[Elecraft] K2 frequency drift

[John, KI6WX]

The K2 drift mentioned by Adam is about 5 ppm for 10 MHz WWV.  This isn't 
bad for a non-TCXO oscillator.  With TCXO oscillators, you can probably get 
down to 1 ppm, but it might cost you $50 extra.  You can do even better by 
mounting the oscillator in an oven (OCXO), but it will cost you a penalty in 
both cost and power consumption.  Neither of these options is feasible for 
the K2 design.

Now a few comments about some of the other comments on this thread.

The PLL reference is the major source of drift in the K2, especially on the 
higher frequency bands.  If you reduce the PLL reference drift enough, you 
might find that the BFO drift becomes more significant.  You can measure BFO 
versus PLL reference drift by measuring the frequency on both USB and LSB.

The thermistor/resistor network for PLL reference frequency stabilization 
was designed to compensate both for warmup drift and drift induced by a 
KPA100.  In receive only, the K2 internal temperature will rise about 10F in 
a half-hour.  Running a KPA100 at high power can result in even larger 
temperature excursions.  Adjusting RA can reduce this drift; just follow the 
instructions in the manual.

Finally, the PLL reference frequency stabilization was tested in an 
temperature chamber from 40F to 120F, and the stabilization works well over 
this temperature range.  In one test, the K2/100 was started cold at 59F for 
30 minutes, and then put in transmit at 20 watts CW output for 40 minutes. 
This increased the internal temperature to 114F.  The total VFO drift over 
this temperature range was 70 Hz; the total BFO drift was 10 Hz.  The test 
unit has an optimized RA resistor.
-John
 KI6WX

----- Original Message ----- 
From: <adamkern at uchicago.edu>

> That being said, I still do notice a temperature-dependent
> frequency drift during normal usage.  The drift is
> predictable, with the zero beat on the dial I get with
> reference to WWV drifting by as much as 70-80 Hz if the
> temperature difference between ambient and the max temp the
> rig hits with the KPA100 under heavy use is large.  The
> frequency at which I can zero-beat WWV does not change
> unpredictably.  The radio consistently zero-beats low (e.g.
> WWV zero-beats at 10.000.07) when cold.
>
> Is this a really big deal?  On SSB, certainly not.  On CW I
> like to zero-beat dead-on with the narrow filters.  Perhaps
> I'm just being a perfectionist.
>