stairs to the apartment level, he met many of the men who had been laid off. None of them said a word, but all of them wore bright, knowing smiles. By Monday morning, however, Burbank was himself again. The rebuff given him by Don Channing had worn off and he was sparkling with ideas. He speared Franks with the glitter in his eye and said: "If our beams are always on the center, why is it necessary to use multiplex diversity?" Franks smiled. "You're mistaken," he told Burbank. "They're not always on the button. They vary. Therefore, we use diversity transmission so that if one beam fails momentarily, one of the other beams will bring the signal in. It is analogous to tying five or six ropes onto a hoisted stone. If one breaks, you have the others." "You have them running all the time, then?" "Certainly. At several minutes of time-lag in transmission, to try and establish a beam failure of a few seconds' duration is utter foolishness." "And you disperse the beam to a thousand miles wide to keep the beam centered at any variation?" Burbank shot at Channing. "Not for any variation. Make that any _normal_ gyration and I'll buy it." "Then why don't we disperse the beam to two or three thousand miles and do away with diversity transmission?" asked Burbank triumphantly. "Ever heard of fading?" asked Channing with a grin. "Your signal comes and goes. Not gyration, it just gets weaker. It fails for want of something to eat, I guess, and takes off after a wandering cosmic ray. At any rate, there are many times per minute that one beam will be right on the nose and yet so weak that our strippers cannot clean it enough to make it usable. Then the diversity system comes in handy. Our coupling detectors automatically select the proper signal channel. It takes the one that is the strongest and subdues the rest within itself." "Complicated?" "It was done in the heyday of radio--1935 or so. Your two channels come in to a common detector. Automatic volume control voltage comes from the single detector and is applied to all channels. This voltage is proper for the strongest channel, but is too high for the ones receiving the weaker signal; blocking them by rendering them insensitive. When the strong channel fades and the weak channel rises, the detector follows down until the two
Prev.
Back
Top