Hi At least back when dirt was new and I got my first phone license the accuracy required for AM broadcast was very loose. What drove them to OCXO's was the need to use X cut crystals rather than AT's. The age of digital dividers had not yet dawned, so if you wanted 500 Kcps you used a crystal at that frequency or some sub-multiple of that frequency.
Bob On May 21, 2011, at 7:43 PM, Lamar Owen wrote: > On May 21, 2011, at 2:51 AM, cook michael wrote: > Le 21/05/2011 08:30, Robert Darlington a écrit : >>> Guys, I gotta ask, what does this have to do with time keeping? Am I >>> missing something? >>> >>> -Bob >>> >> I know what you mean. I was desperately fighting down the urge to reply to >> Lamar's post to query the significance of cows and horses. > > My apologies for drifting even more off-topic than the OP. > > However, there are some serious timing issues present in both AM broadcast > phased arrays. Much of the same techniques are used. > > Obviously the first one is basic oscillator stability. Most of the AM > transmitters I have seen have had ovenized oscillators, with assorted > stabilization circuits. The FCC's requirements aren't quite as stringent to > require more than oven-stabilized quartz, but newer digital stuff does > require much more stability. > > The second is the need for accurate phase monitoring of a phased array. This > gets us into phase-coherent transmission line issues, dielectric variance > (with accompanying change in propagation velocity), as well as being able to > accurately monitor the phase of the RF (at up to 1700kHz) to the FCC's > precision requirements. Can you imagine the precision timing/ frequency > issues an 11-tower (ten phase measurements) phase monitor could have? > > The third is historical, but this group of all groups should grasp some of > the fundamental issues with the old CONELRAD system. The basic idea was to > throw off incoming missile timing and aiming by taking all radio stations on > the AM band away from their normal frequency and to either 640kHz or 1240kHz, > whichever was the farthest away from the station's ordinary frequency, and > 'timeslice' the stations all with the same audio program, on the same > frequency, but at different synchronized times. The wikipedia article goes > into more depth. > > And the fourth area is that of synchronous AM repeaters, to extend an AM > station's coverage using a phase-synced transmitter located at some distance > away from the main transmitter but on the same frequency. > > I'll leave as an exercise the explanation of selective fading in AM, due to > ionospheric scatter. > > There are other disciplines that benefit greatly from techniques that > 'time-nuts' take for granted; high-end analog to digital and digital to > analog converters, for instance, benefit from non-PLL stable clocks to reduce > jitter (at 24 bit samples clock jitter is a significant noise/distortion > issue, at the converters). > > So it is tangental, but just barely so, and I apologize for my off-topic > contributions....._______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
