The following questions are from the WSJTGROUP reflector with some of the responses. Maybe someone will be
helped with this posting. In most cases the names and calls have been removed. This is a good example of hams helping
hams.

QUESTION:
What sort of filter selections do you use?
You need to have the filters wide enough to pass the entire set of tones
used in the mode. FSK441 uses tones from 880-2200Hz, and you'll want a
little headroom on either side. The normal SSB filter on your radio
(2.4 - 3 KHz typ.) is just fine. Limiting the bandpass really doesn't
help you any--in theory you should be able to run with the 10KHz filter
inline and the program will do just fine....in theory :-)
QUESTION:
What about noise limiting? I'm guessing DSP noise reduction would
be bad as it may distort the tones?
Depends on the radio. I haven't had any problem with the NB to
eliminate pulse noise, but others may have other experiences. I've
heard that some NB's like to eat short, loud pings since they look like
pulse noise.The DSP noise reduction is likely to cause you problems,
but you can experiment.
QUESTION:
What about AGC, fast, slow, or off?
AGC should be off--FSK441 detects pings by
looking at the amplitude of the audio. Having the AGC on defeats that!
If you can't turn it off set it to fast.
QUESTION:
What about ALC?
The most important thing is that the individual tones are sent at the
same power output. FSK441 does not require linearity, so "filling the
box" or not shouldn't matter as long as you get the same output with
each tone. I would use the least audio necessary without having the
soundcard audio at either of its extremes.
QUESTION:
I do not understand the following statement.  Try clicking around the ping with
both left and right nouse buttons and observe the decoded text that appears.   
I have played the W8WN file a few times and I have no idea what is ment by click
around the ping.
Try clicking around the ping with both left and right nouse buttons and observe the decoded text that appears.
To 'click around the ping' move the mouse cursor into the upper window
that displays the time trace. There should be vertical spikes (ie. pings)
similar to the examples shown in the manual. If you place the cursor
directly on top of the obvious spike in the time trace, hold it there and
right-click the mouse button, you will force a decode of that signal.
The decoded text will display in the text box.

If you right click on the ping it will ignore steady tones. I have seen strong bursts, which I know is multitone and
it will decode R26, R27 sometimes. The right click will ignore the carrier or whatever it sees in the multitone and
give you a decent decode.

I set W>width to 20 andTol> to 400 to start then
Tol>100 after I find the freq of the signal and tune RIT to correct to
under 100 hertz. The closer you have the receive signal the better
overall...I actually listen to the received signals and manually decode
anything I think might be a ping when possible too.
QUESTION:
Could someone let me know what to do to get Dimension 4 the time program
>to work on a XP Pro based computer?
Disable the Windows Time service. To do this, right click on My Computer,
select Manage then select Services, then scroll down to Windows Time,
double click it and set Startup Type to "Disabled".

Try this http://home.alltel.net/wb5apd/d4fix.html
QUESTION:
I just put up a 6M dipole about 40ft in the air and feel I have
correctly configured the WSJT software. The dipole works as I have
transmitted and received on it using repeater frequencies. I am
running a 756 Pro2. I can easily copy single tone messages like RRR
etc, although they may be birdies, but I cannot copy multitone text
messages. I am monitoring 50.260MHZ. I have not tried to transmit
yet. Is my problem because I am using a dipole instead of a beam or
loop? I was under the impression based on the recent QST article
that a dipole would work. Is this a lost cause or can anyone help me
get going with my current setup?

Hal, please make sure that you are not pouring too much signal into your sound card.
Keep the level down around 0 dB, and not more than 3 dB as indicated in the lower window
on the display. My experience has been, with my system and also with that of a friend
in the Dallas area, that WSJT is very sensitive to signal level.

As I've discovered as a newbie to WSJT, this is not a mode for
impatient people. I had to monitor for many hours before I received
anything, and even at that I'm not sure where it came from. A dipole
antenna on 6 will net you many contacts during E or F propagation on
SSB or CW, but for weak signal modes a few dB makes a big difference.
You'll still make contacts but it won't be as easy, or as often.

The dipole will work, horizontal polarization, but understand it will not perform
as well as a beam. There are many stations that use loops antennas for operating
mobile or portable so it will work, especially on six meters!
QUESTION:
I have been trying to get wsjt to work on my machine. I get no graph
display on any modes. I have run though the set up in the handbook
and still get nothing. It TX fine.I can not set up any recieve even using the
eme echo part of the program. I have the latest program and the update.
The machine here is a 2GHz AMD with 2 sound cards (one for running
Echolink) and the other perminatly connected for data modes.

The WSJT program expects input levels within a certain range. Have you
selected the proper input source in the mixer and set the level to the
maximum. If you are connecting your sound card interface to Line-in
and you still get no input, try connecting your interface to the
Microhone input.

