The choice that puzzled me, I am trying to determine the best way to ground the following audio circuit for best RF immunity. Right now I'm looking at two options, and am not sure if there's even really much of a difference between the two.
Option 1:
Name:  DFNc7.jpg
Views: 394
Size:  58.1 KB
Option 2:
Name:  buG0K.jpg
Views: 369
Size:  57.7 KB

First off: I've been studying up on RFI so please correct me if I'm wrong, but as I understand it any RF pickup by the input cables (V1, V2) will most likely be common-mode, and will be shunted to earth ground through the chassis via the capacitor network (obviously assuming the interference is in the correct frequency range). The chassis is earth grounded.
In option 1, I have two chassis ground points. The caps C11 and C12 would ground to an isolated island on the pcb directly connected to the chassis. Then the signal ground would be tied at a separate location to a separate chassis ground point. I envision this as keeping the RF noise currents off of the ground plane entirely, but there is another separate chassis ground connection to the circuit. Could this possibly result in some kind of loop? I would imagine the RF current is destined for earth ground, and the shortest path is through the CGND at the input, but I'm not 100% sure if it could/would loop back.
In option 2, I've eliminated the second chassis ground connection point and tied the signal ground to the C11/C12 chassis ground point. The signal ground is still only connected to chassis ground in one place, but now there is only one connection to chassis, and the noise currents can flow through the signal ground plane to chassis ground. I would probably make a little half island or lay this out somehow to keep RF noise currents isolated to a small area while still keeping a connection between signal/chassis ground. But there would still be noise currents flowing through this interconnection.
Is either one of these options better than the other?
Update I've added the last half of the schematic as per Dan's response, it was left off initially only because I thought it would make the image too large for some browsers
Name:  yRzg4.jpg
Views: 366
Size:  23.0 KB
Update In regard to the comments in the discussion after DanMills's post, I am including a drawing of how I am visualizing the possible flow of shield currents and why I am having a hard time convincing myself that "option 1" is the better solution.\
Name:  W5D2y.jpg
Views: 305
Size:  42.0 KB
Update Here is an example of how I was thinking to lay out the board to facilitate only one chassis connection for RF noise currents, but keep them off the main ground plane.
Name:  yRzg4.jpg
Views: 366
Size:  23.0 KB
Thanks all guys!