I have been curious wether humidity plays any significant role in the degrading performance of the U800, but I was wrong. I used a hair dryer to blow a stream of hot air through the ventilation slots while placing a fan to draw the air from the rear for a few minutes. After switching the oscilloscope, there was no trace. So it's the heat that causes the U800 to strike!
I placed an air blower type fan against the ventilation slot of the U800, and after a few minutes I turned the oscilloscope on, it worked fine until I pressed the X10 MAG the trace began to fade away to the left after distortion of the wave!
I read some interesting ideas on "baking the U800" and avoiding mechanical stress to it.