Setting
the K2 Dial Calibration
Don
Wilhelm W3FPR
Updated
12/04/2003 to include the "N6KR Method"
The
subject of obtaining better dial calibration has come up repeatedly
over the life of the K2, and usually results in lively discussion on the
Elecraft reflector. Many users have
their favorite methods while there are others who have tried following the
steps in the K2 manual and for some reason find the process cumbersome.
By
creating this document, I attempt to clarify what is happening when you
perform the steps to calibrate the K2, and present various methods to
accomplish the task.
This
is a hyper linked document, and I have included return links in most cases, but
most links will return you to the top of the page.
To return to the place in the document where you linked from, use the
BACK button on your browser. Many
of the referenced documents will open in a separate window, so you will have to
close them when you are finished with that particular document.
This
is the first edition of this document, so please pardon me if there are any
errors, but please do tell me about them so they can be corrected.
First,
a bit of mixed fact and theory (and a few first person opinions) about how the
K2 goes about the task of deciding what numbers to display as the operating
frequency.
The
K2 VFO is a PLL synthesizer design – This particular implementation was chosen
by the K2 designers for 1) low phase noise (and lower noise of other types
too), 2) the design is reasonable to duplicate (very important in a kit), and
3) with parts that are cost effective (making the K2 more affordable).
The
K2 depends on the accuracy of the reference oscillator during CAL PLL, CAL FIL,
and CAL FCTR functions. At other times
the oscillator is required for the K2 to function, but an accurate frequency is
not required at that point. This reference oscillator
is implemented with a crystal, and the long-term stability depends on
temperature, innate stability of the crystal material itself, and other
factors. To achieve the best
calibration, it is critical to set the oscillator as close as possible 4000.000 kHz, shortly before calibrating the
K2.
The
K2 frequency display is handled by the microprocessor from look-up tables, rather
than directly reading the frequency.
The values in the table for the VFO are created when CAL PLL is run, and
CAL FIL creates the values in the BFO table.
In SSB mode (and RTTY, if enabled) the VFO and BFO values are used to
compute the frequency you see on the display.
For CW and CWr, a frequency offset equal to the sidetone pitch you have
selected is also calculated into the display.
The K2 always shows the actual signal frequency – but for this to happen
correctly in CW, the audio tone of the signal you are receiving must match
your selected sidetone frequency.
In
the process of creating a design, some decisions were made. One of these decisions was to select the number of
bits to include in the Digital to Analog Converters (DAC) used in the K2. Since these decisions are already
implemented, there is not much that can be done to improve it (short of selling
your K2 and buying another transceiver). So knowing what limitations these
decisions impose is necessary if we are to avoid ‘beating our heads against the
wall’ while trying to accomplish the impossible.
·
The
dial frequency reading itself is limited to 10 Hz increments.
o
That
really means that we can’t resolve the K2 frequency any finer than that
(we cannot interpolate from a digital dial like we could if it were analog)
·
The
PLL step size is in increments (i.e., the VFO tunes in increments) – and the
increments are larger at higher frequencies.
o
About
3 Hz on the 160 meter band
o
Around
7 Hz on 20 meters
o
Approximately
10 Hz on 10 meters
·
The
BFO increments in steps too – and the step size ranges from 20 to 35 Hz.
·
The
frequency displayed on the K2 dial is the VFO frequency +/- the BFO frequency (and
in CW and CWr modes only) +/- the sidetone pitch.
Given
the size of these increments, after all the math is done, you could determine
that the best one can accomplish in dial accuracy for the K2 is to get it within 20
Hz. Even that takes a lot of work and
care with measurements. If your
expectations are within 30 Hz, the task is a bit easier, but not easy enough to
allow sloppy procedures.
Credits:
Most
of the information above comes from the K2 manual written by Wayne Burdick N6KR
and Eric Schwartz WA6HHQ with additional material derived from a post to the
Elecraft reflector by John
Grebenkemper KI6WX dated April 17, 2003.
Click the link to view John’s post in entirety.
THE
MAIN PROCEDURE -
That
is all there is to the process – yes really.
If
the steps above seem simplistic, I admit – they are. But, these are the steps needed. My presentation is not
meant to over-simplify, but to provide clarity. Each step outlined above can be considered separately. But if more than
one step is performed, they must be performed in order.
The
difficulty lies not with process outlined above, but with various methods to
accomplish step 1 and step 3. The task
becomes increasingly difficult as the desired degree of accuracy increases.
The
hyperlinks below will move you to pages with detailed information for each
step. Step 2 is quite complex, but
fortunately the K2 has the firmware to do this all by itself.
N6KR Method: by none other than Wayne Burdick himself. Wayne has come up with the easiest and (in my opinion) the best method yet. It does require that you be able to receive some other known frequency - like WWV and zero beat it with accuracy, but requires no equipment other than the K2 itself and has been shown to be quite accurate.
Using the N6KR Method with RWM at 9996.00 kHz has been detailed by Vinec EA3ADV and can be viewed by clicking here
Automatically done by the K2 when selected from the menu.
Click here to go to the ‘Run
CAL FIL’ procedure page.