I have been thinking about this problem on and off over the last couple of days.
Would it be better to take the absolute value rather than squaring the signal? I might try some tricky but impractical analog sampling and/or synchronous demodulation recovery method but the Costas loop looks awfully easy to do without any digital signal processing. On Thu, 27 Sep 2012 16:45:16 -0700 (PDT), "J. Forster" <[email protected]> wrote: >You cannot put a narrow filter before the squarer for reasons previously >cited. In a low S/N area, squaring just makes matters worse wrt dynamic >range and clipping. > >-John > >========== > >> Hi >> >> A PLL locks to phase. If the phase switches by 180 degrees, the phase >> tracking switches signs. There's no way to track that. You either need to >> double the frequency (and thus eliminate the modulation) or demodulate the >> signal and lock to the result. If you simply put up a real narrow filter >> and hit it with continuous 180 degree phase shifts, the output will be >> nothing at allÂ… >> >> Bob >> >> >> On Sep 27, 2012, at 7:11 PM, [email protected] wrote: >> >>> Various comments - >>> >>> Hal mentioned SNR for the scheme I suggested. A PLL can be a coherent >>> demodulator of arbitrary >>> bandwidth. Thus the PLL at the output of the doubler can have a small >>> bandwidth since at that point >>> there is no PSK, it having been removed by the doubler. So given a >>> stable VCXO you can probably get down >>> to 1 Hz and thereby achieve a good SNR. There is a lot of stuff out >>> there on phase tracking receivers >>> that do exactly that. You know the frequency so the loop does not have >>> to search far and the BW can be increased >>> for acquisition and closed up for tracking. >>> >>> On writing reams of code - my point was that it is not required to used >>> the admittedly more powerful software >>> techniques to do this job - I noted that one reason to write reams of >>> code is for the fun of it, this is after all >>> a hobby. >>> >>> GPS Antenna Siting - >>> >>> Lets not make this so hard. Mine is at 6 ft elevation and is blocked to >>> an elevation angle of 20 to 30 degrees by a house >>> within 15 ft and a forest of trees. I have room and no restrictions but >>> I also have severe thunderstorms - so the house >>> plays lightning protect for the antenna. My T bolt tracks a Rb to better >>> than 1e-12 over 24 hours with no serious 10 MHz phase bumps >>> as plotted on a strip chart recorder. >>> >>> Sooooo - >>> >>> Put your antenna at 6 ft in back yard. Start out on a photo tripod - who >>> is gonna notice? >>> set up a t bolt at EL=5 AMU=0 Damping = 1.2 and Time Constant = 100 sec. >>> get the t bolt manual >>> get Tbolt monitor >>> get Lady Heather and read all that stuff. >>> >>> Run Lady Heather antenna survey (command SAS) for at least two days - >>> you get a map of signal level in dBc vs elevation >>> Reset the Tbolt elevation mask to reject anything that is shown as >>> blocked using the signal >>> level map. Likewise experiment with the AMU setting to reject the weak = >>> poor signals. Mine works good >>> with AMU all the way up to 10 as fewer good satellites are better than >>> lots of weak ones. >>> >>> The satellites are in high orbits so masking those below 25 degrees is >>> OK and the AMU sets the acceptable signal >>> level - at AMU 10 my setup throws out those below about 40 dBc - the >>> strong guys go up to 50. This is a function >>> of you antenna performance so some experimentation is required. >>> >>> -73 john k6iql _______________________________________________ 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.
