K2 Alignment with Spectrogram

[KO5V]

My K2 S/N 5149 has been up and running well for about 5 years. I finally
decided to check out the filter alignment using Spectrogram. I
downloaded v.16, and am running it through Wine since I'm using Ubuntu
as an operating system. I couldn't get the version on Tom's page to run
this way, and right now, I don't have easy access to a Windows machine.

The parameters on the newer version are a bit different than what Tom
specifies:
1. The slowest sampling rate is 22K, probably not a problem;
2.  To get a 10.8Hz frequency resolution, I need to set the FFT size to
2048 (1024 gives a default of 21.5Hz), again probably not a problem, I
can probably even go to a higher rate and get fewer artifacts on the
signal;
3. Band limits set from 0 to 2840Hz.

I'm using the K2's headphone output to connect to the mic input on the
sound card. Its settings are minimal, but I have selected the correct
input, and am not adding any amplification. The program does appear to
work. I can see the different filter widths, and can move them about the
600Hz cursor to center them.

There is a lot of noise riding on the signal, maybe "5dB-10dB" worth.
There is a Peak Hold function that shows a reasonable shape to the pass
band if I engage it, but the live signal is very noisy. Also, instead of
being way down in power, like in Tom's example, to get the baseline of
the signal at the bottom, I had to set the spectrum level from -10dB to
60dB (This may be a function of my sound card (?)).

I attempted the L34 alignment as also mentioned in Tom's instructions,
and that's where the questions really started. The noise level riding on
the signal is the same, and there are spikes at 600Hz intervals. The
first (@ 600Hz) is about 20dB-25dB above the noisy baseline. The problem
here is that L34 is maxed out up. If I adjust it down, the spikes do
start to descend into the noise. I am worried that there is a problem
with L34, or another component in that circuit that isn't allowing
complete adjustment, and causing the noise riding on the signal. I did
do this by ear, per the manual, and the tone at 7Mhz did get louder, and
the noise did subside, but Tom's comments in the instructions lead me to
believe that there should be enough adjustment in L34 to clean up a lot
of the trash on the signal.

Is there a problem, or am I over-analyzing this, and is the ability to
center the pass bands all that matters? Any comments/criticisms are
welcome.

Thanks in advance. Jim KO5V
Sandia Park, NM



______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[k6dgw]

On 5/3/2011 9:39 AM, Jim Faulkner wrote:

> There is a lot of noise riding on the signal, maybe "5dB-10dB" worth.

I had to unplug my laptop and run on battery because of the noise coming
from the switching "brick" power supply.  The mic input on my sound card
is pretty low level and to get rid of all the noise, I had to mute it
and use the line in jack with a 10:1 divider in the cable and the K2
audio turned up.

73,

Fred K6DGW
- Northern California Contest Club
- CU in the 2011 Cal QSO Party 1-2 Oct 2011
- www.cqp.org
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Don Wilhelm-4]

  Jim,

You may be picking up noise on your soundcard input cable.
The only thing you really need to see are the skirts of the passband (so
you can discern the center for CW and the proper placement of the low
frequency skirt close to 300 Hz for the SSB filters) - most everything
else can be ignored for the purpose of aligning the filters.  In other
words, you are not analyzing all aspects of the audio signal, just
determining where the passband is.  You should have only noise coming
into the K2.  If there is a signal present, it will only confuse the
display and make things more difficult.

For the L34 adjustment, tune to the birdie at 7000 kHz - turn the preamp
on so it is more pronounced.
Use something close to a 700 Hz wide filter.  Then first adjust L34 so
the signal spike goes up as far as possible - then tune around that
point (about a turn either direction) while observing the high frequency
end of the audio spectral display to see if you can reduce the noise
level there without reducing the height of the signal spike.  If you do
not observe any difference, leave L34 set at the peak for the spike.

That is all there is to it.

73,
Don W3FPR

On 5/3/2011 12:39 PM, Jim Faulkner wrote:
> My K2 S/N 5149 has been up and running well for about 5 years. I finally
> decided to check out the filter alignment using Spectrogram. I
> downloaded v.16, and am running it through Wine since I'm using Ubuntu
> as an operating system. I couldn't get the version on Tom's page to run
> this way, and right now, I don't have easy access to a Windows machine.
>
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[KO5V]

Don,

Thanks a lot. It appears that I'm probably over-thinking this. I'm using
my n-gen for a source, and I can identify the skirts. I did as you
suggested with L-34, and then adjusted it again using my ears. With
Spectrogram, as I adjusted the height of the peak of the spike, the
noise on the signal did not ever change. Using my ears I could tell that
the noise was being reduced, so you're probably right about picking up
the noise between the rig and computer.

L-34 is still maxed out. I wonder if that's a problem, and if there
would be more to be gained if there were more adjustment, or should I
just declare victory, and move on?

73, Jim KO5V

On Tue, 2011-05-03 at 13:32 -0400, Don Wilhelm wrote:

> Jim,
>
> You may be picking up noise on your soundcard input cable.
> The only thing you really need to see are the skirts of the passband (so
> you can discern the center for CW and the proper placement of the low
> frequency skirt close to 300 Hz for the SSB filters) - most everything
> else can be ignored for the purpose of aligning the filters.  In other
> words, you are not analyzing all aspects of the audio signal, just
> determining where the passband is.  You should have only noise coming
> into the K2.  If there is a signal present, it will only confuse the
> display and make things more difficult.
>
> For the L34 adjustment, tune to the birdie at 7000 kHz - turn the preamp
> on so it is more pronounced.
> Use something close to a 700 Hz wide filter.  Then first adjust L34 so
> the signal spike goes up as far as possible - then tune around that
> point (about a turn either direction) while observing the high frequency
> end of the audio spectral display to see if you can reduce the noise
> level there without reducing the height of the signal spike.  If you do
> not observe any difference, leave L34 set at the peak for the spike.
>
> That is all there is to it.
>
> 73,
> Don W3FPR

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html