On Jul 6, 2010, at 10:06 AM, Jim Brown wrote: > Because MOST fading is > the result of cancellations between a direct wave and a reflected wave > arriving at some particular point.
HF fading is actually a very interesting phenomenon, and definitely not just a case of signal cancellation. If you decompose an RF signal into in-phase and the quadrature components, and then allow each component to go through a scattering function that has independent Gaussian statistics, the resultant signal vector has what is called the bivariate Gaussian distribution. The most interesting thing about this is the modulus of a bivariate Gaussian has a ta, da.... Rayleigh Distribution! [Gaussian statistics (the "Bell curve") is something very common in science, being the direct consequence of something in mathematics called the Central Limit Theorem.] Phenomena such as flat fading and selective fading (the reason why we need synchronous AM detection and why we use two tones in RTTY) that we encounter on HF occurs in a Rayleigh channel even when the signal is not multi-pathed. Multipath fading (also called Rician fading) is the result of cancellation of signals from different paths. On HF, the scattering function of the ionosphere is enough to produce fading, without the need of multipath. Now, if you want very rapid fading (which we commonly classify as "flutter"), that is a different story -- I have only been able to produce it by introducing multiple paths. If you run an HF Channel Simulator such as PathSim (on Windows) or cocoaPath (on Mac OS X), you will notice that both flat fading and selective fading occur when the ionospheric model consists of only a single path. Both of these programs are free (and written by hams :-). Way back when, Johan KC7WW had written a channel simulator for Linux, but I haven't seen mention of it for a while now. Russ AA7QU has made some recordings of what CW and SSB sound like through an HF channel simulator (you'll need to excuse Russ for some of his proselytizing, he is a good friend and a friend of the Mac OS :-): http://adventure-radio.org/wiki/index.php?title=OOK_vs_21st_Century Indeed, one way you can lab test a receiver for HF conditions and get repeatable results, is to run a signal (CW, RTTY, etc) through an HF Channel Simulator and into an SSB transmitter (use a dummy load, or people who listens would think there is a solar flare on :-). Then tap off the transmitter output into the receiver under test. Channel simulators not only can model the ionosphere but can add a known amount of noise to model different signal-to-noise ratios. 73 Chen, W7AY Some quick references (you can find many more using Google): http://en.wikipedia.org/wiki/Normal_distribution http://en.wikipedia.org/wiki/Rayleigh_distribution http://mathworld.wolfram.com/BivariateNormalDistribution.html http://en.wikipedia.org/wiki/Rician_fading http://en.wikipedia.org/wiki/Rice_distribution http://en.wikipedia.org/wiki/Central_limit_theorem http://homepage.mac.com/chen/w7ay/cocoaPath/Contents/technical.html If you have access to IEEE Transactions (libraries or if you are an IEEE member) I highly recommend reading the Watterson paper that I referenced in the cocoaPath site. ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