Are you putting the program into "Auto is On" mode? Unless you do that it will
never receive a signal. And even then it will not display the received signal or noise
until the end of each receive period. So for FSK441 you have to wait 30
seconds (if you have used the default sequencing), and one minute for the
JT65 modes. Then the received data will appear. (In FSK441 mode you can
also use the "Record" button to engage receive for a single 30 second
period.)

This program is not like a Spectran, or other waterfall display program that
shows the receiver output in real time. When the program is receiving
signal, it will show the word "receiving" just below the graphical display.
When the receive period ends, the "receiving" message will change to
"decoding" for a short time and when that ends, the received noise or/and
signal will finally be displayed.

If you don't want to wait that long to observe what your noise level is,
switch to the EME Echo mode (F9), and engage the Measure function. Then the
received level will be displayed (and shown as a text value) once per
second. You can then easily adjust your levels to achieve the required zero
db. Switching back to FSK441 or JT65 mode, the noise level adjustment will
remain the same.
QUESTION:
I really do need some help picking out a good souncard for a new
computer for digital modes. I don't know good from bad or better from
worse and could use some seriously considered advice here. Can
somebody toss me a line here?
I really do not think it matters, but some have a higher SNR ratio than others. That is on
the audio output though which is what is used on xmit only. You can look here and see the
diff spec on some of them. http://us.creative.com/products/category.asp?
Some are also 16 bi,t and some are 24 bit in the digital porcessors. Again I doubt you would
notice any diff for digital mode stuff. Now if you are wanting the very best audio out for playing music
etc, get a 24 bit with the best SNR you can find.
Many sound card manufactors do not even list any spec's on their sound cards. Creative
Labs has always been considered the standard when it comes to sound cards. They set the
standard for all that followed. A soundblaster 16 PCI should work just fine and be compatable
with all windows systems and software.
QUESTION:
I am new to the group and WJST. I would like to try to use the
software on 6 mtrs but I only have dipoles for 80, 40, and 17mtrs, a
3element beam up 60ft, and a disconne up 30ft. The disconne is the
only antenna I can tune for 6mtrs with my Icom 756 Pro2. I have the
software running ok and using the disconne can receive RR, 73s etc
but no call signs. Am I wasting my time trying to use any of my
antennas or will I have to get a loop or beam to successfully work
this mode?
I would highly suggest not starting with an omni or a moxon type antenna. You will have a
lot better success with a small Yagi. I would suggest 5 elements, but even a 3 element six
meter Yagi will work well for you. You can buy such antennas new from companies like CushCraft
or M2 Enterprises for prices between $125 and $250. You may very well be able to find someone
selling a used antenna for much less. Also, a small antenna like this is quite easy to build
and there are construction details in various handbooks as well as on the Web.
QUESTION:
I've been trying to get my first WSJT contact via meteor scatter, and I may
have detected a potentially interesting problem. The received audio from my
IC-251A tails off dramatically above about 1200 Hz, to the point where the
high frequency tone seems to be nearly inaudible. I hear the pings, but
they don't decode, and looking at them with a spectral analysis shows a
dramatic loss in amplitude. This is apparent even when I look at channel
noise. The problem may also be present on transmit audio.
Anybody got any ideas on how to punch up the upper audio frequencies? (The
sound card is a Creative SB Live.)
Phil, this is only a guess, but I would bet that the problem is in your IC-251 rather
than the sound card. I have a IC-271 that I was using on 2M, and it will not send the
73 tone in FSK441. The audio response just isn't there. Tune in a good clean carrier
and vary the frequency so the output tone varies. Then measure as best you can, the
audio output tone level at various frequencies.

I have/had similar problem with Kenwood TS-60S I used. It rolls
off on higher audio frequencies also. You can adjust the 4 tones for TX
fairly easily as described in the WSJT manual. In short, connect dummy
load and watt meter and see what you get on the A, B, C, D tx buttons.
Select the lowest one on output (usually the D/73 tone) and then play
with "FSK441 Amplitudes" in Setup Options (F2) for the other three
tones. The settings seem to be very nonlinear so it may not take much
reduction below 1.0 to get the levels down to match the lowest response
tone. I didn't have too much trouble with the receive curve on the
Kenwood. You might try IF Shift or something similar if the set has it
to see what effect that might have on Rx response.
QUESTION:
I'm a newbie to VHF+ digital and I'm interested in taking a package along
on a trip and trying WSJT. I plan to take a laptop, a Yaesu FT-100D, and an
array of antennas to Montana in late August. My first question pertains to
the feasibility of using horizontal loops for WSJT on this venture. Would
it be worthwhile, or are directional antennas a must in order to make
contacts? I'd be able to work 6m, 2m & 70cm.

