Issue with 48V and 24V power supply for mic preamp

[chvario]

Hi,

I am building a mic preamp that needs 2 different DC voltages available in one chassis: 48V for phantom powering condenser mics, and 24V for powering the amplification.

I thought I could come in with the 48Volts, and then make a voltage divider that would supply an L7824CV with enough voltage (~26V) to get me a stable 24V at its output.
I got a 48V regulated wall power supply by Netgear, which says it can supply up to 1.45Amps.

First surprise: When I put it together on the breadboard, I noticed that I needed to lower the top resistor connected to 48V significantly, once I connected it to the input of the L7824, because connecting it made the voltage drop at the input.

I came up with a voltage divider like this:

+48V---2k2---L7824-in---6k8---0V-

This gave me the regulated 24V at the L7824 output. So far so good.

But:
Once I connect the amplifier circuit to the output, the voltage at the L7824 input dropped under 18V and kept dropping the longer it was connected.

I tried 2 other setups:
1. When I connect the amplifier circuit to my adjustable power supply, I see it uses 0.08Amps and works fine.

2. When skipping the voltage divider and the 48V power supply, and going into the L7824 input with > 26V, then the L7824 output provides the 24V just fine with the amplifier connected, and my power supply says it uses just 0.05Amps.

I am puzzled, but then I am somewhat new to electronics and am probably forgetting about something obvious.
From my understanding and 'instinct' things should work: The amplifier circuit doesn't appear to be drawing all that much current, and the wall supply promises 1.45 Amps.

Why does the voltage drop? And how to fix my power supply?

Thanks a bunch, Chris

 

[JonSnell Electronic]

The 48V supply is constant current, so any voltage above 20Volts is acceptable as long as it is reasonably smooth. It is usual to produce the lower voltage and then a simple voltage doubler to produce the 48V supply. Have a look at the Allen and Heath mixing desk power supplies for ideas.

 

[BZed]

First the input of the 24 volt regulator needs to be higher. The LM7824 will work up to 40 volts. That means that you only want to drop a little over 8 volts, maybe 10 or 12 would be the correct amount. That lets the regulator work with +- 36 volts as the input to provide the 24 VDC you want. A little ohms law will work for you now. You need to get an accurate current measurement first. Use your DVM in current mode in series with the positive input to the regulator. Set you adjustable supply to 36VDC and power the preamp from the regulator. Once you have that current value you can calculate the resistor value to give the 36 volt input to the regulator. For example 12 volts drop / 80 milliamps = 150 ohms for the series dropping resistor. Don't forget to calculate the resistor wattage. In this case 12 VDC X 80 millamps = .96 watts. You should use at least a 2 watt resistor.

 

[chvario]

Thank you, Jon.

The 48V supply is constant current, so any voltage above 20Volts is acceptable as long as it is reasonably smooth.

Can you elaborate on that sentence? I don't understand where > 20Volts voltage is acceptable, and for what.

Also, could you give me a little more detail on what a simple voltage doubler would look like? I have difficulties finding schematics that would step up 24VDC to 48VDC.

Thanks, Chris

 

[chvario]

Thank you BZed, for the step-by-step. That's going to be helpful. I will do the measurements now and will post what I come up with.

 

[chvario]

My adjustable supply only goes up to 30, so instead of measuring I started experimenting a bit with different value 2W resistors.
The resistor that would get the L7824 input up to around 36V was the 470ohm one. For a while I had that and 24V on the output, going into the circuit. But after around 20 seconds, the voltage at the input started dropping until it got below 20V after a minute or 2.
Again Ohm's law:
This should mean that the current pulled by the L7824 increased from 25 milliamps to 60 milliamps, since the resistor stayed constant, right?
So the amplifier is demanding more current over time?

Does that make sense? And if yes, does that mean that I should go with the 24VDC input instead?

 

[BZed]

30 volts is fine for testing to determine the TOTAL current used by the regulator and the preamp circuit. 470 ohms is way to much resistance to provide the correct input voltage to the regulator. Trying to set the input voltage with just the regulator will result in the problem you are having. What matters is the value of resistor needed to get the regulator input into a range from 27 to 40 volts while in operation with its output loaded by the preamp. I would shoot for somewhere in the middle, 32 to 34 volts for example. TOTAL load is what is important.

 

[JonSnell Electronic]

Thank you, Jon.


Can you elaborate on that sentence? I don't understand where > 20Volts voltage is acceptable, and for what.

Also, could you give me a little more detail on what a simple voltage doubler would look like? I have difficulties finding schematics that would step up 24VDC to 48VDC.

Thanks, Chris

20Volts is acceptable as a phantom power supply. The cheaper mixers use 15Volts. It doesn't need to be too smooth as the noise rejection of the balanced inputs will reject any noise.

 

[chvario]

Thank you, Jon. I want to get it RIGHT, so I am sticking with the 48V.

I ended up building a voltage divider after all, but needed 4W(!!!) resistors to deliver the load consistently. A voltage divider with 100ohm and 820ohm delivers the 34 Volts on the input of the voltage regulator.

I got there with experimentation. Math would have told me I should have 42V with that divider. Does someone know why this is working? I would like to understand what I did before I move on

 

[Ketje]

Why not doing it like this.No loss in big serial resistors.
Depending on the current needed the capacitors (220µ) could be smaller.
Mona

 

[JonSnell Electronic]

The 48V rail must be current limited to 100mA for obvious reasons. Take a look at how A&H and other manufacturers do it.
Have a look here https://www.eeweb.com/blog/extreme_circuits/48v-phantom-microphone-power-supply

 

[Ketje]

At eeweb.com I see a 0.4A regulator and no current limit exept from the IC itself ?
If you use a 78L24 in place of 7824 the current is limited to peak 150mA.Eventualy a resistance between the bridge and capacitor on the positive side so that the voltage drops below 50V (referenced to ground !) when the current reaches 100mA.
Mona

 

[JonSnell Electronic]

Unfortunately, when the ic current regulates, it breaks down and must be reset by powering off then on again. (It is only to protect the gear against idiots using an unbalanced mic lead with phantom power on it). The 783 will sit all day with up to 400mA load.