Jan, dealing with this high frequency stuff is very very difficult. You cannot have a simple look with the scope, change here and there something and find out what's going on! It's too complicated. Everything is mixed together with this high frequency stuff. You can toy days long (as I did!) fumble her and there, but you cannot really make any logic analysis, because you don't have separated issues, when dealing with high frequency.

Only one point: What you see at the scope depends on what probe you have, where you connect the ground pin, what charge is injected to the shield via stray capacitance and how the probe cable deforms electromagnetic fields. By using different grounds for the probe, by holding the probe differently with your hands you can make every wanted and unwanted signal to appear on the scope. There's just no way to learn by slight modifications and by doing simple measurements. For instance, if you only want to measure common mode noise, you cannot do this by a simpe scope. For this task ultra expensive differential probe scopes were built. Because it's so much difficult to see anything useful on the scope.

Why not profit from the experts work and experience? Just to use a solid ground plane? Isn't it enough, that Mahmoods applications does suffer from interference to accept that a solid ground plane must be used? Ok, if you want to learn, then modifying and looking with scope might be helpful to some extend. But what I'm superiorly interested in is, to get an application to work. Fast to work. Not all time worrying about whether it will disastrously faily eventually.

Kai