There are a number of us that are now running mobile and
portable so probably as many opinions as operators! You choice of
antennas will of course depend on a number of factors including how long
will you be in one spot to set up, size of vehicle (or boat!) to tot
them or hang them from. I use loops because I run mostly while
traveling in job related day trips so either truly mobile or in one spot
for short period of time and no good way to haul beam(s) around. Do
they work? Darn good on 6m and just getting into 2m mobile but have
worked K1JT at about 900 miles with stacked KB6KQ 2m loops up about
12-13 feet from a parking lot. Also worked AF4O with JT44 to hand him 4
new 2m grids on two runs to Memphis.

Mostly a 6m person and have about 75 completions (46 initials) from the
truck setup - about 50-50 mobile vs temp portable. I have a KB6KQ 6m
loop but really prefer homebrew copper Squalo for 6m. (except for rain
detuning, if parked - OK mobile with wind blast) Run 100w or less. Best
distance is 1150 miles with a QRO station. I've worked Tip, WA5UFH from
13 oo 15 grids I've been in.

Dave N8OC, Phil N0PB and Mike WB2FKO can share there mobile/rover
experiences and setups on 6 and 2m as can Joe, K9KNW with his recent
Bahamas trip with 2m loops that seemed to do very well for him. I've
not heard reports on anyone running UHF and above but sure there are
some out there

I have a lot of experience with portable 2m WSJT and limited experience
with 6m. I've made about a dozen grid expeditions over the last two
years. It's difficult to be certain because of variable meteor rates,
propagation, and the quality of the station on the other end, but here's
what I feel comfortable stating:

The limitations of a weak portable station can be overcome with an
excellent operating location. I have worked over 1000 miles on 144 MHz
with 35 Watts and a 5-element beam in under 30 minutes in the middle of
the afternoon, when the sporadic meteor rate is supposed to be low. I
look for places that are high in the sky, clear horizon, and free of RFI.

That said, it makes a huge difference depending on who you are running
with...I've had lots of success with K0PW, WA7GSK, WA5UFH, W3UUM, and
K6AAW...these guys are all running pretty serious stations. They
clearly make up for the deficiencies of a puny portable station.

I have experimented with power, pushing as high as 120 Watts, but haven't
noticed anything dramatic. I also tried stacking two 5-element beams on
144 MHz...not sure if it is worth the trouble. I have used exclusively
yagis on 2-meters, so I can't say anything about what happens there with
loops.
QUESTION:
I just got a new Dell laptop and the RX audio from my radio's data
port is to high to get to 0db even with the mic volume set to the
lowest setting. A friend of mine suggested using a resistor to lower
the signal then I reformed that idea changed it to a potentiometer.

My question is will the resistor or pot have a negitive affect on
WSJT ability to decode the weak signals?

I'm sure you have checked but.....
Make sure the audio software does not have the +20db boost marked on...
Move from mic input to line input if it has one....

In my experience whenever I could not lower the gain enough with the
computer mixer controls, it was due to AC hum. I suggest you enable the
laptop speakers and listen to the input you are sending it.

To answer your question though, if you really need to reduce the level a
resistor would do the job just fine and will not cause any loss of
sensitivity as long as you are able to achieve zero DB on noise.

You might consider a pot across the audio to the computer so you can adjust.

PS: Just re-read your post. You should not be using a data port to drive
the computer's audio card! You should use line output, or auxiliary audio
out, or just plain speaker or headphone out instead. This is more likely to
be your problem than anything else (unless you merely mis-spoke when you
said "Data port").
QUESTION:
This morning I ran a sked(FSK441)on 2 meters. On at least two
occasions, I heard a ping and could see it on the graph, but nothing
decoded. Once I "Left Clicked" on the ping, it decoded and there was
useful data there. Qrn is set to 2. Is there something else I need
to do as far as set-up is concerned, or is this just the nature of
the beast?

I often time use the “Left Click” on Single Tones for better decode using FSK441. I run with
QRN set to 5 here. Anytime I hear a ping or have a ping only partially decoded I use the
Right Mouse Click and many times decode is better. If you’re in a quiet location, often time
you might hear signals that do not meet the minimums for decoding to even occur so note the
time mentally and search around on the green line for any indication of a signal and try
clicking on it.
Most the time pings will decode properly but in some occasions being proactive really helps.
Listen and observe the green line, click good prospects. (especially those you hear)
Once the DF is know and the “RIT” is set to reduce the DF to near zero and the TOL reduced to
100 or less the decoding is generally much better.
The last thing to learn is patience. Running for 10 minutes and stopping just because you didn’t
hear anything is a common newbie mistake. A good rule of thumb for skeds under about 1200 miles
is plan to run at least 30 minutes and be happy with anything less. If you’re trying to work
short backscatter paths or longer distances it might take an hour or more.