[Elecraft] N6KR's 4 MHz oscillator cal method for the K2

Thu Aug 28 08:49:01 2003

Michael AG9V gave a detailed report on his investigation into 
frequency calibration. I am not going to paste all of that or give a 
detailed description of my own attempts. The point it shows is that 
the best accuracy you can achieve is within 10Hz on the lower bands 
and rather more on the higher ones.

I have not tried Wayne's method, but I am sure it works. My attempt 
the other day (after my last posting) was briefly as follows:

1. Feed my Gould synthesised generator from the 1MHz reference of my 
Racal counter - the Gould reference has known errors.
2. Set the generator to 9996kHz (RWM) and check the error against RWM 
off air using Argo. The generator was found to be 1Hz low, confirming 
my reference was good.
3. Set the generator to 20MHz and monitor the difference with the 5th 
harmonic of the 4MHz oscillator with Argo. C22 adjusted with a bit of 
delicate tweaking to be within a few Hz, the best I could achieve. 
Refit lid and confirm the 4MHz reference has not moved and is stable.
4. Run CALL PLL on 40m.
5. Go back to 20MHz and check the reference was still in spec - it 
was.
6.Go through CAL FIL and force it to do a measurement on each filter 
setting - this must also be done on 40m though I am not sure why.
7. Note that all the above were done in the CWN 400Hz filter - with 
Argo where you are looking at differences it is irrelevant what 
filter you use.
8. Go through the bands, set the signal generator at suitable 
frequencies, set the K2 to the same frequency, and monitor the 
frequency with Argo. Measure the error from a nominal 700Hz (sidetone 
frequency) .The results below use CWNormal for the lower bands and 
CWReverse for 21MHz and above. I only use CW so it was logical to do  
this in CW mode.

Error: RWM (9996) +2Hz
30m 15Hz, 20m +20Hz, 17m +28Hz, 15m +60Hz, 12m +60Hz, 10m +11Hz

Note that you cannot in anycase set your filters to better than 20Hz 
resolution, so you expect a shift between CWN and CWR and the other 
modes.

Note I can see the (Chinese?) BWM on 10MHz (almost spot on) with Argo 
but no sign of WWV that I can positively identify. An advantage with 
Argo that it can see signals way below the audible level.

Conclusion - as good as I am going to get it.

73 Dave G3YMC

[email protected]
http://www.dsergeant.btinternet.co.uk