[Elecraft] 10Mhz alignment

[Ron D'Eau Claire]

John, KC5KQU asked:

According to my Leader LDC-824S counter, when transmitting with an 
indicated dial freq of 10.000.00Mhz, (Tune mode at .1w) my K2 is
actually TX on 9.999966

Tuning in the reference freq of the counter on the K2, I _almost_ 
zero-beat at 10.000.04Mhz.  When I TX with an indicated dial freq 
of 10.000.04Mhz, the counter reads it as 10.000001.

Am I just being anal/paranoid or could I get this closer?

===================================

Look at it this way, John:

1 - The oscillators in your K2 are set on frequency by the firmware.
When you select a frequency, the firmware looks up certain voltage
levels that are stored as digital data in memory and uses a
digital-to-analog converter (DAC) to provide that voltage to the
oscillator to put it "on frequency". That voltage level was determined
when you did the CAL PLL. The original voltage level stored in memory
was based on two things: 1 - The accuracy of your 4 MHz crystal
oscillator at the time CAL PLL was done and 2 - The resolution of the
digital to analog converters (DACs). The DAC's have a limit to their
accuracy. In other words, the required voltage to put the oscillator
back on the right frequency was "rounded off".  As I recall this process
limits the frequency accuracy to about 30 Hz. That is, your dial may be
up to 30 Hz off from what the actual frequency was when your ran CAL
PLL, assuming the 4 MHz reference oscillator was exactly on frequency
when CAL PLL was run. Any error in the 4 MHz reference oscillator is
added to this 'rounding' error in the DAC's.

2 - All of the above assumes that there is NO drift in your K2
oscillators. The DO drift! They are simple voltage-controlled
oscillators. They are well designed ones to be sure - but they do change
frequency with temperature and with certain ... er... "mechanical"
effects such as squeezing the bottom of the case. So even if the exact
voltage to put the oscillator exactly "on frequency" was supplied by the
logic, your oscillator may not oscillate exactly on frequency due to
temperature and other environmental changes. 

Those two things are what limit the frequency accuracy of the K2. The
frequency error you are reporting that your K2 shows sounds very typical
to me. 

For some, it's an irritation that the frequency can't be set to the
nearest Hz with absolute accuracy. For others it seems unacceptable that
it isn't exactly on the 10 Hz frequency shown in the frequency display.
For me, after years of using a simple crystal marker oscillator to tell
me that I was with in a couple of kHz of the band edge, the frequency
accuracy of the K2 is absolutely astounding! 

Back when CB first took over 11 meters there were a lot of stories that
the FCC soon would mandate that Hams also use "channels" instead of
being able to roam freely over our bands. Most Hams were horrified at
the concept and the FCC and ARRL were deluged with letters of concern.
If there was ever any serious thought about doing that, it was soundly
killed by the Ham protests. Nowadays it seems like Hams are DEMANDING
that we use "channels" -- with one group or another insisting that every
5 kHz of most phone bands "belongs" to one net, group or somesuch who
should have primary use of it. These same ops seem to be DEMANDING that
a good rig is one where they tune the dial to 7265.00  -- and any other
rig in the world that is worth owning will be tuned in exactly if it's
dial shows the same frequency so they will know that they are "on
channel".

To paraphrase the Subject line of another thread on this reflector, I'll
say "fooey!" 

Ron AC7AC
K2 # 1289