Is there truely such a thing as a negative DC voltage (with reference to earth ground)?

I understand the concept of a negative voltage. Ground is a reference point and it's easy to get a negative voltage if your ground reference is above earth ground. Does a differential power supply actually produce a negate voltage (referenced to earth ground) or does it just use a voltage offset from ground as a ground reference.

For example a +-5V power supply can easily be created by using two voltages (+10V, +5V) and GND. If you use +5V as ground, it will make +10V the +5 equivalent. GND will become the -5V equivalent. The problem with such a circuit is that ground isn't compatible with earth ground. For example you can't then tie your newly created (+5V offset) ground to earth ground without creating a short circuit.

Voltage may be interpreted as a measure of the potential energy of a theoretical test charge divided by its magnitude (dimensionally, volts are joules per coulomb). As with many potentials in classical physics, this is a relative quantity, meaning that your measurement of a potential is between some two points or states.

Physically, there is nothing incompatible with the idea that the relative potential of some point P is negative when referenced to earth ground, or any other reference for that matter. All it means is that a theoretical positive charge would lose some amount of potential energy when moving from your earth ground or reference to P.

We can easily realize this by demonstration with a number of examples that should prove that there's "truly" such a thing as a negative DC voltage with reference to earth ground:

1. Electrochemical: Suppose that the ground pin at left is physically earthed by bonding it to earth ground (e.g. by burying an electrode into wet dirt outdoors). The -5V label clearly and demonstrably has a negative voltage with respect to earth ground.

simulate this circuit – Schematic created using CircuitLab

1. Charge pumps and switching regulators: There are a number of circuit topologies that can produce a negative voltage given a positive supply. The simplest would be a charge pump, where a capacitor is alternately charged and discharged:

simulate this circuit

Assume, as before, that the ground symbol may be understood as earthing or the use of any reference convenient to you.

Initially, SW1 and SW3 are closed, causing the potential at A to rise above the potential at B. Then, SW1 and SW3 are opened, and then SW2 and SW4 are closed. Since A is below B, but A is now connected to ground (be it earth ground, your Arduino's GND connection, Mars ground, etc), hence B and VOUT are both below earth ground. With the addition of extra capacitors, and the use of fast semiconductor switches, this is something that is commercially realized as a practical way to produce voltages lower than your lowest supply.

With the appropriate safety precautions and appropriate components and measurement tools, you should be able to verify both of these examples and claims for yourself if you so desire.

• +1 for this answer using most basic elements as an example. – Janka Aug 19 '18 at 0:31

Yes, it is. Your problem is that you're confusing the terminal that is called GND with the big ball of dirt that is the Earth. They may or may not be connected.

In your example, if you hammer an Earth rod into the ground, then connect a wire from that to the terminal labelled +5V, then the terminal labelled GND will be at 5V below the voltage of the Earth. It will be at -5V.

The fact that it is labelled GND is irrelevant.

But if you connect both GND and the +5V terminal to Earth, then there will be a short circuit.

It may be easier to consider doing it with two 5V batteries. Each will have terminals marked + and -, and none of the terminals will be labelled GND. Run a wire from the Earth rod to the - terminal of one of the batteries. the other terminal will now be at +5V relative to the Earth. Now also connect that wire to the + terminal of the other battery. The - terminal of that second battery will be at -5V relative to the Earth.

• My question is specifically about GND being earth ground. – Jeff Wahaus Aug 18 '18 at 22:16
• @JeffWahaus In that case, you need to clarify what your question actually is. To the question in the title, the answer is definitely yes. If you have a power supply where the GND terminal is the most negative voltage, and that GND terminal is actually grounded, then that particular power supply can't produce a negative voltage relative to ground. But that says nothing about the concept of a negative voltage, or how you could go about producing a power supply to produce a negative voltage. – Simon B Aug 18 '18 at 22:26

Each independant power supply provides us a floating output voltage, as it's insulated from the source. Therefore, you cannot measure voltages between 2 different power supplies.

When we are talking about the DC, usually the GND pin (or the common, as it's most often used) is something, which isn't connected to earth's ground and has completely nothing to do with it.

Therefore, when you have a power supply, which has 2 outputs (or 3 output connectors), there are just 2 power supplies inside the same box, which are insulated one from another all the way to the output connectors, where the Vcc of second one's is connected to the common (or GND) on the output.

I hope I replied to your question.

• Well, I interpret what you're saying is that the answer is NO, there is no such thing as a negative DC voltage with reference to earth ground. – Jeff Wahaus Aug 18 '18 at 21:41
• @JeffWahaus I don't think Jakey said that at all. – Simon B Aug 18 '18 at 22:06
• I understand that, but he answered a different question than the one I asked. – Jeff Wahaus Aug 18 '18 at 22:21