# Help me understand negative voltage

#### mcardoso

Joined May 19, 2020
198
Probably a very elementary question, but I've never felt like I had a solid grasp of this.

In audio and analog circuits, I've sometimes seen the use of dual rail power supplies such as +/- 15V. These have been to power op-amps and similar circuits.

If I understand correctly, the positive rail flows current from the source rail to GND. and the positive rail flows current from GND to the negative rail. Is that correct?

My two questions are the following:

1) Why would you need a negative voltage rail? What would lead someone to engineer one into a design?

2) How does a dual rail power supply work? Do you make a 30V supply and tie a rail halfway between the 30V and 0V to GND (thus giving 15V above GND and -15V below GND)?

Thanks!

Joined Jul 18, 2013
23,897
Some circuits require the split supply from Op amps to power devices, the 0v point is a COMMON reference to either of the other two.
This allows for bipolar switching etc, many other reasons.
Max.

• mcardoso

#### crutschow

Joined Mar 14, 2008
28,154
Why would you need a negative voltage rail?
AC signals, such as audio, are often centered at ground so the signal has plus and minus peaks.
So it is simple to handle these signals with circuits that have plus and minus supplies so they can readily handle such signals.
But you can also DC shift such a signal so the negative peaks do not go below ground, and thus use circuits with only a single supply.

#### MrChips

Joined Oct 2, 2009
24,593
There is no such thing as negative voltage (ok I can handle the blowback).

What we measure as voltage is a voltage difference. That is, the voltage measured is the electrical potential between two points A and B, or the potential at A with respect to B.

If the voltage at A with respect to B, i.e. VA - VB is a positive quantity, then VB - VA must be a negative quantity.
This is our reality of a negative voltage.

An electrical circuit may call for a dual supply, +Vs and -Vs. That is so because +Vs is referenced to a COMMON node while -Vs is referenced to the same COMMON node.

You can shift +Vs and -Vs together by any amount as long as we maintain the same potential with respect to the new COMMON. This is done to turn a dual supply circuit into a single supply circuit. For amplifying AC signals this will work fine as long as we understand that all voltages in the circuit are now to be referenced to a new COMMON.

#### MrChips

Joined Oct 2, 2009
24,593
From the perspective of an amplifier, look at it this way.
The amplifier is powered by a power supply that has two connections, VA and VB.
Since you do not want to clip your signals, all signals, input and output should be balanced between VA and VB.

#### hrs

Joined Jun 13, 2014
351
Isn't part of it that circuits such as a power amplifier require a very low impedance ground? Something that is more difficult to achieve with a DC shift than a true split supply.

#### WBahn

Joined Mar 31, 2012
26,398
Probably a very elementary question, but I've never felt like I had a solid grasp of this.

In audio and analog circuits, I've sometimes seen the use of dual rail power supplies such as +/- 15V. These have been to power op-amps and similar circuits.

If I understand correctly, the positive rail flows current from the source rail to GND. and the positive rail flows current from GND to the negative rail. Is that correct?

My two questions are the following:

1) Why would you need a negative voltage rail? What would lead someone to engineer one into a design?

2) How does a dual rail power supply work? Do you make a 30V supply and tie a rail halfway between the 30V and 0V to GND (thus giving 15V above GND and -15V below GND)?

Thanks!
In many circuits, including audio, the information is contained in a signal that varies "up" and "down" relative to the when the information content is zero. In order to accommodate signals that go above and below the "zero information" level, that level usually must be somewhere between the two rails of our power supply. So we bias our system to place the "zero information" value at a suitable point between them, with the midpoint being a very natural candidate.

Since all voltages are always a voltage difference between two points, if we are going to talk about the voltage AT a point, we need a reference point to which that voltage is referred -- what most people sloppily call "ground" but for which "common" is a better name. In many circuits, we use the negative power rail so that all of our voltages are positive -- something which tends to reduce the mistakes that humans make when working with negative values. But it is also natural to want to call the voltage of the signal when the information is zero our 0 V point and if we do that, then our positive rail has a positive voltage and our negative rail has a negative voltage, even though nothing has actually changed other than which node in the circuit we have chosen to call "0 V".