One thing that I would say is that getting an accurate calibration down in the broadcast AM band is pretty tough depending on your tools for measurement. Reason being that what you are correcting for (mainly) is error in the 200MHz oscillator.
That error gets divided down to whatever frequency you are using to calibrate. So, for example, if you calibrate at 1MHz, an error of 200Hz at the LO (200MHz) would show up as a 1Hz error ( 200Hz / (1MHz / 200MHz) ). This means that you will likely find a wide range of acceptable DDS clock values that are fairly accurate because adjusting the clock doesn't make that much difference around 1MHz. For this reason, it is best to calibrate at a higher frequency. If possible, try calibrating up in the 52MHz range for an optimal frequency calibration as the error isn't divided down nearly so far. Obviously this isn't as trivial as finding a strong signal on WWV, but for those that really want a good calibration, higher frequencies on a signal generator are better. Eric Wachsmann FlexRadio Systems > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > radio.biz] On Behalf Of Mark Amos > Sent: Monday, November 06, 2006 12:01 PM > To: flexradio@flex-radio.biz > Subject: [Flexradio] Question regarding commercial AM broadcasters' > carrieraccuracy > > Folks, > > I had the oscillator mod applied this past summer, and I've noticed that > after zero beating with WWV at 10 Mhz it stays put pretty well. > > I thought it would be interesting to take a look at the carriers of some > broadcasters and see where they're at. I tuned in their nominal frequency > in DSB and adjusted the DDS Clock Offset until I got zero beat per the > Phase > display. Most of these were local (Northwest Ohio) stations. > > Below are the results I got. > > Also, I would have thought that they would be much closer to WWV or CHU, > but > they're pretty much all over the place. However, when I plotted > broadcasters > nominal frequences (1310 to 1560 KHz) against DDS offset, I saw an > apparently periodic pattern(!) This led me to believe that the differences > might have something to do with a measurement artifact or DDS inaccuracy > or > some such. I would have expected these measurements to differ randomly if > they were due to frequency misalignment at the broadcasters. > > As expected WWV, CHU and my bench oscillator all clustered nicely. They > were > between -2490 and -2575. > > Any thoughts? > > Mark > > KHz Offset to zero beat > 1310 -1810 > 1330 -3210 > 1340 -2280 > 1370 -1825 > 1430 -1605 > 1450 -1700 > 1470 -2075 > 1490 -2375 > 1520 -2020 > 1560 -1775 > 1800 -1370 > 3330 -2490 > 7335 -2560 > 9450 -2580 > 9625 -2170 > 9985 -2430 > 10000 -2563 > 10000 -2503 (bench TCXO) > 14670 -2570 > 15000 -2575 > > > > _______________________________________________ > FlexRadio mailing list > FlexRadio@flex-radio.biz > http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz > Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ > FlexRadio Homepage: http://www.flex-radio.com _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
