# Is my understanding of electronic ground correct?

I was curious about why the pins on my Raspberry Pi didn't have a positive and negative pole, but positive and ground. I read about it on various web pages, but each article I read seemed to answer a different part of the question. I think I got it, but I'd like to have confirmation.

I will use a metaphor for electricity, as I'm not an expert. I will compare electricity to water flowing down pipes. The height it pours down will be electric potential, or voltage and the amount of water will be the current, or amperage. The product of these two will be power, or wattage. Not a perfect analogy (there's no resistance in this system) but it will have to do.

In reality, electrons flow from the negative pole to the positive pole, not unlike in my water system. So if I continue my analogy, I can call the ground floor (European ground, numbered zero) of a building "ground level," and each floor a volt. Let's start looking at what ground means.

Floor +3V will need to receive a flow of electrons or water. It will therefore need to be underground. It will be three floors beneath ground level. There will need to be a faucet at ground level and a drain (positive pole) three floors below.

And, contrariwise, floor -3V will need to send electrons or water. It will need to be above ground. It will need to be a faucet. That means that, for this floor, I will no longer need a faucet at ground level, but a drain.

So, at ground level, I will need either a faucet or a drain or both, depending on whether the water comes from above or is needed below, or both. In other words, the ground can act as a positive or a negative depending on what the other pole is.

Is that more or less correct?

• Metaphors tend not to be very correct when it comes to EE. Sep 25 at 9:42
• Also in which direction electrons actually move is extremely irrelevant to practical engineering. Let the electrons manage their own business and think of currents instead. Sep 25 at 9:55
• It's the difference of voltages that matters. Whether you measure between +12 and -12 or between +24 and 0 makes no difference for the voltage between those pins. It's 24V in both cases. We simply call one of them 0V so we can later connect all 0V together and have a shared reference.
– Mast
Sep 25 at 17:23
• @Mast This still works. Between floor -3 and floor +3, there are six floors. Sep 25 at 23:04

A major "issue" is that the question has essentially nothing to do with "ground". It relates to the flow of current between points at different voltages. The introduction of 'ground' into the conversation just adds unneded complexity.

A second issue is the use of electron flow rather than "conventional current" flow. This results in the perceive ned for current to flow from "lower levels" for a positive voltage difference to compensate for the negative charge carriers. This is an unnecessary complexity that maty confuse those seeking to answer and also may confuse the OP.

That said, the descriptions are "about right".

I was curious about why the pins on my Raspberry Pi didn't have a positive and negative pole, but positive and ground. I read about it on various web pages, but each article I read seemed to answer a different part of the question. I think I got it, but I'd like to have confirmation.

Mainly it's just a matter of convenience and custom.
"Ground" is a semi-arbitrary choice for the name of tghe negative pole in this case. The RPi DOES have postive and negative connections points. The creators have chosen to arbitrarily name the negative pole ground.

• I will use a metaphor for electricity, as I'm not an expert. I will compare electricity to water flowing down pipes. The height it pours down will be electric potential, or voltage and the amount of water will be the current, or amperage. The product of these two will be power, or wattage. Not a perfect analogy (there's no resistance in this system) but it will have to do.

• In reality, electrons flow from the negative pole to the positive pole, not unlike in my water system.

Far far far better would b to use concentional current flow (CCF). Then positive charge carriers and positive currents flow from positive poles to negatove poles. This avoids current having to eg flow uphill or against positive voltages. While this makes sense it is unnecessarily confusing.

• So if I continue my analogy, I can call the ground floor (European ground, numbered zero) of a building "ground level," and each floor a volt. Let's start looking at what ground means.

• Floor +3V will need to receive a flow of electrons or water. It will therefore need to be underground. It will be three floors beneath ground level. There will need to be a faucet at ground level and a drain (positive pole) three floors below.

As above - much cleaner with CCF - water and current flow downhill / from +ve to -ve.

• And, contrariwise, floor -3V will need to send electrons or water. It will need to be above ground. It will need to be a faucet. That means that, for this floor, I will no longer need a faucet at ground level, but a drain.

Faucets and drains are adding confusing complexity.
It MAY make logical sense in this model to have faucets sending water uphill to a drain, but it's a completely unnecessarily "stand on your head" model that is liable to be very confusing.

• So, at ground level, I will need either a faucet or a drain or both, depending on whether the water comes from above or is needed below, or both. In other words, the ground can act as a positive or a negative depending on what the other pole is.

Yes. Sort of.
At ground level - OR at any other level, a point MAY be considered to be a source or a drain depending on where current is passing from/to (or to/from).

• Is that more or less correct?

As above, sort of.
Change to conventional current, swap all ups and downs. Tidy generally.

