[K3] Spectrogram and the K3

[alorona]

If you have nothing to do today, try this:

Install Spectrogram on your computer so that you can feed the audio output of
your K3 into your computer's line in (or your laptop's mic in).

Scan the input using the following parameters in Spectrogram (I am using version
5.1.7 here):

Sample Rate:  11 kHz
Resolution: 16 bit
Type:  Stereo

Display Type: Line
Channels:  Dual
Scale:  60 dB

Freq Scale:  Linear
FFT Size:  1024
Spectrum Average:  32

Click on OK to start the measurement. Put the K3 into CW mode and vary the WIDTH
control. Look at that fine, flat frequency response that follows your WIDTH just
perfectly. I was using this today to help my son visualize what that control
does. (You may have to set the K3 volume control to about 12 o'clock to get a
good reading in Spectrogram, but this depends on your Windows volume control,
whether you have HI or LO audio gain selected on the K3, etc.)

Set the K3 WIDTH to 50 Hz and notice the response. Beautiful. Now, turn APF on
and watch very closely. Notice what APF does to the peak! Now you know why it
can be so effective under the right conditions.

I asked you to set the Type to STEREO in Spectrogram, but an interesting thing
happens when you set it to MONO instead.

After setting it to MONO, set your K3 WIDTH to a convenient width such as 1.0
kHz. Notice the nice, flat response as before. Now, turn on the K3's AFX and see
what happens. Change the AFX from Delay1 to Delay5 and see the results. For
extra credit, explain what is happening here. Here's a hint: set AFX to "Bin"
and notice that the signal goes away completely! My guess is because
Spectrogram's MONO mode is essentially L + R, and in Bin mode L and R are
out-of-phase and so their sum is zero. Assuming this, can you explain why Delay1
through Delay5 look like they do?

Al  W6LX
______________________________________________________________
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

[alorona]

When you look at Delay1 through 5 in MONO mode in Spectrogram you see peaks and
nulls in the frequency response. Spectrogram is only able to show the magnitude
and not the phase of signals, so from what was learned from AFX = Bin we saw
that the result of adding L + R as vectors can result in cancellation.

Similarly in Delay1 through 5 we are seeing the L and R noise signals adding
together and cancelling each other out at regular intervals -- resulting
in nulls in the frequency response. The *spacing* between the nulls can tell us
something. I measured the spacing as follows:

Delay1 :  500 Hz
Delay5 :  100 Hz

The nulls are caused by two signals, L and R, at the same frequency but shifted
in time, and as they combine their vector sum causes the nulls. The time delay
between L and R is easy to calculate based on our measurements:

Delay1 = 1/500 = 2 msec

Delay5 = 1/100 = 10 msec

That's my best guess as to the actual delay values.


Al  W6LX
______________________________________________________________
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]

Great suggestions.

If you want to see the phase relationship, take a look with Audacity.

Audacity is an editor for audio files, but can be made to work in real
time too (see the help files).
It will show both channels like an oscilloscope.

There are other audio tools that can be either fun or useful in
analyzing many signals in the audio spectrum.  Spectrum Lab,
Spectrogram, Audacity, and Total Recorder are favorites of mine.
There are also Audio Signal Generators that are helpful at my workbench
(but also in the shack at times) - F-G Lite serves most of my purposes,
NCH Function Generator is better, but will only run for a while before
you have to purchase it.

73,
Don W3FPR

On 7/9/2012 6:47 PM, Al Lorona wrote:

> When you look at Delay1 through 5 in MONO mode in Spectrogram you see peaks and
> nulls in the frequency response. Spectrogram is only able to show the magnitude
> and not the phase of signals, so from what was learned from AFX = Bin we saw
> that the result of adding L + R as vectors can result in cancellation.
>
> Similarly in Delay1 through 5 we are seeing the L and R noise signals adding
> together and cancelling each other out at regular intervals -- resulting
> in nulls in the frequency response. The *spacing* between the nulls can tell us
> something. I measured the spacing as follows:
>
> Delay1 :  500 Hz
> Delay5 :  100 Hz
>
> The nulls are caused by two signals, L and R, at the same frequency but shifted
> in time, and as they combine their vector sum causes the nulls. The time delay
> between L and R is easy to calculate based on our measurements:
>
> Delay1 = 1/500 = 2 msec
>
> Delay5 = 1/100 = 10 msec
>
> That's my best guess as to the actual delay values.
>
>
> Al  W6LX
> ______________________________________________________________
> 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
>

______________________________________________________________
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