48v phantom power from 9v AC or DC power supply schematic?

[iandy4]

Anyone know of a schematic for getting 48v phantom power from 9v AC or DC power supply?  I couldn't find one but if a usb powered interface can produce phantom power it must be possible.  

[R.G.]

Anyone know of a schematic for getting 48v phantom power from 9v AC or DC power supply?  I couldn't find one but if a usb powered interface can produce phantom power it must be possible.  

Sure. It all depends on how desperate you are.

With 9Vac, all you have to do is use a voltage multiplier (google "%^&*croft-Walton) to make a DC voltage above 48V, then series regulate to 48V. Unfortunately, three terminal regulators do not do this because they're generally limited to no more than 35-40V inputs, so you'll have to do some other regulator. Also, phantom power usually has a big inductor in series with the voltage so the signal isn't shorted by the power voltage, so there's that you'll need to provide.

With a DC supply to start with, you will have to chop it to AC in some kind of switching circuit, then either voltage multiply it or use a transformer or inductor to increase the voltage to higher than 48V, and regulate it somehow. There are switching regulators which do this directly with inductors as the up-verter element. Then there's still the inductor to provide.

The process is fairly complex, generally more complex than the thing it's powering if done right. This is probably why you haven't found easy circuits for the application. It's a rare-ish application, and not easily done with jellybean-style low voltage ICs.

In general, any power supply voltage, AC or DC, can be converted to any other power supply voltage as long as enough power is available and you're willing to work hard enough at it and pay for the parts and development.

Frankly, the easiest thing to do is to buy a 48Vdc output power adapter, which do exist.

[Hides-His-Eyes]

Just had a play on falstad with the C-W thing; looks like it has good potential for running tubes at high voltage in a pedal; a CMOS 555 provides 100mA according to the data sheet, although you'd need a lot of stages to get one of those to 100mA at 100V, say.

Come to think of it, part of me is saying "that sounds familiar" and realising it's probably what Rick Holt uses.

[Processaurus]

With 9Vac, all you have to do is use a voltage multiplier (google "%^&*croft-Walton) to make a DC voltage above 48V, then series regulate to 48V. Unfortunately, three terminal regulators do not do this because they're generally limited to no more than 35-40V inputs, so you'll have to do some other regulator. Also, phantom power usually has a big inductor in series with the voltage so the signal isn't shorted by the power voltage, so there's that you'll need to provide.

I built a healthy mic preamp last year that uses an LM317 to regulate the +48v. 

I would be really interested to know how modern interfaces and pocket recorders do the DC-DC conversion to get a quiet +48v from 5v or 3v, though.

For that preamp project I was looking at wallwarts that put out ~48v, I remember they weren't as common or cheap as I'd hoped, and was thinking about just getting two 24v's and running them in series.  Ended up just getting a kit with a toroidal transformer from this place instead. 

Good question to ask, though, it would be neat to have a little solid state circuit to add phantom power to things.  I don't think it would need to be very beefy, I seem to remember testing the current draw of a Rode NT-2a condenser mic and it being just a couple mA.

[iandy4]

Hey Guys,
Thanks for your input.  I was doing a lot of research yesterday based on your advice.  I found a few interesting things and I was hoping someone could point me in the right direction.

What I am trying to do is add phantom power to a pedal that I am currently prototyping.  The schematic is based off of an existing design but what I've ended up with is this:



My goal is to incorporate the phantom power into the circuit and use one power supply for both this circuit and the added phantom power.  After doing some research yesterday it seems like it is much easier to accomplish this with an AC power supply rather than DC.  I think I can modify this circuits design to work with a 12 VAC power supply.  So is it then just a matter of incorporating the Coc kcroft-Walton multiplier (last schematic on this page: http://www.allaboutcircuits.com/vol_3/chpt_3/8.html)?

[iandy4]

here is a schematic I just drew of what I meant in the post above.  Would this work?  And if so, what would the diode values need to be?  The schematic I drew should work fine with 12VDC rather than 9VDC as shown in the previous schematic.

[iandy4]

Oops, I suppose the voltage still needs to be regulated going to the mic pins.  I have a feeling that the ac is not as easy to convert to 12dc as I have drawn.

[iandy4]

I built a healthy mic preamp last year that uses an LM317 to regulate the +48v. 

Processaurus,
Do you have a schematic of that?

[Processaurus]

I built a healthy mic preamp last year that uses an LM317 to regulate the +48v. 

Processaurus,
Do you have a schematic of that?

