Julian, The REF CAL adjustment is in 1-Hz steps same as the fine tuning resolution of the VFO. The EXREF will call for a shift to bring the radio frequency back to zero error by introducing a shift command. Typically, the TCXO does not drift much over a short time (TCXO-3 is rated to 0.5 ppm). The error depends on operating frequency and is max at 50-MHz (0.5 ppm = 25 Hz). But it is not like the TCXO is jumping that far. I watched it during the 30-min warm-up of the K3 and it typically moves 1-Hz at a time until it settles out.
So I would not expect corrections over 1 or 2 Hz at maximum and they are happening on a 4-second interval. The TCXO is running continuously and not being stepped. The steps are commands to the digital dividers in the synthesizer of the radio. For example the synth. steps when you rotate the VFO. The advantage of the reference oscillator is that it makes the K3 more accurate in frequency. Frequency stability and LO phase noise are still controlled by the TCXO and not the reference. To have the radio stability and phase noise determined by the external source it would require phase locking the TCXO to the reference signal. I do not know enough about synthesizers or PLL so I am guessing there is a momentary phase shift every time the K3EXREF commands a REF CAL change. For robust FEC codes like WSPR it will not be a problem. The sw expects to have jumps and bumps to the data stream and removes them as noise. The experts out there can correct my interpretation where it is in error. I 73, Ed - KL7UW ------------------------------ Message: 28 Date: Tue, 5 Apr 2011 14:53:00 -0700 (PDT) From: "Julian, G4ILO" <julian.g4...@gmail.com> Subject: Re: [Elecraft] K3EXREF and Trimble Thunderbolts To: elecraft@mailman.qth.net Message-ID: <1302040380604-6243940.p...@n2.nabble.com> Content-Type: text/plain; charset=us-ascii Thanks to Leigh for his detailed explanation. Also to you, Rich, for explaining how the frequency standard is applied to the K3. I still have a couple of questions. If, as I believe is the case, the K3 REF CAL has fairly large discrete steps, is there any benefit in using a reference oscillator that is many times more accurate than that? All one is looking for is something that will keep the K3 as accurate as it can be without the need to perform regular manual checks. A standard that has extra precision is not going to make the K3 any more accurate because the reference oscillator is being controlled using discrete steps. Has anyone tested the effect of these small stepwise adjustments of frequency on the WSPR mode, in which the data is transmitted using a frequency shift of about 1.5Hz? Julian, G4ILO 73, Ed - KL7UW, WD2XSH/45 ====================================== BP40IQ 500 KHz - 10-GHz www.kl7uw.com EME: 144-1.4kw, 432-100w, 1296-testing*, 3400-winter? DUBUS Magazine USA Rep dubus...@hotmail.com ====================================== ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html