On Sat, May 1, 2010 at 2:05 PM, Brian Lloyd <[email protected]> wrote:

 Uh, no. What you have described is the convention used by amateur
> radio operators, not what really happens physically.


Look, you can describe "what happens physically" as an offset from green
light if you want to. The point is, what's the USEFUL thing to do? What
makes SENSE and makes operation EASIER and SIMPLER without compromising
anything, and this is especially so in the case of emergency ops.


> We have 50 stations operation PSK on 20m. All have their radios tuned
> to 14,070 kHz. None of them are interfering with each other. By your
> reckoning, they are all on the same frequency because they have tuned
> the big knob on their transceiver until the number on the display
> reads 14,070 kHz. OTOH, if each of them uses a frequency counter to
> measure the carrier frequency they are radiating, they are each going
> to get a different number, something between 14,070 and 14,073 kHz.
>

So what you're doing is sliding the modem over the passband with the radio
at one frequency, and what you've lost the ability to do is tell the other
station where to tune their radio and reliably receive your signal; instead
they've got THREE tuning steps to start: (1) tune the radio, (2) tune the
modem, (3) tune the IF-shift or put up with AGC pumping. To re-tune, (1)
tune the modem, (2) tune the IF-shift.

OTOH, do it my way, there's ONE re-tuning step to start. Tune the radio. IF
shift remains constant - when the modem freq is constant, so is the
IF-shift. Passband width is the same either way.  There's also ONE tuning
step to re-tune: Tune the radio. There's minimal AGC pumping because the IF
bandwidth is properly set the moment the signal is in the right place on the
waterfall.

Don't see the benefit of your way at all. Doesn't give you more space to
tune around in; makes external filtering hardware you might want to use have
to re-tune; adds tuning steps (or else commits you to use an open IF
bandwidth so that everything in the passband gets through, which isn't the
best idea I've heard recently, unless you like getting your AGC pumped by
that strong signal down-passband... entertainment value? challenge? :)

That is the problem of the receiver, not what goes out over the air.
> To me all the following are the same:
>

The thing is, to an AF filter, to an IF shift, to the AGC circuitry, they
aren't. And it isn't your sensibility about the "appropriateness" of an RF
frequency that controls an these things. It's the actual AF frequency. And -
just offhand - the rate of change of a lower frequency AF signal isn't as
fast as a higher one. That's why most AFSK signals are designed to reside at
the higher frequency side of the passband - e.g., SSTV isn't 300 Hz sync and
500 Hz to 1300 Hz, it's 1200 Hz and 1500 Hz to 2300 Hz. If you move stuff to
the  lower end, you're not doing the frequency response of the channel any
favors that I'm aware of.


> Even fldigi pays homage to this concept.


That doesn't mean it's the right thing to do. It just means it was something
that could be done, and was. No one has yet to put a *benefit* of this
approach on the table. I've pointed out that it adds unnecessary tuning,
compromises optimum IF settings requiring resetting (or expose you to AGC
pumping), can lower the signaling rate, and compromises external filters
requiring resetting.

Whereas all of that is avoided if one treats the signal as an AFSK constant
with known targets.


>  >> Fldigi does calculate and display the
>
> Au contraire! What fldigi reports is that actual carrier frequency,
> the frequency one would measure with a frequency counter connected to
> the transmitter output!
>

And, I ask again, what is the benefit? What step does this save? What
advantage is gained? Why must the operator "know"? By adhering to this
practice, you compromise external filtering gear; you increase the number of
tuning actions without gaining any performance benefits; you have to
re-adjust IF shift to compensate whereas if you were using a locked/defined
set of AFSK tones... one tuning process; the IF shift and bandwidth remain
constant; and external hardware knows what to do; the operator doesn't need
to "know" anything other than (a) is the signal where it belongs in the
waterfall and (b) what they want to send.


> You actually believe what you wrote there, right?


Of course. Please don't act like that. This is a technical discussion of the
merits. If I am wrong, point out how I am wrong, don't attempt to impugn my
mental processes.


> Please tell me the proper modem design point for PSK31, Olivia, DominoEX,
> JT65, etc.
>

Generally speaking, the sensible place for the (high[est] tone) is 2295 Hz,
same as RTTY 170, 450, 800. Space and any intermediate frequencies are down
from there. If one needs to re-tune, it should be done with the radio, so
that IF shift doesn't need re-adjustment, and so that it is one operation.
Such a setting lets you use external AF filtering and other hardware, if
available, and also sets a convention so that the radio may be tuned, just
one step, instead of radio, modem, if shift, external hardware - up to four
or more steps.


> One of the things I used to do was set up VSATs. The frequency we were
> assigned was the center of the transmitted signal, not the frequency
> of the LO injection. As far as I can discern, the only place where
> "dial frequency" is "transmit frequency" is amateur radio. Everyone
> else uses the center of the transmission.
>

That's because we almost always use AFSK. Not FSK. We don't have the same
kind of gear; do you know of a transceiver that lets you tune to 14..
Consequently, the most efficient way to operate is tied to what we actually
have -- not how commercial / military stations operate. Not my fault. Talk
to the FCC. When you do, you might tell 'em I have zero respect for the
shoddy job they've done WRT their mandate: To preserve the RF spectrum for
the most benefit of the citizens of this country. But that's another thread.
:)


> And I agree with Rick, K9AO, 100% on one thing: this is an education
> issue. Operators need to know precisely where their signal is going.
>

Ideally, it's going to the proper place relative to the radio's tuned
frequency. Once you try to add sliding modems, you've added problems with
filters and IF shift (or AGC pumping) you didn't have before. You're just
making it more difficult for them. There's no benefit to be had. People keep
saying "they need to know", and I repeat: WHY do they need to know? If they
already know:

1) They're tuned to the correct frequency on the radio (because the signal
is landing on the right spot on the waterfall), and

2) the received AF frequencies are always the same, so they can adjust
hardware filters ONE TIME and then, for bandwidth and shift and external AF
goodies, leave them be, leaving only concerns about notching and tuning the
radio when signals change,

3) There is only one RF display, and one AF display

..then what's the benefit of trying to say to them "well, you're currently
sending signals from 14.231200 to 14.232300 Mhz"? They're not going to be
tuning any hardware filters to those frequencies. The numbers they NEED are
1200 Hz, and 1500...2300 Hz. And 14.230 MHz. These numbers give you (1) the
correct radio receive tuning, (2) constant setting for filters, (3) the
correct radio transmit tuning, (4) constant AF conditions so that IF shift
remains constant if a signal comes up on another frequency.


> Here is another way to look at it. When I see a 500Hz blank spot in my
> spectrum display, I want to be able to tell my buddy to transmit in
> the middle of that blank spot, not have to calculate a bunch of
> offsets in my head.
>

Say you're on 14.230 Mhz. You see the blank spot. The Waterfall reads out
from 0 Hz to 3 Khz (for example.) You're on USB. The blank spot begins at
2000 Hz. So, you tell you buddy to "tune up 2 KHz." No calculations
involved. If you *want* to do so, you can say 14.232 KHz, but we're all used
to moving a few hundred or a few KHz... aren't we? I know I am. "Meet you up
(or down) x KHz" is an unbelievably common AR form of message, I have
trouble believing anyone would balk at saying such a thing. In this case,
it's "meet you up 2 Khz." No big deal.

--Ben
AA7AS
_______________________________________________
fldigi-alpha mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/fldigi-alpha

Reply via email to