An Analysis of a HSMS QSO between W8WN and K0SM

by Doug VE5UF



Date: Tue, 07 Oct 1997 22:25:10 +0000
To:    Rein W6/PA0ZN 
From:  Shelby, W8WN 
Subject: Re: 4000 lpm ping
From:  Doug, VE5UF 
 
   At 03:30 AM 10/6/97 +0000, K0SM, Andy wrote:
  
   Here's a good one I got this afternoon.  I hit the RIT at the end, so
   that explains why the tone goes up.

    Andy K0SM EN10

    Attachment: 
    
     W8WN on HSMS, part of his received signal at K0SM [.wav]


 That's a great ping, Andy.  Looks like the SNR was 20dB or so.  A joy
 to decode by any means.  Thought you guys might be interested in the 
 attached GIF files.  Shel/Andy - One of you should be adding a 
 description of this 4000LPM work to Rein's page.  Feel free to use
 these GIFs if you want.

    CoolEdit Display #1: HSMS Envelopes as received by K0SM  

    CoolEdit Display #2: Closeup of short dots and dashes 
    
 I think the closeup points out the difficulties in using 4000LPM with 
 anything but high-SNR pings.  As Andy already pointed out, there's
 only one sinewave (plus a little bit) in a dot at this data rate and
 it will take a sophisticated decoder to pull these little guys out 
 of the noise.  Perhaps that's what is intriguing about Andy's graphical
 decode scheme using COOL96 - it adds tremendous sophistication in the
 form of human intuition and decision-making which would be extremely
 difficult to replicate in a computer program.
 
 Another point, somewhat more technical - 
 In the spectral analysis, the "sidebands" of the CW beatnote "carrier" 
 are visible and if you play around long enough, you can determine that
 for this one, the sidedebands occur at 1260Hz (the carrier) +/- 315Hz
 and again at +/- 630Hz from the carrier and again at +/- 945Hz from 
 the carrier.  The 315Hz number is very nearly the dot repetition rate
 at 4000LPM.  The implication here is that one needs to recover the
 carrier plus at least the first set of sidebands and preferably the 
 second set as well to be able to extract the info from the carrier since
 this is classic AM in the form of On-Off modulation of that beatnote.
 
 To get the second set of sidebands requires at least a 1300Hz bandwidth
 and to get the third-order sidebands requires an additional 600-700Hz
 of receiver bandwidth.  Tough to maintain a good SNR with a full 2KHz of
 receiver bandwidth but with a loud ping like this, it just doesn't 
 matter :)   Unfortunately, there won't be many of these on most paths.
 
 On the bright side, it wouldn't take many loud, short ones to complete a
 QSO, though, and maybe you guys could be using 30-second sequences to
 double the odds of a useable short ping happening in any one particular
 direction during a sked.  Just a thought...
 
 Cheers
 
    Doug  VE5UF

    Receiver Pass Band at K0SM [CoolEdit]

    4000 lpm FFT Spectrum Analysis [CoolEdit] 








 Return 



Comments:  Rein, W6/PA0ZN 



 Top Page