[RTTY] FSK is bad?

[Kok Chen]

> On Oct 25, 2016, at 2:22 PM, Thom <ki8w at ki8w.com> wrote:
> 
> and exactly how do I change that shaping?

With FSK, you are at the mercy of the rig manufacturer.  Some rigs use the "FSK" facility in the synthesizer chips, so even the rig manufacturer has no control of it, the chip manufacturer does.  And the "FSK" requirements from the chip consumers are different from ours (we are a tiny market for the chip manufacturers).

With AFSK, it depends on the software modem designer.  Among popular modems, 2Tone and fldigi are currently the only two I know that individually waveshapes the Mark and Space signals before combining them for transmission.  

The AFSK modulators that use a bandpass filter to limit the keyclicks from a composite Mark/Space signal will inevitably be wider than the signals from 2Tone and fldigi.  See here to get an idea of how wide the transmit filters need to be so that the receiver's demodulator at the other end is not impacted:

http://www.w7ay.net/site/Technical/RTTY%20Transmit%20Filters/index.html

By the way, since all filters are in a cascade (RTTY generator -> transmit filter -> receive filter -> demodulator), the above also applies to how narrow you can make a superhet's receive crystal filter before you start to impact error rates.

FSK keyclicks are no different from CW keyclicks.  They are (a) not useful for decoding, (b) they interfere unnecessarily with nearby stations, and (c) they eat up transmit power needlessly.  If your PA is limited by RMS power instead of PEP power, and given the same amount of average transmit power, a well filtered AFSK signal will deliver more useful signal power into the demodulator's filters at the receiving end.

73
Chen, W7AY