struggling with gMFSK setup

struggling with gMFSK setup

[Haines Brown]

Despite some generous help, I'm not doing well setting levels for
transmit from my FT-817. I am using a SignaLink interface and am
following the suggestions that came with it. 

When I set gMFSK to Transmit/Tune, I hear a tone. Although neither my
Master volume nor PCM volume are muted, increasing the playback levels
should turn on the SignaLink PTT LED, but that never
happens. Reception is fine. 

I can't use their test for the soundcard because my card does not
support full-duplex. Also, gMFSK says it does not support full duplex
in/out.  

What I'd like to ask about here are my gMFSK preferences. 

Under Devices, Sound, 

      Sound card device: /dev/dsp
      Requested sample rate: 8000 (default) 
      Tx rate offset (ppm): 0 (default)
      Rx rate offset (ppm): 0 

I checked 8-bit sound in/out, and left stereo and full duplex
unchecked.  

Under Devices, PTT, for device I use /dev/ttyS0, and for PTT line
(serial), I have both RTS and DTR checked. I leave unchecked the Use
inverted PTT logic.

Part of my problem is that I have little or no idea how to make these
choices. 
-- 
 
       Haines Brown, KB1GRM
        
	 
        

Re: struggling with gMFSK setup

[Tomi Manninen]

On Sun, 2006-10-29 at 15:50 -0500, Haines Brown wrote:

> Despite some generous help, I'm not doing well setting levels for
> transmit from my FT-817. I am using a SignaLink interface and am
> following the suggestions that came with it. 

I don't know signalink but it seems it's based on some sort of VOX 
(voice operated relay/transmit) to operate PTT (not a good idea if you
ask me btw).

> When I set gMFSK to Transmit/Tune, I hear a tone. Although neither my
> Master volume nor PCM volume are muted, increasing the playback levels
> should turn on the SignaLink PTT LED, but that never
> happens. Reception is fine. 

If you hear a tone, then gMFSK should be ok. Now you have to figure
out why your signalink doesn't "hear the tone", ie. why doesn't it
trigger PTT for you.

> Under Devices, Sound, 
> 
>       Sound card device: /dev/dsp
>       Requested sample rate: 8000 (default) 
>       Tx rate offset (ppm): 0 (default)
>       Rx rate offset (ppm): 0 
> 
> I checked 8-bit sound in/out, and left stereo and full duplex
> unchecked.  

8-bit sound should not be needed for any modern hardware. If you
don't get any errors then don't check that. (It's a good idea to 
start gMFSK from a command line, ie. type "gmfsk" to an xterm or
equivalent. That way you see if it gives any warning messages etc.)

> Under Devices, PTT, for device I use /dev/ttyS0, and for PTT line
> (serial), I have both RTS and DTR checked. I leave unchecked the Use
> inverted PTT logic.

PTT settings in gMFSK should not be relevant to you as your interface
relies on VOX for PTT. Just set PTT device to "none".

-- 
Tomi Manninen / OH2BNS / KP20JF74

Re: struggling with gMFSK setup

[Haines Brown]

> > Despite some generous help, I'm not doing well setting levels for
> > transmit from my FT-817. I am using a SignaLink interface and am
> > following the suggestions that came with it. 
 
> If you hear a tone, then gMFSK should be ok. Now you have to figure
> out why your signalink doesn't "hear the tone", ie. why doesn't it
> trigger PTT for you.

Thank you for the advice on gMFSK preferences and for the assurance
that a tone at least shows the gMFSK is sending something to the
soundcard. 

I'll seek some advice from Tigertronics/SignaLink. 

-- 
 
       Haines Brown, KB1GRM
        
	 
        

Re: struggling with gMFSK setup

[Haines Brown]

Kind of a progress report on my initial query. 

I had to do some stumbling about to get myself set up my soundcard to
record so that I might perform a test on my soundcard. I was
eventually able to record and playback sound from my transceiver, and
so the soundcard is working. 

