From: Rich Lentz KE0X
Dated: July 28, 2003
Subject: K2 Frequency Alignment

-------------------------------------------------------

I have had very good luck with a modified method A that Steven Gibbs,
GU3MBS, suggested (I run CAL PLL after step A.4 and/or after B3). Copy
follows.  Performed it (and started this discussion yesterday) And again
several more times and finally got the reading for WWV at 10 MHz to less
than 10 Hz, and CHU is less than 4 Hz. 

I like Michael, have found that the 4.000 MHz oscillator has to be a VERY
small amount low to get proper readings.

Things to remember.

1. YOU MUST re run CAL PLL after you change the clock frequency.  As I
understand it, the K2 saves voltages, based upon frequency, and does not
save frequency.  It does not measure frequency (like some of the outboard
digital indicators).  Thus if you change the clock and do not build a new
table (run cal pll) then when you select a frequency in the dial the MCU
will ask for the same old voltage over again. And you end up chasing your
tail trying to calibrate the K2.

2. I have found that anything that is connected to the K2 test points (TP1,
etc) will "pull" or affect the frequency of what it is measuring.

3. I have found that I can not touch the 4.000 MHz oscillator with a probe,
it will pull it.

4. W1AW is not a frequency standard !!!!!  WWV and CHU are, and if you need
more do a search on the web there are lots of them, "One in your
neighborhood."

5. A calibrated receiver with its audio displayed in MixW or DigiPan can
give you readings to less than 1 Hz. Plus, you can see signals way down in
the grass - some I don't even hear.  Compare with WWV at 5 and 10 and
confirm that your "calibrated" rig is indeed calibrated.  This allows use of
a small length of wire just stuck inside the K2 (insulated and not touching
anything) and connected to the ant connector.  Rather than use the CW mode,
I use USB and then set the dial off by 1 KHz.  Make sure you factor the 1KHz
offset into you calculations.  This is also the best way to "Calibrate" that
rig to WWV.  (Unless you are sure you are looking at the 400, 500, 600, etc.
tone.)

5. DO NOT assume that the frequency display in Spectrogram, MixW, or
DigiPan, etc. is accurate until you confirm it. On many slower PC's it will
not be accurate, unless you have entered in a correction factor.

6. Even with your clock accuratly set for 4.000000 MHZ, your side tone is
NOT what you set it for. (Its close but NOT EXACT.)

7. I really don't think that you need to re run CAL FILL unless you
drastically changed the frequency. You will not see the difference, but the
instructions in the manual say to do it.


Rich