Here's my plan:
1) All relays will be normally open except the selected signal relay, which will remain normally closed so the signal doesn't have to pass through the coils of the relay
2) Since I don't like dealing with too many parts mounted to the chassis, I'm going to use a simple 6-position rotary switch (I ordered the $20 GrayHill one from Welborne Labs
) as an input selector - but all this switch does is connect a "high" signal (5 VDC) to any of 6 input pins on a microcontroller (I have an Atmel AT90S2313 ordered). On power-up/reset, all output pins that trigger the relays are first high (to disconnect all input signals). The microcontroller will check to see which pin the rotary switch is signaling high, and will select that source as an input (outputting a corresponding pin low to un-trigger the relay for that source). However, to implement the remote control, the output of the IR's microcontroller from Reynolds Electronics will set 1 pin of a 2nd set of 6 pins high. The microcontroller at this point will use a "latest-input policy" to select the output. So if the rotary switch is set to one source, then the IR IC outputs another source, the microcontroller will switch the relays for the source picked by the IR chip. But if you go and move the rotary switch, the microcontroller will change the relays for that setting.
The microcontroller output to activate the relay circuit will have to be experimented with, since the microcontroller pins can only output like 20 mA, and the relays that I'm looking at are 28-40 mA or so.
Edit: I just realized that the AT90S2313 doesn't have 18 I/O pins... I'll probably test the algorithm with that chip then move to the 40-pin AT90S8515, since I have a few of those already.
3) Volume control using the IR chip, the 16-pin DIP L293D chip from STM electronics (ordered from Mouser), the 10k Alps motorized pot from PercyAudio.com, and the corresponding PCB available from Percy for the motorized pot. I'll have to experiment with the circuit involved with the h-bridge to see what the current limit should be for small motor movements (hence small volume adjustments).