[Elecraft] K1 Internally Generated Signal (Long Response)

[Mike Morrow]

Jim / KI9M wrote:

> Page 58 refers to "internally-generated band edge signals".
> 
> I can't seem to find it. Does the K1 really have an internally
> generated signal?

There is no such signal there that is purposely designed for such
use.  But there is a signal at the *lower* edge of a band beginning
at an integral MHz value that is marginally useful.

Here's where it comes from:

The front panel board has a 4000 kHz crystal X1 to clock the operation
of the MCU U1.  The MCU generates an internal clock that is one-quarter
of X1...or 1000 kHz.  That would generate a bothersome spurious signal
at the bottom of most HF ham bands, so the actual operating frequency
of X1 is padded down below 4.000 MHz with C1 and C2 to keep these spurs
below the ham bands.  

On my K1, and others typically, X1 operates very near 3999.5 kHz.  
The spurs thus occur every ( 3999.5 / 4 ) kHz or every 999.875 kHz
throughout the tuning range of the K1.  For my K1:

The spur just below  7000 kHz occurs at  7 x 999.875 =  6999.125 kHz.
The spur just below 10000 kHz occurs at 10 x 999.875 =  9998.750 kHz.  
The spur just below 14000 kHz occurs at 14 x 999.875 = 13998.250 kHz.
The spur just below 21000 kHz occurs at 21 x 999.875 = 20997.375 kHz.

It can be seen that the difference in spur frequency from exact band
lower edge frequency is not a fixed value, but one that gets larger
as frequency is increased.

To use these spurs to calibrate the K1 frequency display it must be
remembered that, properly set up, the K1 LCD should show *transmitter*
frequency.  The K1 receiver functions in LSB mode on all bands with
existing filter board designs.  This means that if the receiver is
zero-beat with 10000 kHz WWV, the transmitter should actually be set
to transmit a signal that is *below* 10000 kHz by the amount of CW
offset for which the K1 is adjusted.  My K1 uses an CW offset of
600 Hz.  That means that when my K1 is zero beat with 10000 MHz WWV,
the frequency displayed should be 600 Hz below 10000 kHz, or 9999.6
kHz.

Thus, the frequency displayed on my K1 when:

the  6999.125 kHz spur is zero beat should be  6998.525 kHz, 
the  9998.750 kHz spur is zero beat should be  9998.150 kHz,
the 13998.250 kHz spur is zero beat should be 13997.650 kHz,
the 20997.375 kHz spur is zero beat should be 20996.775 kHz.

If you are using a offset value that is not 600 Hz, substitute
your offset in place of the 600 Hz used in my example above.

For better accuracy in calculating the spur frequency, the
oscillation frequency of X1 could be determined to greater
accuracy and precision than I have used above, with an accurate
receiver.

If the K1 is using the 80 kHz VFO span option, on 30m it will
not be possible to tune to hear the spur just below 10000 kHz.

If the K1 filter board is tuned to 17m, it will not be possible
to tune to hear the spur just below 18000 kHz.

If the K1 filter board is tuned to 80m, it will not be possible
to tune to hear the spur just below 4000 kHz.

These spurs are fairly weak, and the K1 has a pretty good four-pole
crystal filter that makes it impossible to hear the other side
of the spur to make an accurate zero beat.  And obviously one isn't
going to get closer than a hundred Hertz for any frequency that is
displayed on the K1...never the nice exact numbers used in the
examples above.  A much better approach is to use a much stronger
and accurate external calibration signal.  I use a ham transceiver
that has a TXCO and can generate strong cal signals where ever I
want with very good accuracy.

Also, due to the weakness of these signals, they are completely 
useless for performing an initial alignment of a K1 filter board.

> I'm very happy with my new rig. My CW is a little rusty, but will
> improve with practice.

The K1 is an fantastic rig.  Number 175 is still my favorite after
13 years.

73,
Mike / KK5F