DC Power Polarity Auto-detect

[R.G.]

Aaaaaand we get back around to the MOSFET polarity protector. A MOSFET, a resistor or two, and some juggling, and you can get series polarity protection that takes even less forward drop lost than a Schottky, which has about a half-a-silicon-diode drop across it.

This was a little impractical when small MOSFETs were $1.00, but now that they're pennies...

[Rixen]

micro USB 5V power input with isolated DC-DC converter. Except for noise from the converter (not insurmountable), all problems solved... and users can power from one of the hundreds of old chargers they have lying around, as well as portable power banks.

[reddesert]

My main reason in even entertaining this subject was largely academic. I always use a series schottky in my builds and 90% of the time is fool proof. However, the 10% that plug AC power supplies into DC pedals is why I am going down this rabbit hole. IME, when an AC supply is mistakenly used it kills or moderately damages so many things, it's just better to replace everything. If I could find an inexpensive way to avoid that scenario, I'd implement it. Mind you I would not change the labeling on the pedal. It would still say 9-18VDC center neg

I may be missing something. If a user plugs a common 9VAC or 12VAC wallwart into a pedal with a series-protection diode, won't the diode half-wave rectify the supply? The circuit should never see the reverse voltage. Depending on the supply bypass capacitor, it might even smooth the half-wave enough to work. Of course, there are a few chips that could cook due to overvoltage from a 12VAC supply rectified to 12-15 VDC (some charge pumps), but protecting against all overvoltages is a more complex problem.

[PRR]

> From the datasheet it appears they want you to shoot for a 30mV drop across each mosfet.

"They" don't know what YOU are doing. Common audio with (often) common power supply.

30mV drops may (or may not) mess-up the ~~100mV audio, depending upon details your customers would not understand.

[R.G.]

I may be missing something. If a user plugs a common 9VAC or 12VAC wallwart into a pedal with a series-protection diode, won't the diode half-wave rectify the supply? The circuit should never see the reverse voltage.

Yes, that is what would happen.

Depending on the supply bypass capacitor, it might even smooth the half-wave enough to work.

That's a solid "maaaaaaybe". It usually takes more capacitance than a common value of bypass to make this happen. But it's possible.  The pedal would then be eating power with a strong ripple voltage on it, and most pedals would show that up as unacceptable hum.

Of course, there are a few chips that could cook due to overvoltage from a 12VAC supply rectified to 12-15 VDC (some charge pumps), but protecting against all overvoltages is a more complex problem

And maybe the bypass and other caps in the circuits, as well. A 12Vac supply is rarely just 12Vac from the necessity to make the zero-load voltage on a small transformer larger than the nominal voltage so it will still have the nominal volltage when it sags down under full load. It's not uncommon for the unloaded or lightly loaded voltage on a small wall wart to be 15-20% higher than nominal. The rectified voltage from a sine wave AC source is 1.414 times the RMS, so a 12Vac source rectifies to 16.968 volts, minus a diode drop in this scenario. Add 15% for low loading, and you get to 18.8V.

You're right, protecting against all overvoltages is a more complicated issue, and probably an insoluble one. There is no way I can think of to do it; after all, a lightning strike is one manifestation of an overvoltage. But for modest over voltages, maybe up to 40-50V, it can be done. A series regulator with a MOSFET pass device for low forward drop comes to mind.

[PRR]

> lightly loaded voltage on a small wall wart to be 15-20% higher than nominal.

For some values of "small" (times some values of "cheap"), no-load can be 35% higher.

No matter how careful you are, some penny-pincher will go a little further.