[RTTY] Pseudo FSK

[Kok Chen]

On Nov 27, 2014, at 9:46 AM, Dennis Egan wrote:

> I have been a convert to Pseudo FSK for over 2 years now, using the FSKit.

FSKit does not use pseudo-FSK (pseudo-FSK is pioneered by fldigi).  FSKit actually implements full AFSK demodulation between *two* tones from the computer to threshold the FSK keying line.

Pseudo-FSK decodes a *single* on-off tone from the computer.  

As such, the FSKit should produce a Mark bit duration that is very close to the Space bit duration.  

However, if you ensure that the single pseudo-FSK tone has plenty of amplitude and the hardware detector is properly constructed (rise time == fall time, if thresholding at half amplitude), it should be close enough to what the FSKit does.  

Simple pseudo-FSK circuits may or may not produce constant Mark and Space periods -- in fact, I have seen circuits that will bias in one direction when it receives two or more consecutive Marks (thus produce a kind of deterministic jitter).

One advantage of the FSKit is that it works with any software modem that puts out AFSK.  To use pseudo-FSK, the software needs to (1) turn off one of the the two AFSK tones, and (2) hits the audio output with a tone risetime that is as short as possible. 

For those with modems like the ST-8000 that can do Mark-only demodulation, the analogy is the FSKit hardware module demodulates using Mark/Space, while pseudo-FSK module demodulates using Mark-only.

Indeed, you can feed an SSB transmitter with the on-off keyed pseudo-FSK tone and be able to decode using a Mark-only modem (or Space-only, depending on polarity).  But do that only onto a dummy load/directional coupler, since you will be transmitting with keying sidebands that are as wide as non-phase continuous FSK :-).

Gobble, gobble; remember that you are what you eat!  (And I do have lots to be thankful for.)

73
Chen, W7AY