What is ground in a physical circuit?

I'm trying to build this oscillator + modulator circuit. It works fair enough but I am having trouble interpreting it to build into a physical circuit.

In particular I am rather confused by how the ground works because in all of my previous circuits all the ground was just connected to the negative terminal of the battery. I don't understand what that means for my two battery power source.

I thought that this is just how it works and connect the negative and the positive of the opposite batteries together and connect everything else at that point.

• I would like to point out that while all the answers seem great and true in the EE sense, none is in the home design sense. In that "ground" is either a metal stake buried next to the house or the metal piping connecting to the rebar of the apartment. Some remote villages even use only a single wire for power transmission and return through "ground" instead of "neutral". Commented Jul 27 at 12:39
• Be advised that a modulated oscillator is not a good design. Much better to have an unmodulated oscillator and a modulated power amp. Commented Jul 28 at 1:38

Ground is what you define as the 0V point of a circuit.

So you simply have a +9V supply in reference to 0V GND and a -9V supply in referenve to 0V GND.

The total supply is two 9V batteries or 18V.

For the purpose of the circuits, it does not even matter what you define the ground. If you define the ground to be 1009V then that is fine, the voltages are still 1000V and 1018V but it does not change the fact that there still is two 9V batteries.

It may be easier to explain with a 18 foot stick and earth. It's a 18 ft stick. You drive it halfway into ground level so it's other end is 9ft above ground (positive) and the other is 9ft below ground (negative). But your ground level might still be 1000ft above some other reference such as sea level.

• What do you mean by absolute voltage? Is that a well-defined concept? Commented Jul 23 at 16:18
• @Nayuki I've clarified it a bit. No because voltage is always based on reference, but you are free to set the reference to anything you want, and in daily EE activities you simply talk about voltage and assuming it's always referenced to local 0V ground even if that local ground is 230VAC mains Live wire referenced to Neutral which is referenced to PE. Thanks for requiring a better, more pedantic answer. (If you are who I think, I love your website btw). Commented Jul 23 at 18:09
• Thanks for reply. I did some Googling later and found that absolute voltage might exist after all if you define 0 V to be the electrostatic potential infinitely far away. e.g. khanacademy.org/science/electrical-engineering/… , electronics.stackexchange.com/questions/639739/… . As for me, I linked my website in my StackExchange profile, so you probably got the right person. I'd be happy to converse with you by email. Commented Jul 24 at 18:34

"Ground" mainly is an abstract concept. It is the voltage node you use as a reference point for all other voltages in the circuit.

Because when we say this point/trace/pin is charged to voltage X, this is always referenced to some other point ("voltages" are actually always voltage differences).
Most of the time it makes sense to use the negative side of our power supply (0V) as the reference/ground (assuming a single positive supply). If you have a dual supply you will probably use the center point of the positive and the negative supply as the stable 0V reference/ground.

In practical circuits "ground" is mostly the common return path from all of the circuit parts.

In the physical context, all the ground symbols represent the same node. Every node which is connected to ground, using the same triangle symbol , is physically connected to every other node with that symbol. In fact, all nodes with the same label would also be physically joined, such as the $$\V+\$$ nodes in this schematic.

It makes sense to designate as "ground" (0V) the node that most of the rest of the circuit shares as a "common" reference point, of fixed potential. This de-clutters the schematic, too, by avoiding the need to draw lines everywhere, joining them together.

In the context of the algebra that engineers would use to analyse and describe this circuit, this symbol says "this node has 0V potential". Every other potential in the circuit would have some value quoted relative to this zero reference point.

For instance, in your circuit, the ends of the batteries not connected to ground are labelled $$\V_+\$$ and $$\V_-\$$. They would be 9V higher in potential than the ground node, and 9V lower, respectively. I italicise "higher" and "lower" to emphasise how their potentials are actually just relative to ground. However, since that ground is 0V, we would quote their "absolute" values as: $$V_+= 0V + (+9V) = +9V$$ and $$V_-=0V + (-9V) = -9V$$

It's arbitrary which node you choose to call "zero volts", but it makes sense to choose a point which facilitates the algebra and the understanding of the circuit's function. For circuits with op-amps, it's often useful to put that "zero" mid-way between the extremes of the supply voltages (+9V and -9V here), and have the op-amp's input and output signals oscillate about that zero "mid-point".

To relate this to the physical construction of your circuit, if I redraw a small portion (the top-left part) of your circuit, using no ground symbols or other node annotations, it would be like this:

simulate this circuit – Schematic created using CircuitLab

That's a good representation of the physical wiring of the thing, but imagine trying to draw all the other parts as well, being forced to join all the ground nodes, and power nodes, with lines. Nightmare.

To summarise, all nodes sharing the same symbol or label (the ground symbol in this case) are physically joined. The schematic is less cluttered using this convention. All nodes marked as ground have the same potential, because they're all joined. That potential is "zero volts" because it's convenient, both algebraically and descriptively.

The terminology commonly used in electronics is a bit unfortunate* because (at least in my opinion) it makes this concept harder that it needs to be. An alternative term for "ground" is "common", which may help your understanding.

Voltage is a relative measurement. You don't measure the voltage of a point, you measure the voltage between two points. When you draw a circuit, you select some point to be the "common" node and then make all of your voltage measurements relative to that point (i.e., all of your voltage measurements have that point in common). It's not a physical entity with any special properties**, it's merely a label that you use to say "this point is defined to be 0V and all measurements will be taken from here".

It doesn't matter which node in the circuit you choose to be the common node. In many cases, the common node is selected such that no voltages in the circuit would measure negative values as this makes the circuit a bit easier to work with (some equipment, particularly older analog voltmeters, don't handle negative voltages very well). Your circuit has the common node between two batteries. That's a perfectly valid arrangement as well, and is fairly common when opamps are involved. The circuit works the same regardless of which node you label as "common", so choose whichever node makes the most sense based on your particular circuit and use case. Think of the common node just like you would your V+ and V- nodes. It's a way of indicating that everything connected here has the same voltage (in this case, a fixed 0V).

*Electricians have better terminology here IMO. They use the term "common" for this, and the term "ground" refers to an earth ground (a wire directly connected into the dirt). It's extremely common for circuit designers to use an earth ground as the common node, which has lead the electronics world to use the two terms interchangeably. This often causes confusion when the common node is defined to be something different.

**The common node is sometimes constructed differently than other nodes in physical circuit boards. This isn't because it's special, it's simply because the common node is typically used more often than any other (even your simple circuit has 10 different references to it). It needs to be designed slightly differently to accommodate the large number of widely-spread connections and the amount of energy flowing through it. Any other node with a large number of connections (like a V+ rail) will likely have similar requirements. This is also why the common node gets a special symbol on schematics. It's used so often, a schematic would turn into an indecipherable jumble if we drew it like any other node.