[TVARC] WJST-X: Just How Much Bandwidth is Being Decoded?

[Anthony Hackenberg]

Hi Members of TVARC's Maillist (especially those interested in FT8 via
WJST-X,

My basic question is just how much bandwidth (range of frequencies) is
actually being decoded by the WSJT-X software and what determines it?  Is
it the entire waterfall's window width (range of frequencies)? OR, is it
more complicated, and really depends on your rig's "bandwidth" (filter)
setting?

ONLY read on if you want my overly detailed description of my amateurish
experiments to try and determine an answer, which at best, IMHO, is
inconclusive.  Hopefully, one or more readers of this posting will have the
real (& right) answer.

My inconclusive experiments appear to say WJST-X's decoding range really
depends primarily on your rig's bandwidth (filter) setting and not on the
waterfall window's width.  However, since my knowledge and experience with
the software and it's underlying technology is extremely superficial, I
cannot categorically say my experiments proved anything.

Here's what I've observed watching the waterfall display and tweaking my
rig's bandwidth settings. When using WJST-X software with my ICOM IC-7610
rig, this software takes control of my rig.  It puts both of my rig's 2
independent receivers into "USB-D1" mode & into "Split" frequency
operation. So, pursuant to the Baker Island DXpedition's band plan for 17
meters, the "RX" one is set to a dial frequency of 18.095.00 MHz; while the
"TX" one is set slightly less at 18.094.50 MHz.  However, the software also
selected the "standard" filter ("FIL2") for both receivers and on my rig
this selected filter's bandwidth is ONLY 500 Hz!  On WJST-X's waterfall
display (width set to display a frequency range of about 4 KHz), I only
noticed signals (red/pink/yellow cascading) in a narrow 500 Hz part of the
waterfall display.  In normal (not FT8 DXpedition mode), I never really
paid attention because in normal mode, your transmitted signal is going to
be relatively close your receive frequency (the other guy/gal's transmit
frequency).  However, in FT8 DXpedition mode, they are VERY fall apart (the
FOX is below 1000 Hz & the HOUNDS are above 1000 Hz).  My (hound) TX
frequency (red bracket-like symbol) on the waterfall display fell clearly
inside this narrow 500 Hz range; but my RX frequency (green bracket-like
symbol) where I was listening for the fox (his/her transmit frequency) was
down near 300 Hz (clearly way outside that 500 Hz range).  Thus, there was
a much larger span between these 2 frequencies compared to when you're
operating in normal (not FT8 DXpedition) mode. As a consequence ONLY 1
frequency (my hound TX frequency @ the waterfall's red bracket symbol) was
within the rig's pass band (bandwidth); while the RX frequency (@ the
waterfall's green bracket symbol), where I was trying to hear the fox, fell
way OUTSIDE that 500 Hz rig bandwidth.  I believe (but cannot prove) that
WJST-X was only decoding signals that fell within this narrow 500 Hz rig
bandwidth (passband).

So, now the question of just what determines how much frequency range is
WJST-X really decoding and what does it depend on becomes REALLY
important.  The only thing I was seeing in the waterfall that looked like
"real" signals was an intense band of red/yellow cascading waterfall but
this ONLY appeared within that narrow 500 Hz bandwidth.  Nothing that
looked like signals appeared anywhere else on the waterfall display.

By trial & error, I finally discovered how to adjust the bandwidth settings
on my rig.  FYI, on the ICON 7610, you depress & briefly HOLD the "Filter"
button (near the top right side under the "Twin PBT" knob) until the
bandwidth setting display pops-up on the bottom half of the rig's main
screen. I selected "BW" on this display and then "FIL1" (which was
originally set to 1.2 KHz).  I used the main tuning knob to increase it
from 1.2 to 3.6 KHz (my rig would not let me go above 3.6).  So, now my
SPLIT operation, was configured like this on the rig:

   [Receiver side]   SPLIT      TX [Transmit side]
   VFO USB-D1 FIL1              VFO USB-D1 FIL2
   18.095.00 (MHx)              18.094.50 (MHz)

So, now the first independent receiver in my rig is using the revised
"FIL1" bandwidth setting of 3.6 KHz; while the second independent receiver
(TX frequency setting) stays with the "FIL2" setting of only 500 Hz. This
made a HUGE difference in what I viewed on the waterfall display.  Now,
cascading colors (hopefully signifying real signals that WJST-X was now
decoding, were in evidence across much more of the waterfalls 4 KHz range
(width).  Now, some apparent decoding was happening in the 300 to 1000 Hz
range where the fox's transmitted signal should be.

I have to thank Pete/K2PS for putting a brief but tantalizing hint in his
excellent summary of how to configure the WJST-X software in FT8 DCpediton
mode. Here's his hint that clued me into doing this investigation:

"Expand your bandwidth to up to 4000Hz if you can. That will let you get
into a portion of the band where you’ll have more chance of finding an open
spot from which to call him [the fox], since many folks [hounds] won’t be
looking up that high – but they [fox] will."


So, again, I pose the question: what determines how much bandwidth WSJT-X
is decoding across?

73,
Tony/K4QR

P.S. Sorry about the length of this posting.  It's a case where a picture,
or video, would have truly been worth a "thousand or more words" but
although I have some photos & movies of parts of my experimentation, I did
not want to attach them to this posting.  So, I had to paint a picture with
way TOO many words.



-- 
Anthony W. Hackenberg
Email: anthony.w.hackenberg at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qth.net/pipermail/tvarc/attachments/20180701/2132e126/attachment.html>