[Elecraft] KPA freq drift

[lhlousek]

I did some further experiments with frequency drift of my K2/100.

In a cool morning shack and after being in RX mode for about 20 minutes or so I measured the internal temp of the rig at
23C.  The K2 read 10Hz low on 40m and 340Hz high on 10 meters.  Later, with a room temp of 81F and internal rig temp of
33C 40m was 40 Hz low and 10m was 150Hz low.  After 20 minutes of continuous CW TX at 100W in a 84F shack the internal
temp of the rig was 50C (heat sink was 58C on the K2's display) and 40m was 120Hz low with 10m was 430Hz low.

Taking the temp and frequency extremes this works out to ~-0.6ppm/degC at 40 meters and ~-1.0ppm/degC on 10 meters.
These numbers are in line with frequency drift measurements of other K2s as listed in John's (KI6WX) "Reducing Frequency
Drift in the Elecraft K2" applications note. http://home.pacbell.net/johngreb/reducing_k2_frequency_drift.pdf

Granted, it is highly unlikely you would be operating under circumstances where the internal temp of the rig would go up
by 27C but I have had it go up by 15C during the course of a long CW ragchew at 100W.    That translates into a 60Hz
drift on 40m and 480Hz drift on 10 meters.

I can think of a number of ways to address this:

John's mod might be of some help.  He did a lot of work to improve his drift by a factor of 2.

Stick it in an EC2 (and don't set the K2 on top of it).

Change how the rig is ventilated.  The way it is now, the fan draws cool air in at the rear, blows it through the PA
section where it heats up and is then mixes through the rig before exiting at the front edge of the heat sink.  In the
process, it heats up the rest of the rig.  If the fan direction were reversed, the front of the heat sink blocked, and
vents added near the bottom of rig, cool air would be drawn into the rig and then flow through the PA before be
exhausted by the fan. Warm air would not circulate through the rig heating things up.  The vents would have to be in the
side and/or rear panels above the level of the RF PCB instead of below it or in the bottom since there is no path
around.  With this setup the heat sink would loose the benefit of having the air blown through the front edge of the
fins but it might work reasonably well without blocking them. Or, perhaps, move the fan so that it blows into the rig
below the level of the PA shield.

Lou W7dzn