[RTTY] Distortion

[Kok Chen]

On Aug 24, 2004, at 9:23 PM, Bill Turner wrote:

> I would be wary of trying to narrow the RTTY signal by reducing the 
> sidebands.
> RTTY is a form of FM (think of a carrier modulated by a square wave) 
> and as
> such, sidebands are necessary to permit the receiver to recreate the 
> original
> signal.  Reducing the sidebands causes rounding of the square wave and 
> thereby
> reduces the number of milliseconds the output wave spends at full 
> power.

Without knowing it, you probably have already worked some on-air 
signals produced this way, Bill :-).

 From the demodulation viewpoint, not many sidebands of 22.7 Hz 
(=45.45/2) are needed to get a waveform that integrates in a matched 
filter to close to that of a full bandwidth FSK signal.  The 9th 
harmonic of the keying signal, 204 Hz away, is almost 20 dB down and 
thus at very low FM modulation index (thus the component produced by 
the 9th order Bessel function is small).

 From the bit clock extraction viewpoint, my experiments with cocoaModem 
indicates that you do not need more than the third harmonic to get 
excellent bit clocks.  Anything more than the 7th is essentially wasted 
power, but the -17 dB signal can still QRM the heck out of the adjacent 
station.

Additionally, even with regular on-air FSK signals, your sound 
card/TNC/TU is already seeing narrow signals if you are using narrow IF 
filters for reception.  I regularly copy RTTY with a cascade of two 250 
Hz INRAD filters.  Only occasionally, a 400 Hz/500 Hz cascade provides 
better copy than with the 250 Hz IF filters.

> A 170 Hz shift signal requires at least 250 Hz of bandwidth for good 
> linearity,
> and 300 Hz is probably better.

I just checked, my AFSK transmit bandpass is a 128-tap FIR filter that 
is 510 Hz wide (it just includes 7th harmonic of the keying signal to 
each side of the mark and space).  I will probably reduce it someday 
after I have time to experiment more, although 510 Hz is already very 
narrow compared to an unconstrained RTTY signal.

Even though generating a clean signal on the air can be very 
satisfying, being clean is not necessarily good strategy in a contest.  
Generating a narrow signal allows someone else to feel they can squeeze 
closer to you. The problem is that his signal may not be as clean and 
therefore he QRMs you even if you don't QRM him.

73
Chen, W7AY