However, I found that if my transceiver (FT-817) audio was at all up,
it triggered the PTT LED on my SignaLink interface, at which point it
blocked sound getting to my sound cards' Mike conenction. As a result,
the volume level on my transceiver had to be turned most of the way
down, and as a result, only a very weak signal was recorded. However,
by cranking Playback volume levels up very high for my soundcard
driver and the Play volume high on my sound recording software, I was
able to hear the signal. It seems the soundcard is working, but not
absolutely sure the levels were adjusted right.

I set my FT-817 to Dig Mode and have my SignaLink interface connected
as directed. For my soundcard driver, I set the Playback mike level to
zero, master volume to 75% and PCM/Wave Volume initially to zero. I
then advance it to the point the SignaLink PTT LED goes on - it never
does. 

So I'm back to where I started, only with the assurance that my
soundcard _seems_ to be working. I'm about to check with SignaLink
tech support.

-- 
 
       Haines Brown, KB1GRM        
	 
        

Re: struggling with gMFSK setup

[Dave Platt]

brownh@hartford-hwp.com (Haines Brown) wrote:
> Kind of a progress report on my initial query. 
> 
> I had to do some stumbling about to get myself set up my soundcard to
> record so that I might perform a test on my soundcard. I was
> eventually able to record and playback sound from my transceiver, and
> so the soundcard is working. 
> 
> However, I found that if my transceiver (FT-817) audio was at all up,
> it triggered the PTT LED on my SignaLink interface, at which point it
> blocked sound getting to my sound cards' Mike conenction.    As a result,
> the volume level on my transceiver had to be turned most of the way
> down, and as a result, only a very weak signal was recorded.

It's common for a soundcard's mic input to have two different audio
routing pathways associated with it.

When the input is in the "play" mode, the signal coming from it is
mixed into the outgoing (speaker or headphone) audio output, at the
level controlled by the input's software volume control.  The
signal may also be digitized and captured.

When the input is in the "record" or "capture" mode, the signal
coming from it doesn't go to the outgoing audio output, but 
only to the digitize-and-capture logic.

The same is usually true for the line-level input(s).

On some cards, these two signal pathways have a common volume
control setting, and there's just a mode switch (record or
playback) which controls the particular pathway which is enabled
at any given time.  On other cards, the two pathways have
individual enable/disable switches as well as individual volume
controls.

What you want, for computer-to-rig hookup, is to put whatever input
you're using into the "record" mode.  This will prevent the rig's
audio signal (fed to the sound card) from being fed back into the
rig via the speaker pathway, and will ensure that you won't get
accidental or inappropriate triggering of the vox (whether that's
on the rig, or in your audio interface box).

When using ALSA, you can manipulate the input's mode via the
GUI in the "amixer" program.  What you'd want to do, I think, is
select the input that you're using (MIC, probably), hit the M key
to mute it (the display will show MM to indicate that it's muted),
and enable capture with the space bar (the display will show
CAPTUR rather than ------).

This should let you set your rig's audio output level high enough
to permit a high-quality signal capture and digitization, without
inadvertently feeding back the audio into the transmit side and
triggering the vox/PTT.

Re: struggling with gMFSK setup

[Harry Popov]

Hi,

> I set my FT-817 to Dig Mode and have my SignaLink interface connected

Did you chose the proper mode for DIG? I use USER-U. Original setting is
RTTY with great shift which is unusable for computer generated modes.

73, Harry LZ1BB

Re: struggling with gMFSK setup

[Haines Brown]

Sorry for the delay in responding, for I swapped out my audio card
(turns out to have been unnecessary), and had other distractions as
well. 

But your information was _invaluable_! It explained very nicely what
was going on with my soundcard, and how to set the linux alsamixer
controls. Now the PTT LED on my SignaLink goes on as it should. Not
tried it on the air yet, but about to do so.

