solid rocket motor manager for the Marshall Space Flight Center in Huntsville, Ala., said "on a scale of this undertaking it's a relatively minor problem. It's unfortunate, but we do have spare equipment."

Once the test is completed, engineers will need several days to disassemble the giant rocket for a detailed inspection. If no surprises are found, the booster segments that will be assembled for use with Discovery will begin their trip to Florida and the Kennedy Space Center around Jan. 1.

Challenger was destroyed Jan. 28, 1986, by the rupture of a booster fuel-segment joint, a failure triggered in part by freezing temperature the morning of the ill-fated blastoff.

The temperature had risen to 36 degrees by the time Challenger took off, but investigators found that colder predawn weather had caused two rubber O-rings to stiffen, preventing them from sealing the faulty joint at ignition.

A wind-chill factor of minus 10 degrees was recorded at the Morton Thiokol test stand in the high Utah desert at 10 a.m. today, but the giant rocket was equipped with heaters on each joint to maintain the O-rings at a minimum of 75 degrees.

Mitchell described the test as the most ambitious since the Challenger exploded.