I'm designing a musical instrument in which two oscillators (one based on a Quad NAND gate, the other on a 555) are controlled by a signal from a condenser microphone. At the moment the mic is being phantom powered via my computer audio interface, then coming from an output of the interface and entering the circuit at line level.

I now need to add a 48v phantom power circuit so that the mic can be powered directly from the instrument. I've purchased a 12v-48v DC-DC converter (this one) which seems to work fine. My question is about how to incorporate this without damaging the subsequent circuitry.

This is essentially my proposed circuit. Phantom power works by putting both signal pins of the XLR at 48v with respect to ground; the mic signal is amplified by a LM386N-4 power amp and this signal controls the oscillators via an LED-LDR arrangement. I haven't included the rest of the circuit, since I'm mainly worried about whether the phantom power will damage the LM386N-4. (I'm also not too bothered about audio fidelity, as the mic signal is only used to control the oscillators, and won't be directly audible - it just needs to be high enough going into the rest of the circuit):

phantom circuit

What I'm particularly worried about is blowing up the LM386N, perhaps by it internally referencing either of its inputs to ground (and hence seeing 48v across some of its internal structure). Will this circuit work, in the sense that it will amplify the mic signal without damaging anything?

Incidentally, the phantom power circuit proposed here suggests putting a pair of zener diodes between each signal line and ground, as shown below - would using these in my proposed circuit prevent damage to the LM386N-4, or any subsequent circuitry?

enter image description here

  • \$\begingroup\$ Yes, the Zeners (or clamping diodes to power supplies) are necessary. If the capacitors are charged up and then something is plugged into it that shorts them to ground, the Mic+ Mic- nets will be at -48 V and damage whatever comes after them (which then produces cool random crackly noises) \$\endgroup\$
    – endolith
    Jul 5, 2018 at 18:45

1 Answer 1


The second schematic is right. The first has polarity the wrong way around and a dead short across the signal lines. You've also got the polarity wrong on the LM386 so be careful!

The Zeners are a good idea. When the circuit is powered up the left side of the capacitors will lift to, maybe, 24 V (depending on the load presented by the mic) and the right side will try to as well. That's the characteristic "thump" you hear on powering up an amplifier with a single supply rail. The Zeners will clamp the signal to a value that won't damage the opamp (but may still offend your ears).

I didn't know if the LM386 could be used in differential mode as you've drawn it. A quick image search for "LM386 differential" shows that others have at least drawn it so you won't be alone in trying!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.