[RTTY] Yep Boooo

[Kok Chen]

On Jul 22, 2011, at 7:56 AM, Tom Osborne wrote:
> It can't be the setting on the radio or AFSK would be SSB, also--the function  is the same.

Direct FSK is also just an FM mode.  

Direct FSK and AFSK are just two different ways of generating a frequency shifted signal on the RF spectrum.  The Direct FSK uses a brute force shifting of the carriers, while the AFSK approach is more mathematical.

AFSK makes use of what is called the Convolution Theorem in mathematics (http://en.wikipedia.org/wiki/Convolution_theorem).

When you multiply two signals in the time domain, in the frequency domain they look like the convolutions (http://en.wikipedia.org/wiki/Convolution) of the Fourier transforms of the two signals.  In this context, you can visualize "Fourier Transform" as "spectrum."  

Now, if one of the signals is a pure sine wave, then the convolution theorem simplifies to a case, there the second signal is simply shifted up in the spectrum by the frequency of this oscillator.

One way this "multiplying in the time domain" is done is through the use of a mixer.

A mixer is also called a product detector, a term that is less confused with the "mixers" use in the audio recoding world -- I am sure you had stumbled into that in the Windows world when playing with sound cards.  The audio mixer is not a multiplying device, but simply produces the sum of two or more signals at different gains.

The multiplication process actually produces two "sidebands" at the output, one being a "mirror" of the other.  The sine wave oscillator itself does not appear at the output.  We call this "mirrored" signal the double sideband suppressed carrier (DSB) signal.  If you now also filter away the lower portion of the mirror, you end up with an upper sideband suppressed carrier (USB) signal.  If you filter away the upper component, you end up with a lower sideband suppressed carrier (LSB) signal.

Engineers realized early on (through the convolution theorem, actually) that to transmit human voice (or music) on the RF spectrum, you do not need the carrier (if you design an appropriate demodulator).  Indeed, you do not even need to transmit both the sidebands, and therefor get further improvement in SNR while reducing spectrum usage.

So, voice started to be passed through an SSB modulator.

In essence, and this is all important, so I will say it slowly -- SSB_does_not mean_voice_transmission!  SSB is just one way that people use to modulate a voice signal into the RF spectrum efficiently.

Similarly, you can use SSB to take an audio FSK signal ("AFSK") and shift it into the RF spectrum with the use of an SSB transmitter.

The pure beauty of a SSB modulator is that the input can be as complex as you want.  So, you can use SSB to send voice, you can send CW (J2A emission mode), you can use it to send SSTV,you can send it an audio PSK signal,  you can send MFSK, and you can send it an audio FSK signal.  You can even send a windowed AFSK signal, which reduces co-channel interferences between two RTTY signals.

What a SSB transmitter produces is a function of what you feed it.  Feed it voice and you get voice in the RF spectrum.  Feed it an audio SSTV signal and you get SSTV in the RF spectrum, feed it audio FSK (AFSK) that uses 5 bit Baudot character encoding and you get RTTY.

73
Chen, W7AY