• I consider "sort of" a success. If someone had said, "You got it exactly right" I would have been extremely suspicious. But "sort of correct," I actually can believe. Since I started looking into this, I saw many references to conventional current flow. The way my mind works, I'm still a bit uncomfortable with how it does not match physical flow, but from now on, I'll learn to ignore that. In concept, I have no problem with water running upwards. The whole thing was supposed to be an abstract representative anyway. But I get how it's unnecessarily complicated. Thanks! Oct 4 at 2:02
• @eje211 Its all based on a fluke of history. Bfore electrons surfaced they guessed at the polarity of the charge carrier. 50:50 chance, But, they got it wrong. IF they had guessed correctly it would all be ridy. BUT now we have conventional flow and electron flow. The difference is ONLY in sign convention in a few places. Negative electrons GO or positive got-it-wrongs come. And vice versa. It's far easier to use conventional flow knowing that the electroncs go the other way THAN to turn your building models upside down, as it were. You will otherwise confuse others, at least. Oct 4 at 2:10
• @eje211 The world runs on "sort of" :-) Oct 4 at 2:10

"Ground" is just a name used for a convenient reference point.
It is often at or near local physical ground potential but this is not essential.

• GROUND is just a convenient name for a reference point.

It may be arbitraryily assigned, but usually corresponds to some system feature.
Most usually it is the negative pole of the main power supply.
And very often, but not always this negative pole that we call ground is connected to a wire leading to local physical ground. But, it doesn't have to be.
Often several power supplies may share the same "ground".
eg in a typical historical PC supply "Ground" is at the connection of

• the +5V negative terminal
• the +12V negative terminal,
• the -5V POSITIVE terminal, and
• the -12V POSITIVE terminal.

Draw that and it should make sense.

BUT - It's just a name.
A convenient reference point.
It may be connected to "real" ground or not, but it has no "meaning".

Neither current in a system nor water in a water model necessarily HAVE to flow "up" or "down" to ground. Ground may be at the top (most positive), Bottom (most negative) or somewhere in between.

In the ECL logic system described below "ground" is at the "top" most positive point AND ALSO the logic used is inverted to normal so that effectively negative currents flow upwards to ground. If that sounds confusing, it is. But that's partially because the usage is arbitrary, and in the case of ECL the supply conventions AND the logic conventions have both been arbitrarily inverted compared to normal. This is valid because it's all arbitrary anyway - but most systems use the same arbitrary choices.

With modern logic families and system, ground is usually (but not always) the negative reference for the main power supply, and is often shared if there are several power supplies.
eg a PC supply of the recent past may have had a main +5V supply with it's negative pole "grounded" and also a +12V supply with negative pole grounded. BUT also a -5V and -12V supply with their positive poles grounded (which is required to make their outputs negative relative to ground.)

Modern PCs and other computer equipment may have very high current low voltage supplies (typically around 1 - 2 V range) to power processors and similar. These are often derived locally with buck converters from the eg 5V suppy, but their negative output will be connected to system ground.

Older and/or specialist logic families and devices may have employed negative main power supply rails relative to ground or multiple rails above and below ground.
An example of such a logic family is ECL (|Emitter Coupled Logic) which employed a 5.2V supply, with the positive pole grounded so that the operating supply was effectively -5.2V.
The system uses "negative logic" (high = 0, low = 1) to complete the 'unusualness'.
ECL is 50+ years old but modern balanced variants of it are still in use in niche applications.

This ECL gate

from here demonstrates a +ve ground used with a -5.2V supply.
All bypass capacitors will be returned to the +ve (ground) rail.

• It's not just a "convenient reference point" but it also can source and sink currents well, so you can put all of your bypass capacitors to ground and expect them to source or sink current spikes from ground in a manner that improves rather than exacerbates the problem. Sep 25 at 9:46
• @user107063 Yesish, but ... :-). In the context of "what does ground mean" it's 'just a reference point'. What you describe is a manifestation of that. The bypass capacitors et al sink and source to the system zero level .= the reference point. In a single supply eg 5V2 only system you could return all bypass paths to the positive rail and regulate the supply in the negative rail. You'd probably then call the +ve rail "ground" and suddenly th distinctions would vanish, apart from the signs of the potentials involved. ... Sep 25 at 9:56
• This ECL gate from here demonstrates exactly this. Sep 25 at 9:57
• Typical ECL logic levels were something lik high (1) = -0.8 V and low (0) = -1.6 V. I never saw a version that had logic 1 at a lower potential than logic 0. Of course a designer might choose to use inverted signals if it reduced the number of gates required for a design, but that was done in TTL and CMOS designs also. Sep 25 at 14:48
• @RussellMcMahon That's interesting, but it does not answer the question. Also, given the simplicity of the way I phrased things, an answer with terms like, "local physical ground potential," "usually (but not alway)," "buck converters," and other similar terms without explaining them will not be useful for me. This is certainly clear for people who already understand the issue well, but not for me. It mostly shows that the poster is very knowledgeable, but it's not actually useful at all as an answer. Sep 25 at 23:08