> I've never seen scientists burn out over failed experiments.
It's never a single failed experiment.
I can't think, off the top of my head, of a scientist that I've known professionally that I didn't see burn out at some point or another (besides the PIs). For those that I have been watching their experiments, it usually happens after a slew of experiments fail (anywhere from 1 month - 6 months worth). I've probably met about 75-ish scientists that I've watched burn out.* Of these, about 25 or so I have working knowledge of what they were doing around the time of burnout.
*one guy, after three successive failed projects that his PI gave to him, had a very characteristic pattern. He would come in really late (4pm or so), then play game boy tetris on a java applet for about 3-4 hours, then do a little bit of science, and come back and do more tetris. Some of the things he would do - using pipette tips in geometric patterns.
I mean it makes sense. These little achievements provide dopamine hits reinforcing self-worth after completion of a minor task.