Thank you for your reply.
What I've tryid so far is
- another PSU (750W, not a year old)
- 2 other graphic cards (MSI 9600 GT, Sapphire GTX 570)
- minimal setup on cardboard
- in BIOS/UEFI disabling the sound device, disabling Intel-C-states
- using both RAM modules one at a time
this all doesn't affect the noice. It does persist.
The noice does not relate to cpu load (tested with Prime95) or RAM usage (tested with MemTest86+).
It does relate to graphic load or as I believe load on the PCIe bus - it changes characteristics and intensity depending on load and on which PCIe standard is in use (1.1, 2.0, 3.0).
I don't to barely get the noice when using integrated graphics instead or using the PCIe 2.0 slot (only tested with the MSI card, since the others' fan frames conflict with the next slot below), while i can make sound with mouse cursor movement.
And the noice can made audible (with not grounded stereo) by grabbing it between
- audio out left/right and ground // with ground != audio out ground
- ground_A and ground_B // with ground_A != ground_B
with ground is out of
- MB back panel
- graphics card back panel
- grounding on power bar
- audio out ground.
As I recall - what might be wrong? - my cpu's working the PCIe lanes via the Z77 chipset, which is why i wonder if the cpu is to be ruled out or must be considered potentialy being the cause of that noice - maybe per current leak? But i don't know to decide on that question.