[Elecraft] TCXO drift and calibration

[Wes]

Your first number is bogus, but generally speaking if your other numbers are 
correct, that does seem excessive.  That said, crystals do age and some of this 
depends on power on/off cycles.  I'm now to the point of leaving my K3S powered 
continuously.

I recently ran some drift tests and found that it takes about 25 minutes for my 
reference to settle to a point limited by measurement jitter.  There is an 
initial positive increase of a few Hz at power up (at room temperature) then a 
non-linear decrease of about 25 Hz.

(Direct measurement of the synthesizer output with a FA-2 counter by BG7TBL, 
with a Bodnar GPSDO reference)

Wes  N7WS

On 12/4/2019 1:33 PM, Gary Peterson wrote:
> Last February, I purchased a factory assembled K3S including a 1 ppm TCXO.  The transceiver is serial number 11803 and the firmware was updated on May 5th of this year.
>
> Last July, I noticed that many of the signals on 50.125 MHz sounded as if they were a bit off frequency as they were higher pitched.  This led me to believe that the reference oscillator may have drifted, possibly due to crystal aging.
>
> Subsequently, I used method 2 and WWV, as described on page 53 of the Elecraft K3S manual.  The REF CAL frequency was lowered from 49.999.996 to 49.379.448.  That seemed to make the signals on the 6 meter calling frequency sound like they should.
>
> A few days ago, I caught a six meter opening and signals on 50.125 sounded higher pitched again.  I repeated the reference oscillator calibration again, using method 2.  Now, the REF CAL frequency is set to 49.379.110.  I have not had the opportunity to hear any six meter signals, since I changed the reference oscillator.
>
> I am wondering if the REF CAL frequency change that I have seen in less than a year is normal, or whether I have a TCXO problem.
>
> Any thoughts would be most appreciated.  Thanks in advance for any time taken to reply.
>
> Gary, K0CX
> Rapid City, SD