> When doing temperature testing, I always put the PCB in a small closed > container lightly packed with crumpled up newspaper to effectively > eliminate drafts. This is the only way I have been able to measure > temperature drift under 100 ppm reproducably. Which newspaper? ;)