OK !
One trouble is solved :
The osciloscope was grabbing the sampling frequency.... Thank you HP for your good osciloscopes....
This morning i used an analyser and can't see the FS at the input...I used the same cables, the same power supply.... So that the 360-170-~400Hz is no longer in the spectrum. I still have a 1.1kHz, 3dBV over the background noise but i think i can manage it.
The aliasing filter is certainely helping.
I will keep you in touch in the case this annoying noise is here again.
Yann
