[Elecraft] Frequency calibration points

[John Grebenkemper, KI6WX]

Lenny;
The interaction between the two K2 oscillators makes it difficult to predict
the frequency accuracy at any one frequency.  From my measurements, you can
make the following statements:

- The best accuracy that the K2 can achieve across all bands is limited to
about 20-30 Hz.  This is due to the design of the phase lock loop which
greatly reduces spurious signals over most designs by giving up some
absolute frequency accuracy.

- The K2 will achieve its best frequency calibration when the 4 MHz
oscillator on the control board is adjusted to precisely 4 MHz.  There are
several ways to do this.  After you re-adjust it, you need to perform a CAL
PLL and reset all of the CAL FIL frequencies.

- The K2 will drift with ambient temperature changes.  A rate of around -1
ppm/C seems to be in the normal range for a K2.  This means that a 10F
temperature rise (about the amount that a K2 will increase its internal
temperature from a cold start) in a stable thermal environment will cause a
drift at 10 MHz of around 75 Hz.

Your numbers are within my expected range for a K2's accuracy that has been
reasonably calibrated.

-John
 KI6WX

----- Original Message -----
From: "lenny wintfeld" <w2bvh@comcast.net>

>     About a week ago, Ken Brown invited people to try and use CHU as a
> frequency reference in checking K2 calibration. I made 2 checks on my
> K2  (sn1520): One on CHU at 3330.00 and one on WWV on 10000.00 KHz.
>
>     I did the check in cw mode, matching (by ear) the CHU/WWV carrier
> note to the K2's internal sidetone, looking for the slowest beat note I
> could. I'd estimate the beat note was 10 Hz or less in both
> measurements.
>
>     The results are:
>         1. for CHU, my K2 reads 3330.05 (or 50 Hz higher than CHU's
> actual transmit frequency)
>         2. for WWV, my K2 reads 9999.91 (or 90 Hz lower than WWV's
> actual transmit frequency)
>
>     I'm sure not going to sweat these differences. They're plenty close
> enough to keep me in band (and in band-segment).
>
>     But I was wondering a) if this is typical of K2's in general;  b) is
> a different calibration error to be expected on each band and c) is my
> measurement method valid.
>