On Aug 24, 2008, at 11:35 AM, Bill, W6WRT wrote:
> That is true when AFSK is done properly. Done improperly, you can tell
> the difference easily.
The same can be said about poorly generated FSK.
When not waveshaped, an FSK (with instantaneous rise and fall times)
is very wide. It is only about 20 dB down from the carrier at 220 Hz
away from the mark and space tones, and barely 40 dB down 2 kc away
from the RTTY signal. The quality of the signal from an FSK
transmitter depends on the attention paid to FSK keying waveforms.
On the other hand, good AFSK generators can apply waveshaping in
software and you are limited only by the IMD performance of the SSB
transmitter.
See for example, Alex VE3NEA's article "CW Waveshaping in DSP
Software" in the May/June 2006 issue of QEX. Alex, by the way is the
author of the "CW Skimmer" and some of us might remember his
comparative demodulation curves for some RTTY software on this
reflector a few years ago.
RTTY behaves the same way as described by Alex in his QEX article.
You are turning on and off two different carriers, instead of a single
one.
As someone who writes my own software to work digital modes, I prefer
SSB modulation ("AFSK") since I can more precisely control the signal
which I transmit. With FSK, I am at the mercy of the rig manufacturer
as to how they control the keying waveforms, and in many cases such as
FSK done in DDS chips, they are opaque to the experimenter.
73
Chen, W7AY
P.S. I highly recommend Alex' QEX article mentioned above for anyone
even remotely interested in digital modes.
More recently, another very interesting article pertaining to digital
modes is "Signal Resilience to Ionospheric Distortion of HF Digital
Chat Modes" by HB9TPL in the November/December 2007 QEX. The latter
article shows very clearly how PSK31 beats the pants out of RTTY when
there is no multipath and flutter, but RTTY with sufficiently large
power will still punch through flutter while not a single character
will get through multipath no matter how much power you use on PSK31
(which is sort of obvious since the phase error is a constant no
matter how much power you use).
_______________________________________________
RTTY mailing list
RTTY@contesting.com
http://lists.contesting.com/mailman/listinfo/rtty
|