So you're using Autotune as an effect on an aux channel, then using a send from the vocal channel?
I'd put autotune as an insert on the vocal channel.
Alternatively, you could change the output of the vocal channel to go to a bus rather than an output.
Or if you're working how I assumed you...