One other (OT) question. One usually runs a mode like PSK31 at well
under half power because it is nearly 100% duty cycle. However, my
FT-817 manual says nothing about this. Am I to assume that it can
handle 5 watts at 100% duty cycle, or should I bring power down to 2.5
watts (50%)?

Another OT question. I've seen it said that more than 25 watts makes no
sense in PSK31 because it can reach anywhere at that power. Is this
really so? I suppose the quality of one's antenna has some bearing,
and so, if the statement above is true, it assumes an efficient
antenna. 

Haines Brown KB1GRM 

Re: struggling with gMFSK setup

[Dave Platt]

brownh@hartford-hwp.com (Haines Brown) wrote:
> Sorry for the delay in responding, for I swapped out my audio card
> (turns out to have been unnecessary), and had other distractions as
> well. 
> 
> But your information was _invaluable_! It explained very nicely what
> was going on with my soundcard, and how to set the linux alsamixer
> controls. Now the PTT LED on my SignaLink goes on as it should. Not
> tried it on the air yet, but about to do so.

Excellent!  I'm glad that it all sorted out in the end.

> One other (OT) question. One usually runs a mode like PSK31 at well
> under half power because it is nearly 100% duty cycle. However, my
> FT-817 manual says nothing about this. Am I to assume that it can
> handle 5 watts at 100% duty cycle, or should I bring power down to 2.5
> watts (50%)?

There are a couple of different issues which create the
"50% rule" for PSK31.

One of these is the issue of the transmitter's ability to handle
full-power transmission at 100% duty cycle without overheating.
You certainly don't want to bake your finals... it could 
eventually result in all of the Magic Blue Smoke escaping.
A transmitter's ability to handle (and get rid of) the heat
depends on its heatsinking, airflow, the type of power
transistors used, and probably a bunch of other factors.

5 watts ERP, from a typical final running in Class B, won't
produce all that much waste heat in the finals... somewhere
between 5 and 10 watts at a rough guess.  If you've got
good airflow to the 817's heatsink, and perhaps a small
external fan, the 817 would probably be able to get rid of
the heat well enough.

Something you could check, in the manual (I don't have a copy
here) is whether Yaesu states any cautions about full-
power operation in RTTY mode, or extended full-power key-down
testing.  If so, honor the cautions - they'd apply to PSK31
as well.  If not, then the chances are (I think) that the
817's heat dissipation capability is adequate for this.

The other issue is linearity.  PSK31 uses an amplitude-and-
phase modulation scheme, and (like other amplitude-sensitive
modulations like AM and SSB phone, but unlike FM and CW) its
intelligibility can and will be affected by nonlinearity.
If the strength of the RF carrier envelope is not quite close 
to being a linear function of the PSK31 audio signal at any
given moment, then you'll get distortion of the RF envelope.
This will tend to look like a "flat-topping" of the RF carrier
if viewed on a station monitor or oscilloscope, and it appears
in a PSK31 waterfall display as a broad pattern of bright
lines on either side of the primary PSK31 carriers.  Many
PSK31 receiving programs can calculate and display the percentage
of IMD (intermodulation distortion).

A lot of ham transmitters exhibit good linearity up to
roughly half-power, and then begin to become less linear at
peak levels higher than this.  The nonlinearity isn't
terribly noticeable when speaking via SSB phone (which already
has a bandwidth of several kHz) but it's enough to generate
a significant amount of splatter around a (narrow-band)
PSK31 transmission.  This could make one unpopular with other
users of that portion of the band... it's considered rude to
splatter across a full kHz or more of the band when using a
modulation which is nominally capable of conducting a good
QSO on a fifth of that.

My guess is that if you limit your PSK31 transmissions to half
power, and have a good battery or power supply setup and
adequate airflow to the heatsink, your rig will be perfectly
happy and your PSK31 signal will be clean.  If you want to
go above that level, check for overheating, and have another
ham (or several) give you a signal quality report on the air
to indicate whether you have a problem with nonlinearity-
splatter or not.

