[Elecraft] Re: K3 Audio

[Andrew Forrest]

Alexandr Kobranov made an interesting post showing a spurious response
in the line out audio

    "I just made two basic samples using K3 and TS-850 using Firebox
    external soundcard and Winrad SW.

    You can see screenshot from K3 line-out audio on

    http://www.ok1kei.org/K3_audio/line_out_firebox_winrad.jpg

    This is K3 with 400Hz roofing filter on some stable carrier.
    LINE OUT setting on level 9. "


Looking at the plot the spurious cannot be harmonics because the
frequency scale of the ghost response is the same as that of the normal
response. It is not doubled or tripled as would be the case with
harmonics. I posted similar results in March which I noticed using PSK
on a computer. Quoting myself:

    "The second thing I am
    noticing is ghosts in the waterfall display of MixW. I assumed that I
    was driving the audio card too hard and getting harmonics, but this is
    not the case. The ghost signals do move faster across the waterfall
    than
    the tuning rate implying harmonics but I can decode the ghosts, so the
    bandwidth of the ghost signals has not changed which it would do with a
    harmonic. Ghosts are about 20-30db down on the correct signals"

It appears that this effect is nothing to do with clipping or
transformer saturation but some kind of mixing. On the positive side I
have not manged to make a ghost appear in the passband where it would
damage the decoding efficiency, although it may still be there under the
noise. This is an unusual effect that I have not seen before. Anybody
any ideas of how these ghost signals are being generated?

Andrew Forrest
M1KAZ (K2 #3679,K3 #0499)
--