48V phantom supply for two old mic preamps - picture inside

[HeaD]

Hi all,

I'm placing into a rack enclosure two old microphone preamps pulled from an odd vintage console. These beasts sound really good, and I would like to add a 48V phantom supply circuit, so I found and already built this little phantom powering project:



My problem is that these Mic Preamps are positive grounded, so they carry the supply voltage (24V DC) to the microphone XLR GND. I need some help here in order to make this circuit work in this odd scenario. Can someone please give me any hint?

Thank you so much.

[Seljer]

Ground is ground, its not +24V, it's that the preamp's own supply is -24V. Reference your phantom power supply to the same ground.

You can generate +48V  with a simple diode & capacitor voltage multiplier circuit https://upload.wikimedia.org/wikipedia/en/thumb/b/ba/Stacked_Villard_cascade.svg/400px-Stacked_Villard_cascade.svg.png , you can feed it off the either AC side of the main bridge rectifier of the main PSU. Or use something like a 555 timer other IC on the already rectified negative DC supply to get a higher frequency square wave so you can get away with smaller capacitors in the multiplier circuit and its easier to smooth out the ripples (at 50Hz you'd probably have to use a bunch of +100uF electrolytics)...of course this then brings up the issue of having an many kHz square wave right next to sensitive mic preamps.
It may be easier to just add a second AC transformer just for the phantom supply (doesn't need to be very large 2 or 3VA is enough for the current draw of the phantom power of just 2 microphones).

[HeaD]

Hi Simon, thank you for your reply.

I actually built my 48V generator using this schematic:


So, do I just need to reference this circuit to the same ground (+24v)?

Thanks

[Seljer]

Yes, but use a separate transformer than the one that is powering the rest of the preamp.

Transformers provide galvanic isolation. And all voltages are relative so it's up to up to decide which point to connect to "ground" for your reference of 0V.

[HeaD]

Ok, I will try this.

Thank you so much!

[PRR]

> 48V phantom
> these Mic Preamps are positive grounded
> work in this odd scenario.

It is not odd just unusual.

All DC power supplies are like batteries.

As Simon mentioned, you need separate "battery" for each supply.

So take a 48V battery and connect negative to common ground.

Take a 24V battery and connect positive to common ground.

Take the + end of the 48V to Phantom live power.

Take the - end of the 24V to preamp live power.

[HeaD]

Thank you Paul, it's even more clear now.

[HeaD]

Hey guys,

I just ordered a brand new transformer to use for the phantom power supply and I'm waiting for it to be shipped. In the meanwhile I was experimenting using a single 24V 50VA transformer, like in the diagram below.



The two PSU boards works properly if tested separately, but when I connect the 48V negative and the 24V positive to common ground, the R2 resistor in the phantom psu circuit blows up. I guess using two separate transformers that should not happen... or maybe I'm doing something wrong here?

Thank you!

[Seljer]

If it's wired like that diagram you're not using two separate transformers! You're shorting out the power rail directly to ground hence the blowing up of resistors. I'd take some measurements to make sure you didn't damage the main bridge rectifier as well by any chance!


By using a separate transformer for the phantom supply, you accomplish galvanic isolation, meaning you can ground it at whatever point you like you like to establish your reference to what 0V is in your circuit and nothing will blow up.

[PRR]

> experimenting using a single 24V 50VA transformer

Simon and I both emphasized TWO separate supplies, so that "you can ground it at whatever point you like" (need).

> 50VA transformer

One phantom mike is absolute-maximum 14mA or 0.014 Amps. Raw supply is 72V. This is close to one VA. You need to reound-up a bit because AC/DC conversion is tough on the transformer, but 2VA would be plenty ample. Transformer costs are usually minimum around 10VA... bigger cost more but also very-small costs more.

BTW, for initial tests, five 9V batteries in series will power one Phantom mike for days, and avoid ALL trouble with hum and grounds.


You have not revealed what preamp this is. Few Germanium-era preamps ate more than 30mA at 24V. This again is about 1VA raw DC or under 2VA of transformer.

So 50VA is about 10 times more than you need. It may be cheap; however you WILL have hum problems between power transformer and mike input transformer, and a larger core may be much worse.

There is probably a way to derive both supplies, correctly grounded, from one winding. However if done wrone it smokes, and I need more coffee before attempting that.

[Seljer]

I've been messing about with LTspice voltage doubler circuits and came up with this. Please someone here chime in if I'm doing anything horrendously inappropriate. In the simulation it seems to hold up (with the chosen values of caps and such for a "worst-case" 50mA load on the doubled positive supply the voltage drops down to 52V). Follow it by the voltage regulation circuitry from the sound.westhost.com phantom supply to get it down to 48V and get rid of the ripple and you may be in business.

[PRR]

I know you already have your fancy Phantom circuit, but---

This is the way _I_ would be thinking. It only requires one winding. It is very much in keeping with Germanium era thinking. It does put an unbalanced current in the transformer, but only 0.2A RMS which a 24V 10VA core can handle well. The preamp side is half-wave, high ripple; but with the low cost of large caps today, and a several-stage filter, the ripple at the preamp is just below what I find in my 1956 Transistors book. The Phantom side uses simple Zener regulation-- while this is often not a best path, in this case the mix/max currents are known, and the possible variation in raw voltage will not stress low-cost resistor and Zeners.

-24V is NOT regulated-- that would have been un-common in the day. But you do have to adjust the two series resistors R7 R8 to get the output very-near 24V. Start with 470 or evem 1K, work down gradually.

Crayon-sketch and sim-grab. After flogging the sim I decided some values could change, but they are not critical. And if you wish, use 100V caps everywhere... it may be cheaper to get ten all the same than two at 100V and four at 50V.