> Another OT question. I've seen it said that more than 25 watts makes no
> sense in PSK31 because it can reach anywhere at that power. Is this
> really so? I suppose the quality of one's antenna has some bearing,
> and so, if the statement above is true, it assumes an efficient
> antenna. 

It can be so.  It really depends on the band conditions.

When whatever band you're using is propagating well, a surprisingly
small amount of power can complete a PSK31 QSO.  I've worked from
California to Antarctica (Polish research station) on 10
meters, using 35 watts on a simple dipole in my back yard.
My brother Chris managed to complete a PSK31 QSO from Oregon
to Argentina on 20 meters, using a little QRP PSK31 radio kit 
operating at about 2 watts.

When the band's not propagating well, you can't work from here
to anywhere outside of ground-wave distance with a kilowatt 
and a six-element beam :-)

I think that PSK31 can be thought of, in terms of required power,
as being more like CW than SSB phone - you don't need all that
much power to communicate - with the added proviso that whatever
power you _do_ have needs to be as clean as is possible.
An overdriven PSK31 transmitter is no more welcome on the bands
than a CW station splattering key-clicks around.

Oh... another thing to mention... when operating PSK31, make sure
that you turn _off_ your rig's audio signal processor (compressor)!
If left on, it will deliberately introduce signal-dependent
nonlinearity and will create an impressive amount of bandsplatter.
For similar reasons, it's probably best to leave your rig's
noise reduction system turned off when receiving.

Re: struggling with gMFSK setup

[Haines Brown]

Again, I must thank you for your patient and thorough attention to my
question. You really have illuminated things nicely.

> > Now the PTT LED on my SignaLink goes on as it should. Not
> > tried it on the air yet, but about to do so.
> 
> Excellent!  I'm glad that it all sorted out in the end.

Today, because I fiddled with soundcard mixer settings, my soundcard
does not function at all, even to play a simple sound file, so I'm
back to scratch, but at least I know what to do once I get my card to
word again.

> Something you could check, in the manual (I don't have a copy
> here) is whether Yaesu states any cautions about full-
> power operation in RTTY mode, or extended full-power key-down
> testing.  If so, honor the cautions - they'd apply to PSK31
> as well.  If not, then the chances are (I think) that the
> 817's heat dissipation capability is adequate for this.

The FT-817 manual says nothing about limiting power for digital
modes. So apparently it's safe to run a full 5-watts, but based on
what you say about IMD, 2.5 would be a good idea.

> When the band's not propagating well, you can't work from here
> to anywhere outside of ground-wave distance with a kilowatt 
> and a six-element beam :-)

Which brings up another naive question. On 20-meters, communications
principally rely on F2 layer skip, and that means I can hear stations
from about 500 miles to several throusand miles away. That's great,
but it would also be nice to communicate with folks in my own state
(within 100 miles) by PSK31. Is there any ground wave communication?

Haines Brown KB1GRM 

Re: struggling with gMFSK setup

[Harry Popov]

Haines,

> The FT-817 manual says nothing about limiting power for digital
> modes. So apparently it's safe to run a full 5-watts, but based on
> what you say about IMD, 2.5 would be a good idea.

All said befor is corect in principe. Particularly for FT817 do not wory
so much for IMD and thermal problems. Just have in mind the radio is
prepared for FM chatters on VHF. If it is engeniered to survive 5W FM
carier it should do on PSK too.

I use FT817 for several years on digi modes HF and WSJT VHF. I worked
DXCC on all bands except 80m with 5W and 20m LW just about 5 meters
above ground. Here is how I adjust PSK31 level. I turn down audio from
sound card. Go on TX and encrease gradually until I have full 5W. There
is a point where more encrease of audio do not produce more power. That
is so called "point of compression". You should always be just on it or
strace down. IMD is good due to good ALC of transmit unit of FT817.
Later if I need QRP down I just use FT817's own power levels without
touching any more the audio level from sound card. At least I tune in
this way. I never had splatter complains. And I have several neighbours
just about 200 to 800m away from me.

