[Elecraft] K3: Firmware 3.03 introducing artifacts?

[Richard Ferch]

James,

These fast tuning birdies depend on the mode selected, the roofing filter
selected, and the position of the shift control. If any of these was changed
between the last time you looked and this time, that could account for a
change between the two versions.

The birdie removal is not as simple as just finding the birdie and turning
the knob. If you are tuning at the fast VFO rate (50 Hz per step), you may
not be getting quite enough information to see and control what is
happening. If you change the rate to 10 Hz per step with the Rate button, it
is easier to see what is happening and easier to get the birdie removal
working right.

Here is an example. With the filter bandwidth set to 2.8 kHz, as you step
through frequencies at 10 Hz steps you might hear birdies with the following
pitches:

 28135.30 - nothing heard.
 28135.27 - 2700 Hz.
 28135.28 - 2400 Hz.
 28135.29 - 2100 Hz.
 28135.30 - 1800 Hz.
 28135.31 - 1500 Hz.
 28135.32 - 1200 Hz.
 28135.33 - 900 Hz.
 28135.34 - 600 Hz.
 28135.35 - 300 Hz.
 28135.36 - nothing heard.

At the 50 Hz tuning rate, you would only hear this birdie at two spots:

 28135.30 - 1800 Hz.
 28135.35 - 300 Hz.

I assume that's what you meant by a "two-tone birdie".

OK, with the dial tuned to 28135.30, use the SIG RMV feature. Depending on
whether you turn the VFO A knob to the left or to the right, the birdies
might be moved either up or down in pitch. Let's suppose you turned the knob
in the direction that moves the birdie up in pitch, and that the shift is
1500 Hz (BTW, I have no idea whether that's the right amount, this is just a
made-up example). Raising the pitch by 1500 Hz would remove the birdie at
28135.30 all right (it would now be at 3300 Hz, outside your bandpass), but
it would change the birdie at 28135.35 to 1800 Hz, which does not remove it.
This is the wrong direction to apply the correction for this particular
birdie. If you tune at 10 Hz steps, this correction will produce new birdies
at 28135.36-28135.39 that weren't even audible before.

If you turn the knob in the opposite direction, though, the pitch of the
birdie at 28135.30 will be lowered to 300 Hz, and there will be no birdie
audible at 28135.31 or higher frequencies. This is almost, but not quite,
good enough; so, turn the knob one more step in the same direction and that
will eliminate the birdie from the entire range 28135.30 - 28135.39 .

Now tune down one 10 Hz step to 28135.29. The birdie is still there at 2100
Hz - the previous adjustment did not affect it. Turning the knob to lower
the pitch, like you did in the previous step, will not work - it will make
this birdie worse. You need to turn the knob in the opposite direction, to
raise the pitch. A one-step rotation will eliminate this birdie from the
entire range 28135.20 - 28135.29 .

So for this particular birdie, you had to apply two steps in one direction
in one 100 Hz segment, and one step in the opposite direction in the
adjacent 100 Hz step.

Now suppose the birdie goes in the opposite direction, e.g.:

 28135.27 - 300 Hz.
 28135.28 - 600 Hz.
 28135.29 - 900 Hz.
 28135.30 - 1200 Hz.
 28135.31 - 1500 Hz.
 28135.32 - 1800 Hz.
 28135.33 - 2100 Hz.
 28135.34 - 2400 Hz.
 28135.35 - 2700 Hz.

In this case, you will need to use the opposite directions for correction
compared to the previous case. You will have to raise the pitch of the
birdies in 28135.30-39 by 2 steps, and you will have to lower the pitch of
the birdies in 28135.20-29 by one step.

I hope this helps you understand the birdie removal process better.

73,
Rich VE3KI