Just doodling with an eye towards building something like this. I would probably use the dead bug technique if I did. It would be used on my lappie or similar for microphones. Low power and small size is a must.
So I'm throwing it out on the forum and see if it sticks.
View attachment 95993
Not that want to, it just does not matter in the real world. I have two amps, I need two amps, so I used the minimum parts configuration.Why do you want to invert the audio?
Because for this application I do not need them. Remember, size and low power is a requirement.Perhaps you should add at least a volume control and maybe some tone adjustment?
The quiescent current of your op amp is 4mA. There are many op amps available with similar specs and much lower quiescent current (0.2mA in this one - giving 20x battery life - most of the time you are not likely testing your voice so current consumption is likely mostly based on quiescent time).Not a bad thought, but I generally like using virtual grounds. The power supply runs down for both sides equally.
You may want to verify that there is no software solution. Here's my experience: I've played around with plugging my guitar directly into various computers, and in general it doesn't work because the signal is just too low. BUT, with software meant for this application, I can get pretty good results without any preamp. I believe the software can increase the gain of the computer's built in amp beyond what is normally available in the system settings. This experience is on a Mac, mind you, but might have a parallel on whatever you're using. Just something worth a look.... the several microphones I have for the use are too quiet.
The virtual ground (signal ground) is developed by the two unbypassed 1meg resistors R1 and R2. U1 is shown for the power supply connections. And U1a and U1b are the same device, showing the signal connections.I do like the idea of not adding an extra op amp to make a virtual ground and splitting the batteries to make your split supply
The bias resistors (R8+R9) are there hold the Vcc/2 voltage on the 10uF cap. It is the cap that creates the low impedance virtual ground. If the cap was not there then the current from the op-amp feed back would drag the virtual ground around, causing distortion and odd gain instabilities. The cap is the important part (not the resistors) and should be sized based on the resistance of the feedback resistors and lowest frequency. Actually in my circuit the decoupling cap (10uF) is too small for a 20 Hz signal with a 412 ohm feed back resistance, but it wasn't designed for a sub-woofer.Note that Les's centerpoint ground is a) 100 times lower impedance; and b) AC decoupled. Without the decoupling there will be significant crosstalk between the two audio channels...