It's that kit I linked to, the schematic is the last page of the assembly instructions pdf.

[iandy4]

I built a healthy mic preamp last year that uses an LM317 to regulate the +48v. 

Processaurus,
Do you have a schematic of that?

It's that kit I linked to, the schematic is the last page of the assembly instructions pdf.

O ok thanks, now I see.

[iandy4]

Can anyone tell me if this schematic that I just pieced together would work or if it's even in the right direction for getting regulated 12vdc and 48vdc from a 15vac supply?

If it is, can you tell me what the values of the caps should be in the regulation stages and if they are supposed to be electrolytic?



I've tested the rectifier part of the circuit by using a 9VAC power supply going through  four 1n4007 diodes and then to the original schematic that I posted and it seemed to function just fine without the regulation part of the circuit proceeding it.  The original schematic should function fine between 9vdc-15vdc so do I even need regulation for that?

[Hides-His-Eyes]

I would go with 10u on the left (electro), 100n film or cermaic on the right.

The one on the input has to be higher to stop reverse voltage damage, and the one on the output has to be not too big otherwise it causes problems inside the regulator.

[iandy4]

I would go with 10u on the left (electro), 100n film or cermaic on the right.

The one on the input has to be higher to stop reverse voltage damage, and the one on the output has to be not too big otherwise it causes problems inside the regulator.

Thanks Hides-His-Eyes!
I take it that means the rest of circuit looks like it should work?

[PRR]

> a CMOS 555 provides 100mA ... ... get one of those to 100mA at 100V, say.

A BJT 555 will chunk 100mA, true; but levered from 10V to 100V you also lever from 100mA to 10mA (no free lunch).

Minus losses!

> Rode NT-2a condenser mic and it being just a couple mA.

There's a limit and a spec.

P48 uses 3.4K resistors so the worst load (shorted cable!) is 15mA. That leaves zero at the microphone so is not too useful.... but your supply should not blow-up if this happens.

P48 specs used to say 2mA max, now say 10mA max. Many mikes are just a couple mA, but you don't want your Phantom to collapse if you plug some 9mA pig into it.

PaIa has an old common-design using a hex CMOS and diode/cap chain to make some dozens of volts from 9V-12V DC supplies.

Trying to get your multiplied 48V _and_ your 12V from the same AC winding can be very frustrating: easy to get both separately but when you bring the grounds together something shorts-out.

Why run the opamp on 9V? Most will take 36V. 24VAC into a doubler gives 72VDC. Regulate 48V and 24 or 35V from that.

You want another output DC blocking cap.

The one-opamp diff-amp with two 1K input resistors is not lowest-noise. Probably fine for large condensers, but annoying hiss with dynamic mikes. But when you come down to it, a "Good!" mike preamp is a non-trivial build, and it can be very educational to try all varieties (gad, I'm edju-cated).

[iandy4]

> a CMOS 555 provides 100mA ... ... get one of those to 100mA at 100V, say.

A BJT 555 will chunk 100mA, true; but levered from 10V to 100V you also lever from 100mA to 10mA (no free lunch).

Minus losses!

> Rode NT-2a condenser mic and it being just a couple mA.

There's a limit and a spec.

P48 uses 3.4K resistors so the worst load (shorted cable!) is 15mA. That leaves zero at the microphone so is not too useful.... but your supply should not blow-up if this happens.

P48 specs used to say 2mA max, now say 10mA max. Many mikes are just a couple mA, but you don't want your Phantom to collapse if you plug some 9mA pig into it.

PaIa has an old common-design using a hex CMOS and diode/cap chain to make some dozens of volts from 9V-12V DC supplies.

Trying to get your multiplied 48V _and_ your 12V from the same AC winding can be very frustrating: easy to get both separately but when you bring the grounds together something shorts-out.

Why run the opamp on 9V? Most will take 36V. 24VAC into a doubler gives 72VDC. Regulate 48V and 24 or 35V from that.

You want another output DC blocking cap.

The one-opamp diff-amp with two 1K input resistors is not lowest-noise. Probably fine for large condensers, but annoying hiss with dynamic mikes. But when you come down to it, a "Good!" mike preamp is a non-trivial build, and it can be very educational to try all varieties (gad, I'm edju-cated).

Hey Paul, thanks for your response!  I asked the same question on the diyaudio forum and got a helpful answer similar to yours.  Here is the forum thread for anyone interested, http://www.diyaudio.com/forums/power-supplies/187737-would-work-15vac-12vdc-24vdc.html