Rick, Good questions.
My only response right now is "I dunno". <g> Back to the books. The QEX article based its results on a rate of 2% character error rate. PSK-31 with AWGN needed -11 dB. Crunching the numbers that at -10 dB you need a bandwidth of 227 Hz for 31.25 bps. At -11 dB would need somewhat more. Pushed to give some kind of answer I wonder if (1) since our received bandwidth is much wider than 31.25 Hz perhaps the sidebands are helping the situation and (2) is the reported SNR accurate? Additionally, for the latter is the SNR for just the 31.25 Hz bandwidth or for the entire received bandwidth? Rud Merriam K5RUD ARES AEC Montgomery County, TX http://TheHamNetwork.net -----Original Message----- From: digitalradio@yahoogroups.com [mailto:[EMAIL PROTECTED] On Behalf Of Rick Sent: Friday, October 26, 2007 7:53 PM To: digitalradio@yahoogroups.com Subject: Re: [digitalradio] Re: QEX Article on HF Digital Propagation My understanding is that the Eb/No is more of what you would find at the antenna terminals, without the bandwidth of the receiver? Using your data on your web site, how does this relate to say, PSK31 modulation? Would the SNR also be at zero with the 31 bps baud rate with the B/C (Bandwidth in Hz divided by the Channel capacity in bps) at ~ 1.0000? Then how do you get the much lower SNR ascribed to a mode such as PSK31? ( ~ 10dB or so?) According to your chart it would need about 7 times the B/C ratio? I had thought the ratio would be somewhat fixed at about 63 Hz BW to 31 bps or around ~ 2.0000. What am I missing? The BW is actually much wider than the number we usually use for PSK31 to get the much lower SNR? How do you make a wider bandwidth for a given mode? Isn't the bandwidth based on the baud rate to begin with? 73, Rick, KV9U