[Elecraft] Fwd: Re: Frequency Counter for Calibrating K3S

[Mike Harris]

Slip up at this end, I didn't send it to the list.

And whilst I'm resending I might as well mention that I am using the Leo 
Bodnar GPSDO to drive the REF IN port on my K3. I have found that the 
unit locks up within 30-45 seconds from power up with the antenna in the 
shack, near the north facing window with a clear half hemisphere view.

Totally unnecessary for general HF operation but having driven a 
commercial Intelsat standard B 11 metre dish earth station with all the 
necessary test equipment you can become a bit anal about this stuff.

So for the second time:

The "Method 2" using WWV is fine, however, there is a slight gotcha.

I have the 1ppm TCXO in my K3 and have been using WWV since new (#1400). 
Checking the REF CAL regularly I've noticed the setting changes slightly 
each time, ageing of the TXCO I was thinking, there will be some truth 
in this but it isn't the whole story and once I started using the 10MHz 
ref-lock feature fed from a GPSDO reference a whole new story emerged 
from monitoring the change in the REF CAL parameter, the TCXO frequency.

My TCXO will drift HF approx 28Hz within the first four minutes after 
switch on, perfectly reasonable. It will then plateau for about 26 
minutes and thereafter start to drift a further 27Hz HF for the next 90 
minutes. For the next two hours it will drift HF another 5Hz and stop.

All the time whilst this drift is taking place the front panel is 
heating up, mine is fitted with the sub RX DSP board so fully loaded. On 
RX only the temp changes from 23C at switch on to 40C over the four 
hours. The bulk of this temperature change taking place in the first 
hour, only a few degrees there after.

I suspect that on RX the front panel is the main source of heat within 
the box and over time is the main factor contributing to TCXO drift.

I observed that TXCO drift is not translated directly Hz for Hz to 
calibration accuracy. I did check it on 28MHz and and found that cal 
drift was approximately half that of TXCO drift. Wayne will know the truth.

So from all this I would suggest that a one to two hour warm up before 
doing a REF CAL using WWV might be a good idea.

Using WWV you can do your own TXCO drift measurements. Just do a "Method 
2" REF CAL as soon as possible after switch on and again at 30 minute 
intervals for the next couple of hours making a note of the REF CAL 
readings as you zero beat WWV. Armed with that info you can set the REF 
CAL to a number which indicates a good stability has been reached.

I'm not saying all K3(S) will have the same drift pattern as #1400. My 
TXCO is well within the 1ppm spec and if the first four minutes is 
discounted it is nearer 0.5ppm which has been reported as typical and 
hence no need for a "Method 3".

Clearly checking the REF CAL at random times after power up will give 
varying set points. This is what has made me offer previously that 
setting REF CAL is not a one off deal. Given a bit of ageing of the TCXO 
then maybe it is over a sensible time scale of a couple or more of years.

Regards,

Mike VP8NO