Help former CB'er understand how the waterfall works!!!

[koØm]

I need a little help understanding our wonderful and wondrous Digital communications modes. First off, consider me an OF who started out on AM in the 70’s, went to SSB in the 1980’s, CW in 1985 and, the rest is history.

Help me with this “Waterfall” display that is interfaced to the soundcard of my PC; I am familiar with Pan-Adapters and Spectrum Analyzers but, how does this soundcard generated a video display to pinpoint a 31 hertz wide signals and transmit a signal on that exact “RCH” of a frequency so that the other guy’s soundcard can decode my message sent thru my (pretty wide) SSB filter?

Although I’ve messed around with it a little I was not really impressed with sending the canned Spam that the macros offer and, I really didn’t get into radio to stare at a keyboard and type, it just isn’t “radio”.

Can I interface a “Speech-to-Text” program to any of the popular Digital programs and turn it back into ‘communicating’?

ETA: I cross-posted this to 'Da Zed looking for answers.

.

[w3bny]

This may help a bit.

http://www.mymorninglight.org/ham/psk.htm

[WØTKX]

When you talk on SSB, you produce radio energy of a range frequencies that is as wide as your voice is, keeping in mind that the transmitter will only transmit as wide as it is designed (or set) to. If you transmit a single tone over the microphone (same SSB scenario), you will produce radio energy at a single frequency. PSK is Phase Shift Keying, PSK-31 uses two tones very close together (+/- 15Hz) from the center transmit frequency. You can hear this in the 30 Hz warble (frequency change) when you listen to it.

The waterfall shows the whole audio passband of your soundcard, 30-100 hz to 18-25khz. This varies in the quality of the soundcard and such. The transmitter will only produce a tone within it's transmit passband, typically 2.5 to 3 Khz wide. So you set things up in the software so your waterfall is as wide as your transmitters passband. Then you can transmit or "decode" (receive is wider than decode, it's the whole waterfall) anywhere in that width by clicking on a spot in the waterfall.

Hope that helps.

[ab1ga]

Help me with this “Waterfall” display that is interfaced to the soundcard of my PC; I am familiar with Pan-Adapters and Spectrum Analyzers but, how does this soundcard generated a video display to pinpoint a 31 hertz wide signals and transmit a signal on that exact “RCH” of a frequency so that the other guy’s soundcard can decode my message sent thru my (pretty wide) SSB filter?[/FONT]
...


The sound card is acting like the input stage of an FFT-based spectrum analyzer.

The output of your rig is basically a the slice of the spectrum as wide as your rigs receive filter, frequency shifted down to audio; the lower edge of your rig's passband is at about 0Hz.

Your rig's audio output is fed to the sound card, which digitizes the signal, and then performs at least two digital signal processing functions:

1. A spectrum analyzer function. In the old days, spectrum analyzers used swept filters and either storage tubes, long-decay phosphors, or film cameras to convert amplitude vs. time to amplitude vs. frequency data. Today's spectrum analyzers collect time domain data in chunks, convert that to frequency domain data using an FFT, and then display that data on the computer screen. The longer the chunk of data collected, the shorter the screen update rate, but the better the frequency resolution.
In your rig, a slice of the spectrum as wide as your receive bandpass is frequency-shifted to the audio region, with the low end at about 0Hz, but not necessarily spot-on. The sound card performs the digitization of the analog signal, and sends a data stream to the computer, which then displays the frequency domain data. You're looking at your whole receive bandpass at one time, but with narrow digital modes more than one station's signal will fit into an SSB or even CW bandpass.

2. When you click on a specific signal, the software configures digital filters to pick that narrow slice out of the audio passband and send it to the data decoder. The center frequency of the filter depends on the frequency you clicked on with the mouse. The bandwidth of the filter is determined by the mode you're running. Both of these parameters are configurable on the fly in software, which is why digital filters are so popular. But a warning: The digital filter is acting at audio, not in the earlier stages of your receive chain, so if you're getting front-end overload desensitizing the receiver, digital filtering won't help you one bit.

3. After the digtal filter, there are decoding routines in the software which convert the signals into character data; a single piece of software can decode as many modes as the programmers want to put in, it's all numbers from here on.

4. On the transmit side, the text information is converted into a synthesized audio signal in the sound card, at a frequency which matches either the received signal you clicked on, or any number you care to input to the software. That goes to audio in. If you aren't accidentally running split, you should be zero beat with the received signal you clicked on.

This is a coarse overview, and at each step there are probably several different implementations possible, but since you mentioned you understood spectrum analyzers I thought it best to stick to that idea.

73,