[Elecraft] AM with FM filter

[Guy Olinger K2AV]

Not all DSP processing, by any stretch of the imagination, is merely a
digital "analog" of an analog technique.  Some stuff simply CANNOT be
done in analog circuits of any kind.  It is true that back a ways, the
way we got started was simply digitizing analog techniques and saving
chip count by doing that, but these days there is so much digital
without any analog "analog," that there is no going back, and no
convenient nearby analog version to use as a mental image of what is
going on.  Like someone really learning a language, you know you are
there when you both speak and DREAM in digital.

One such K3 process is Wayne's proprietary TX voice amplitude
management scheme as a replacement for compression by RF clipping.

Some of this stuff got no analog daddy to point back at.

73, Guy.

On Mon, Nov 7, 2011 at 10:16 PM, Joe Subich, W4TV <lists at subich.com> wrote:
>
>> Since modulation involves imposing one signal on a second signal to
>> obtain the wanted third signal, technically there is no modulation in
>> a K3.  The desired signal is directly formed digitally via algorithm
>> and placed without further processing into the linear TX string at the
>> 15 kHz IF via a DAC.
>
> I don't know how Lyle does it but one "DSP Modulation" technique is
> simply to convert the incoming audio to digital create a quadrature
> version of the digital audio stream, multiply each of those streams
> with either an in phase or quadrature version of the "carrier" and
> combine (add) the resulting data streams.  In this case the "modulator"
> is identical to a phasing SSB modulator only in the digital domain.
>
>> The K3 uses firmware directed direct signal generation.
>
> It's still nothing more than doing mathematically what was previously
> done in circuitry.
>
>> It's comforting to call that voltage scalar pot plus an ADC the "RF"
>> gain. But it's real hard to find the RF gain lead going to the CPU,
>> cause the digital variable gain advice is time division multiplexed
>> with a bunch of other stuff sent on a single lead to the CPU. It's
>> "advice" because it has no direct affect on anything unless the
>> firmware and CPU want it to, in very severe contrast to our daddies'
>> analog radios.
>
> Again, it's simply doing in mathematics what our fathers' radios did
> in circuitry.  Simply because we can no longer point to a specific
> diode and capacitor that converts a series of RF waves of varying
> amplitudes into an audio frequency doesn't mean there isn't an envelope
> detector (or an AGC detector depending on the size of the capacitor).
>
> 73,
>
>   ... Joe, W4TV
>
>
> On 11/7/2011 9:37 PM, Guy Olinger K2AV wrote:
>>
>> On Mon, Nov 7, 2011 at 8:23 PM, Joe Subich, W4TV<lists at subich.com>  wrote:
>>
>>> Since modulation is performed in DSP, there is no change in the TX
>>> width.
>>
>> DSP modulation is an oxymoron of sorts.
>>
>> Since modulation involves imposing one signal on a second signal to
>> obtain the wanted third signal, technically there is no modulation in
>> a K3.  The desired signal is directly formed digitally via algorithm
>> and placed without further processing into the linear TX string at the
>> 15 kHz IF via a DAC.  And we don't normally refer to up-conversion to
>> the TX frequency as "modulation."  So there is no modulation.  The K3
>> uses firmware directed direct signal generation.
>>
>> We have SUCH a hard time walking away from our analog roots.  It's
>> comforting to call that voltage scalar pot plus an ADC the "RF" gain.
>> But it's real hard to find the RF gain lead going to the CPU, cause
>> the digital variable gain advice is time division multiplexed with a
>> bunch of other stuff sent on a single lead to the CPU.  It's "advice"
>> because it has no direct affect on anything unless the firmware and
>> CPU want it to, in very severe contrast to our daddies' analog radios.
>>
>> 73, Guy.
>>
>