[Elecraft] K3 CW and RTTY tones

[Guy Olinger K2AV]

> Now this is a long-winded explanation of more-or-less how the K3 handles tone offsets.

Not exactly.  Have to lean pretty far on the less...

What was wrong with the prior is that "audio" is not used in the
ordinary sense, the last or "lowest" analog stage in a K3 is a 15 kHz
IF for both transmit and receive.  The only audio is on the headphone,
microphone, speaker, line in/out leads to/from amplifiers, which are
all independent digtal to/from audio points for individual lines.

In particular, there is no balanced modulator in the K3.  The 15 kHz
IF SINGLE sideband signal, without a carrier, is directly generated
from a number stream in a digital to analog converter.  The number
stream was generated in firmware related to any audio input in use at
the time.  The audio input was converted to digital from mic/line in
inputs, and ALL processing done in firmware.

Data input from RS232, or DC states on input leads are also used to
generate number streams.  In particular, the DC wave shape of keyed CW
does not generate the transmitted wave. It interprets the DC CW wave
state to determine the events Key Down and Key Up.  A firmware routine
uses these to create a digital number stream representation of a
pristine waveshape from formula and memory.  This number stream comes
out clean as a whistle into the 15 kHz transmit IF and depends on the
linearity of the analog TX stream from that point forward as does
every other transmitted K3 signal.

In the sense that this scheme, from number stream at TX DAC to 15 kHz
TX IF forward is used for ALL modulation types, your notion is
correct.  To the extent that you couch it in non-existent (in the K3)
analog circuit methods, the K3 ain't your daddy's analog radio and
it's time for all around here to get all brains on DIGITAL standard
time, and out of the analog ages.

There is no balanced modulator in a K3.  What is actually done in the
digital number stream by the firmware is quite more simple than an
analog phasing or filter SSB generator.  Once firmware writers are
freed from the tyranny of emulating analog circuits in numeric
methods, modern digital methods are surprisingly more simple in the
numbers for a given end than in analog circuitry. Many digital methods
have no known analog equivalents.  We only keep using analog in our
head because that's all we have as a reference, and no other words to
use.   Therefore, to the question posted earlier:

> When operating CW or Rtty there is a split tone that you
> have to know
> about.  I have been reading on the HRD forums to
> figure this out to get
> the software set up correctly.  I know it has
> something to do with
> carrier supression, but it just does not make sense to me.

The answer is that the K3 does not use a balanced modulator to create
it's signals, therefore there is no carrier suppression to worry
about. The prior poster is correct if you ignore his analog
terminology and concentrate on what he is saying about frequency
offsets.   The K3 can be thought of as having a virtual carrier
reference frequency that shows up real only when you use AM.  While
most analog circuits can't do this, just consider that the virtual
carrier frequency is what would happen if your audio signal had a DC
voltage added to it, and DC was always converted by the digital radio
to the carrier frequency that a misaligned balanced modulator would
produce in an analog radio.

73, Guy.

On Mon, Jan 3, 2011 at 3:06 PM, Wes Stewart <n7ws at yahoo.com> wrote:
> Don,
>
> Stripped of a lot of stuff, the essence of it is this:
>
> When we transmit using a SSB transmitter (and except for true FSK, all modern radios use SSB for all modes, including CW and digital) all that is really happening is the up-conversion of an audio frequency to a radio frequency.  When we receive, we do the opposite, convert a radio frequency to an audio frequency.
>
> Although there are (usually) intermediate steps, the process can actually be done directly and with some switching, the same hardware can be used to do both.
>
> For example if there is a CW signal on 7100 KHz and we would like to hear it in our headphones with an audio tone of 1000 Hz, we can use a frequency mixer in which we combine the 7100 KHz signal with another one which we generate on 7101 KHz.  When we subtract the first from the second, we get the tone that we desire.
>
> Now if we put enough "smarts" in the radio, we can say to the radio, "When I listen to CW, the "pitch" of the tone that I prefer is 1000 Hz."  The radio then knows that when you (or it) tune the signal so the pitch is 1000 Hz, the actual frequency of the incoming signal is offset by that amount from the locally generated conversion frequency.
>
> In the case of a synthesized radio like the K3, it already knows the frequency of the conversion oscillator (the "LO") so it just does the math and in this case, subtracts 1000 Hz from that number and presents it on the display as 7.100.000
>
> To transmit, we do the reverse: subtract 1000 from 7101 KHz and get 7100 KHz. But here it's not so simple, the mixer also adds the two signals and we get 7102 KHz too.  Plus, the 7101 KHz LO also appears in the mixer output.  So we have a mess, 7100, 7101 and 7102 KHz.  The 7101 LO is also called the "carrier" since it is the vehicle that "carries" the audio signal to the radio frequency spectrum.  The other two signals, 7100 and 7102 are called the "sidebands" since they reside equally spaced around the carrier.  The FCC (and our soon to be enemy ham friends) would really frown on us transmitting this signal in the 40-meter CW band.  So we have to do some more work.
>
> When was hams say, "SSB", what we really are indicating is, "Suppressed-Carrier-Single-Sideband."  So with our messy signal we can use a special kind of mixer called a balanced-modulator that does the suppression of the carrier, that's the 7101 KHz signal in this case. Now we are left with the two sidebands, when we want only one.
>
> We can deal with this several ways, the most common are to either use phasing techniques to cancel one sideband while enhancing the other, or to use a narrow filter to remove the unwanted sideband.
>
> The phasing technique can be performed over a band of frequencies, however, the narrow filter we would need has to be done using crystals, hence we would be stuck on 7100 KHz.  So another frequency conversion is usually used with filter radios so that a single filter at an intermediate frequency can be utilized.  This frequency conversion operates just as before, with the same math used and the same, or more, problems associated with unwanted frequencies.
>
> Now this is a long-winded explanation of more-or-less how the K3 handles tone offsets.  I have tried and rejected HRD so I can't begin to explain how it accounts for this.  But maybe this fundamental explanation will help you sort it out or at least formulate more specific questions.
>
> Good luck,
>
> Wes  N7WS
>
> --- On Sun, 1/2/11, goldtr8 at charter.net <goldtr8 at charter.net> wrote:
>
>> Dear All,
>>
>> When operating CW or Rtty there is a split tone that you
>> have to know
>> about.  I have been reading on the HRD forums to
>> figure this out to get
>> the software set up correctly.  I know it has
>> something to do with
>> carrier supression, but it just does not make sense to me.
>>
>> So the more I read there the more confused I become.
>> I have also tried
>> google searches but I still have not figured it out yet.
>>
>> So if someone can direct me to some reading sources to help
>> me
>> understand this I would really appreciate it.
>>
>> Thanks
>> Don
>> KD8NNU
>
>
>
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>