[Lowfer] Phasing exciters
Ed Phillips
[email protected]
Thu, 13 Nov 2003 08:21:38 -0800
Stewart Nelson wrote:
>
> Hi Ed, John, and all,
>
> By using both output channels of the sound card, and
> doing the audio phase shift in software, a phasing
> network is not needed. I believe that the exciter
> should only require two chips, a dual flip-flop and a
> dual 4-input analog mux, e.g. 4052. (That does not
> include the LO and output amp.)
>
> Unfortunately, I have not yet built/tested this
> design, but it is pretty straightforward, and I
> wouldn't expect any serious problems. I'm traveling
> and don't have a decent drawing program; here is
> a textual description:
>
> Start with a source at 4 times the desired LO freq.
> Feed it into the clock inputs of the flip-flops.
> Connect the flip-flops as a two-bit Gray counter.
> Drive the multiplexer address inputs from this
> counter.
>
> Let's call the soundcard outputs I and Q. Wire the
> multiplexer inputs to cycle through them as below:
>
> (TIME) ADR IN 1Y IN 2Y
> ------ --- ----- -----
> 0 0 I GND
> 1 1 Q GND
> 2 3 GND I
> 3 2 GND Q
>
> The outputs 1Z and 2Z, taken differentially, form the
> SSB signal. You just need a transformer or an op-amp
> to convert to single ended. You also need some low pass
> filtering, because there will be strong spurious signals
> near odd harmonics of the LO.
>
> There are several ways one could use such a design.
> For the 160-190 kHz band, you could have a fixed LO
> near 175 kHz, and (using audio frequencies up to 15 kHz)
> tune the entire band in software. A possible problem
> with higher audio frequencies is that sound card time
> base stability becomes more important.
>
> Alternatively, you can use an audio frequency of a
> few hundred hertz, and vary the LO for coarse tuning.
> A synthesizer with, say, 100 Hz steps should be ok,
> with fine tuning done by varying the audio.
>
> 73,
>
> Stewart KK7KA
Very interesting thoughts. How does an ignoramous do "the phase
shifting in software"? The rest is straightforward enough.
Ed