[OZAPRS] HF APRS Frequency accuracy/stability

[Richter, Mike W]

I have limited knowledge in this area, so I need to ask this question.

Why does your frequency need to be stable to dozens of Hertz for HF APRS ?

I assume the demodulated audio tones go into the TNC and the FSK is
detected using a PLL. Assuming there is some preamble to give the PLL time
to lock, then as long as both tones are well within the receiver passband,
then they should decode easily.

Assuming a Rx passband of 300-2700 we have 2400 Hz available but to keep
away from distortions at the edges of the filters by 400 Hz, lets say we
have 1600Hz. With +/-100Hz shift the "carrier" must not drift outside a
1400Hz range of the receive radio, so the stability need is +/- 700Hz. 

But both Tx and Rx frequencies can drift, so we need to share this drift
allowance, so both the Tx and the Rx can drift +/- 350Hz. Assuming a
calibration accuracy of 100Hz, we are left with a requirement of tuning
+/- 250Hz.

This is 100's of Hertz, so if TNC's do use PLL decoding, the only
explanation I can see for needing dozens of Hertz accuracy/stability is
because Rx calibration is well off ????

Mike VK2BMM
_______________________________________________
ozaprs mailing list
ozaprs at marconi.ics.mq.edu.au
http://marconi.ics.mq.edu.au/cgi-bin/mailman/listinfo/ozaprs