# Multiple-microphones and mic-switcher

Trying to make a round-robin (time-slice based) array of microphones, which feed to the microphone input of a PC.

There are 2 things I am looking for some ideas / pointers on:-

1. The distance between mic and pc can be upto 50ft. What can I do to ensure that signal reaches PC with acceptable fidelity ? Do I need to do a pre-amp and feed to line-in of the PC instead ?

2. I need to run upto 20 mics, all connected to the same PC, with a single ordinary sound-card. The mics inputs are switched near the PC, based on a selection logic which is running on the PC. Clearly, I need only input from one mic at a time. Are there some pit-falls I should watch out for with this approach ?

Any existing schematics for such purpose would be really useful, but I do intend to upload a schematic for community validation shortly.

The microphone signals are probably single-ended, and then you want pre-amplification at the microphone end of the cable. If you would transfer the low-level signal over the cable first and then amplify you would amplify any picked up noise as well. Your S/N ratio could be pretty bad. Balanced signals are far less susceptible to this, but usually only professional microphones use this. Most of the time they'll use an XLR connector.

For the switching you could build a summing amplifier like this

and use a 74HC4066 analog switch for each $R_{IN}$.

• Thanks @stevenvh. The purpose of analog-switch is quite clear, although the purpose of summing amp, and it's usage along with the switch isn't. Are V1, V2, V3... coming from the pre-amp stage at the far-end (mic end) ? Sorry, but you might have to dumb it down a bit more... if it's not asking for too much. – icarus74 Apr 7 '12 at 11:57
• @icarus74 - Yes, Vx are the inputs from the microphones. Theoretically you could just tie the switched signals together, but the summing amplifier buffers the signal and can also take care of extra amplification. See also this excellent explanation – stevenvh Apr 7 '12 at 12:09
• Thanks @stevenh. I guess the summing amp would be excellent if I had "mixing" requirement (something I might need in future), but do you see any problems if I tie the switched signals together ? I am not sure I need the extra amplification. – icarus74 Apr 7 '12 at 12:40
• @icarus74 - depends on what what comes after the node where you tie everything together. If that's an high impedance (like > 10k$\Omega$) input you won't need the buffer. – stevenvh Apr 7 '12 at 12:44
• The summing amp is a good way to bring together inputs, even if only one is on at a time. The reasons is that it provides isolation between the circuits. Even if V2 and V3 are off, and only V1 is live, the impedances in the circuitry for V2 and V3 could mess with V1. But it won't if this virtual earth mixing point is used. – Kaz May 7 '13 at 3:48

Fall back to a 555 timer and some discrete logic to feed the computer a signal from the external circuit. The 555 can be the clock, and each tic can be the next mic that switches on, and at the end, it starts back at the beginning. Your code already does this in the computer, but it's faster done outside most likely, depending on how bloated your favorite programming language is. The circuit can feed the computer a blip each time the next mic connects so you know which one is active.

Just a thought.