How do I split signals without crosstalk?

[Sage]

This is a circuit I designed to mix together four different inputs.  Input 1 is a line input from an amp modeler (Kemper, Line 6, etc.).  Inputs 2, 3, and 4 come from individual Mic preamps, so they should also be at line level by this point.  The idea is that the three Mic preamps will have their signals split into wet/dry signals.  Opamp U1 in this schematic mixes all the dry signals together, including the signal from Input 1.  Opamp U2 mixes the Mic signals together and feeds them into an Accutronics/Belton BTDR-3 reverb brick for a wet signal.  Opamp U3 combines the wet and dry signals back together.

Here's the issue: Input 1 should be unaffected by the reverb.  However, it seems to me that the only thing preventing Input 1 from leaking into Opamp 2 in this circuit is a couple of resistors.  Is that enough to isolate that signal, or do I need to be adding something else?  How do I split Inputs 2, 3 and 4 without allowing crosstalk from Input 1 to enter into the reverb chain?

[ElectricDruid]

This will work as drawn. The reason it does is because the op-amp end of R1 is a "virtual ground", a point held at ground by the op-amp. Op-amps try to keep both inputs at the same voltage, remember, and the +ve input is grounded, so the op-amp tries to keep the -ve input at ground too. If we assume that it manages that feat, there is no voltage at that point that can bleed through the other resistors to the reverb mixer.

Tom

[PRR]

What Tom says. Also any signal sneaking "backward" through R2 meets the output of the mike preamp which is typically low-Z. So there are *two* severe cuts in the sneak path.

Put numbers on that thing. The virtual ground mixer node is typically below 1 Ohm for low audio and a couple Ohms at the top of the audio band. Pencil 1 Ohm. Fed with 10K this is about 10,000:1 cut of voltage.

That wee bit does pass "back" through R2, and meets the preamp. This will often be around 100 Ohms, So we have another 100:1 cut along the sneakage path. Result may approach -100dB which is more than good enough for the BBC (which may have two *different* programs flowing in the same desk). Real-world is never this good; but should not be the least problem unless you listen for it in testing.