[Elecraft] Tuning L34 on K2 - The 'Signal to Noise Ratio' method

[Ron D'Eau Claire]

Tom, N0SS, wrote:

... I was dinking around with L34 (no recollection why), and even 
'earballing' the adjustment, I did notice what I perceived to be a
distinct 
drop in relative background at a point just inside the point of maximum 
received signal... maybe 1/2 turn (into the can) before reaching max 
signal. I was so surprised to hear it that I tried it several times,
just 
to confirm that I was hearing what I thought I heard, and each time I 
noticed the same thing. So I now do believe that there was some thought 
behind the statement.

NOTE that I have NOT (yet) performed any quantitative measurements to 
confirm my beliefs.
==========

I've heard that too, but decided that it was an artifact of my hearing
based on the frequency spectrum of the noise. Looking at the K2 audio
spectrum I can see the noise peak produced by L34 move across the audio
range as the coil is adjusted. This seems to a produce a subjective
response in our ears (and in an voltmeter on the audio amplifier output)
that makes the best signal-to-nose ratio appear when L34 is tuned off of
its real peak. 

To see the effect on the background noise using Spectrogram requires me
to turn the Attenuator On to put the 7 MHz birdie down near the internal
noise level so I can see both the peak of the birdie signal and the
noise floor on the display without driving the K2 into compression. It
is important to turn the K2 AGC Off so it isn't trying to maintain an
even audio output level as I make the adjustment. 

Then, as I adjust L34 I can see the "hump" of noise move back and forth
in frequency as the spike of the "birdie" signal moves up and down in
amplitude. To my ear, the background noise sounds lower when L34 is
tuned to the low-frequency side of the peak. That produces the
impression that the signal-to-noise ratio is better there, but it looks
the same on Spectrogram. That is, the difference in dB between the
signal peak of the birdie and the peak of the noise stays the same on
the Spectrogram display. Only the frequency spectrum of the noise is
slightly different. 

I found that trying to measure the "peak" with an audio voltmeter at the
speaker jack produced a false peak when L34 is tuned off on the
high-frequency side of the actual peak.  The voltage the meter shows is
the total voltage at all frequencies across the spectrum that is being
amplified by the audio amp in the K2, and it appears that by detuning
L34 to the "high frequency" side of the actual resonant point produces
greater wide band noise for the meter to register, producing the false
"peak". 

It seems to me that our ears, like that a-c voltmeter, are responding to
a wide spectrum of noise that masks the actual "sweet spot". 

And, bottom, line, I can't see how that has any detectable effect on the
K2's performance in the "real world" on the HF bands. These noise levels
seen when doing the adjustment are WAY below what is coming in from my
antenna in actual use. Still, it's a satisfying feeling to have the
adjustment "peaked". 

Ron AC7AC
K2 # 1289