So I am the very pleased owner of KX3 #3899, purchased used from an absolute gentleman of a ham (WB4OQX). I only just recently acquired the KXPA100 amplifier for the KX3 and sold off my FT-950, IF-2000, LP-PAN2, and LDF AT100Pro2. So far with the KX3 as a base station I’ve only run SSB, but I want to get the PKS31 and JT-65HF modes up and running.
To help out with this, I performed the Extended VFO Temperature Compensation Procedure two nights ago. I made some mistakes and performed the procedure again last night with much better results. I’m going to share my notes and observations so others can avoid my pitfalls.
Note: I used the XG50 from Elecraft, which I assembled from a kit. I am relatively inexperienced with a soldiering iron, but I have a good soldering station (fine tip, temperature control) which made it rather painless. I highly recommend the XG50 for this. I also used a hairdryer for the heating element.
1. On my first try, I applied the heater (hairdryer) to the radio as soon as I started data logging after removing it from the refrigerator. This was a mistake. This caused the temperature to rise too fast from the coolest point and the data gathering did not get a nice, wide spread of data points in the cooler temperatures. The second time around, I let the radio naturally equalize to room temperature (as observed before cooling it) and only then applied the heater.
2. On my first try, I kept the hair dryer too close and the temperature rose too quickly to the 52 C mark. How did I figure this out? When the procedure was complete and the data logging finished, I switched off the heater and let the radio equalize naturally back to room temperature. I watched the temperature drop and saw that the baseline frequency oscillated over 20Hz from the 550Hz. It should have been maybe 2Hz off, so I knew something was wrong.
3. On my first try, I attached the XG50 directly to the KX3. Doing so allowed the XG50 to be heated by the hair dryer due to its proximity to the KX3. I believe this affected the signal output, making it inconsistent. On my second attempt I attached the XG50 to the KX3 using a three foot coax cable which kept the XG50 well away from the heater.
After the second attempt at the procedure, I let the radio cool naturally and watched the temperature drop and saw that the frequency varied at most about 2-3Hz all the way down to room temperature from the 550Hz baseline. This was much better the the first attempt.
So now I have a reasonably temperature stable KX3, primed and ready for JT65, JT9, PSK31 and just about any digital mode I can throw at it.
Next up, my first foray into PSK31.