According mith that PSK31 does not need power. The explanation is only
one. Majority of stations use between 10 and 30W and our 5W QRP is
compeating very well with other stations. It should be the same on CW
and almost the same on SSB if all used the minimum necessary power. You
know reality is much different. Every one and his brother use QRO and
that is only problem for QRP stations to compeat during pile ups. But
not impossible!!! While using QRP one gain knoledge and one day take the
pleasure when overshout all the QRO gung. That is real pleasure. Beleave
me I had it many many times.

Excuse my spare English. I think you got the point.

73, Harry LZ1BB

Re: struggling with gMFSK setup

[Harry Popov]

BTW...

> The FT-817 manual says nothing about limiting power for digital
> modes. So apparently it's safe to run a full 5-watts, but based on
> what you say about IMD, 2.5 would be a good idea.

One good hint for FT817 users. Radio is getting hot on 5W. But most of
heat is coming not from power transistors but 7V inside voltage
regulator. Solution is to power it from low voltage. I use around 10V.
With 10V you still have 5 Watts power but much less heat. If your supply
is with fixed 13.8V output just connect 4 silicon diodes in series.
Connect one end to the power suplay and second to radio. You have
roughly 0.7 V dropping in a single diode. So 4x0.7=2.8V.
13.8-2.8=11V. You will be surprised how cool is FT817. Diodes should be
capable to pass 2A. I use 5A/30V diodes.

73, Harry LZ1BB

Re: struggling with gMFSK setup

[Dave Platt]

> Today, because I fiddled with soundcard mixer settings, my soundcard
> does not function at all, even to play a simple sound file, so I'm
> back to scratch, but at least I know what to do once I get my card to
> word again.

I've had a similar experience enough times, that I've taken to
storing my sound card configuration setup in a script file (e.g.
running "amixer" with a bunch of "set" commands) so that I can always
get the card back to a known good state.

> Which brings up another naive question. On 20-meters, communications
> principally rely on F2 layer skip, and that means I can hear stations
> from about 500 miles to several throusand miles away. That's great,
> but it would also be nice to communicate with folks in my own state
> (within 100 miles) by PSK31. Is there any ground wave communication?

In my experiences, groundwave on HF is good for very short distances
only.  It'll sometimes work out to a few miles, or to a few tens of
miles, but poops out after that.  It doesn't have much range advantage
(if any) over VHF.

Groundwave propagation works much better for vertically polarized
signals than for horizontally polarized.  That's one reason why
HF verticals have a reputation for being noisy - they pick up the
vertically-polarized groundwave QRM from manmade sources (power poles,
neon lights, etc.).  Horizontal antennas tend to reject this V-P
noise, and since horizontally-polarized noise (and signal) doesn't
propagate well by groundwave the horizontal antennas tend to notice
local QRM sources less.

There are a number of propagation tricks you can try:

-  Local groundwave from a 20-meter vertical.  Probably works best if
   both stations use verticals.
   
-  NVIS (near-vertical-incidence skywave).  A horizontal dipole at a
   low height (say, 1/8 wavelength) radiates most of its energy
   upwards, and enough may refract back for local and area
   communication.  The military has become fond of NVIS HF for
   tactical communication, I understand.
   
Sometimes NVIS works, sometimes it doesn't.  20 meters is probably a
bit harder to work via NVIS than 40 or 80, especially in the low part
of the sunspot cycle.

I've had some odd experiences doing local PSK31 on 20.  On one
occasion I was barely able to raise a station 15 miles away, even
though we were both working at about 50 watts.  His signal strength
seemed OK and there wasn't much noise, but my software couldn't get a
decent lock on it.  He had a similar problem copying me.  My best
guess is that we were receiving a combination of groundwave and NVIS,
and that the two pathways were interfering enough to mess up the phase
of the PSK31 carrier and prevent proper decoding.