I set up a breadboard in order to conduct tests on diodes to compare clipping harmonics. The idea is to see if there is really a difference between different types of silicon diodes when used as clippers.
I adjusted the test rig to drive the diodes to 10% distortion to make the comparison. While adjusting the setup, I noticed that after 5 to 10 seconds, the percentage of distortion would slowly go down. The change was very slow but noticeable. I repeated the adjustments several times with the same results. It then struck me that maybe the diodes were heating up and the forward voltage was changing ever so slightly as they warmed.
I started the oscillator and waited for the diodes to begin the slow change. Once the steady decline started, I blew on the diodes to cool then off. Sure enough, the distortion change stopped then slowly reversed! Once I stopped blowing on them, the slow decline began again.
The change I was looking at was very slight but easily seen since I was looking at the distortion percentage down to 3 decimal places. I doubt the change would be audible but it is interesting to note that the effect is definitely there.
In the end, I got around it by sending a short burst of the test signal and measuring the distortion before the diodes had a chance to heat up.
