7

When I press a key on my keyboard, it will alter the flow of electrons in a wire connecting to send data to my computer. Let's say it sends the bits: 100000000100000000100000000100000000, which is a few letters encoded in binary and ASCII.

These electrons somehow move down the wire, to a chip, that detects where my data should go with transistors, or as I understand it, little switches that can be controlled with another wire. The chip will detect where my data is heading, and send it in the right direction.

So the question is, if I make a HTTP request from Europe to a server in USA, do some of electrons from my PC, in 200 ms the response takes, travel across the Atlantic ocean to USA and come back to me?

Anonymous Penguin
  • 3,750
  • 8
  • 35
  • 57
Markus von Broady
  • 173
  • 6
  • 1
    No. The electrons move quite slowly compared to the "bump wave" which propagates the signal. Then there are many insulated MOS gates where no part of the input is included in the output. Also your example is likely fiber optic in the middle. – Chris Stratton Sep 04 '12 at 19:22
  • This looks like an answer. To summarize this, how long do You think, an electron can actually travel? – Markus von Broady Sep 04 '12 at 19:52
  • As a rough estimate,if we take the typical distance from Europe to the USA of around 6000km and a 'typical' electron velocity of 0.1mm/s, the transit time for an electron would be around 1900 years (assuming a steady dc current)! – MikeJ-UK Sep 05 '12 at 08:32

4 Answers4

9

The signal is the movement of charge, which happens much faster than the movement of electrons.

The two are compared on Wikipedia: http://en.wikipedia.org/wiki/Speed_of_electricity From the Wikipedia article:

When a DC voltage is applied the electrons will increase in speed proportional to the strength of the electric field. These speeds are on the order of millimeters per hour. AC voltages cause no net movement; the electrons oscillate back and forth in response to the alternating electric field.

The signals used for communication are AC.

By analogy, consider the "Newton's cradle" desk toy:

enter image description here

The signal moves through the balls faster than the ball itself.

markrages
  • 19,905
  • 7
  • 59
  • 96
  • 2
    That's great analogy, I didn't know that! But the question remains, do the electrons journey? Or they go back like the balls? – Markus von Broady Sep 04 '12 at 19:29
  • How can you distinguish "movement of charge" from "movement of electrons" when electrons are charge carriers? I understand it's the signal that moves fast, but it seems like the charge itself (the electrons) are not moving fast, which makes it difficult to determine what "current" is really referring to. – Michael Feb 17 '14 at 15:31
9

When you push an electron into one end of a wire, they all jostle around a bit, and a different one pops out the other end.

In a conductor (e.g., any metal) there is a certain fraction of the electrons that are not bound to any particular nucleus and these roam freely throughout the conductor all of the time, even when there's no external voltage applied. Applying a voltage simply causes a net "drift" of this sea of electrons in the direction of the electric field. The speed of this drift is orders of magnitude slower than the speed the electric field propogates (which is basically the speed of light).

The answer to whether any particular electron might make it from one end of the wire to the other depends on the length of the wire and the signaling frequency. The drift speed divided by the signaling frequency would give you the average distance an electron might travel in one cycle. (I hesitate to call this "wavelength", but it's the same concept.) If the wire is shorter than half of this distance, some electrons are making the whole trip.

Signaling frequency is important, because in the example you give, both USB and Ethernet use AC signals at rather high frequencies. The electrons really don't move very far at all.

And to address your other point, about electrons from the input of a module appearing at its output — no, this is rarely true. Capacitors, transformers, tubes, FETs and most bipolar transistor circuit configurations prevent this. There are many places in your transatlantic trip where the direct flow of electrons is deliberately blocked, so no, "your" electrons definitely don't make it across the Atlantic.

Dave Tweed
  • 168,369
  • 17
  • 228
  • 393
6

You need to separate the concept of voltage from current.

The canonical example is a hose connected to a faucet. For simplicity, assume that the hose is already full of water, but the faucet is off, and the hose does not leak (i.e. it is aimed toward the sky).

Now assume that you just barely turn on the faucet. A given molecule of water (representing an electron for the analogy) will take quite a long time to propagate from the faucet to the end of the hose. This is like "current".

However, the water molecules at the end of the hose will almost immediately begin to fall out of the hose. Importantly, these water molecules are not the ones coming from the faucet, which are still way at the other end of the hose. You could say that an "aqua-motive force" from the faucet is pushing the water out the end of the hose. This force travels much faster than a given water molecule. This is like "voltage" (or "electro-motive force")

To recap: current represents discrete "somethings" moving past some point, and voltage represents the force which is pushing those somethings. They move at vastly different speeds; I believe that while voltage propagates through copper at roughly 2e8 meters per second (approximating for permittivity), the actual electrons move at a glacial 1 meter per hour (and that's assuming DC, AC will move even more slowly as the electrons "slosh" back and forth)

Why is this? Well, in a vacuum (or a superconductor) the electrons would fly by at about the same speed that the voltage propagates. But in real world conductors, the electrons slam into other electrons, into a nucleus, they get excited by thermal energy, etc. So they actually end up bouncing back and forth at high speeds, individually, but collectively they all move slowly in one direction.

The voltage never has this problem, it's just an electric field. That is why we use voltage and not current for signaling. The signal from across the Atlantic (ignoring fiber optics) is not carried by individual electrons, but instead is sensed as the force which pushes those electrons. That is why electronics is the art of directing voltage, not current. The wire is never "empty", it is always full of electrons; the question is whether there is a force applied or not.

ajs410
  • 8,381
  • 5
  • 35
  • 42
  • I see. The wire is like a hose, always full of electrons, the information sent is carried through electromagnetic interactions between electrons, and electron's actual movement towards the recipient is more of a side effect. 1 meter per hour is very surprising! Thanks for a professional answer. – Markus von Broady Sep 05 '12 at 07:43
  • There isn't anything magic about a vacuum that makes electrons travel at the speed of light. An electron, baseball, planet, or star can move at any speed from \$0\$ and up to, but not including, \$c\$. This is true of everything observed so far, vacuum or not. – Phil Frost Aug 15 '13 at 13:26
1

Definitely no; you overlooked something huge:

It is uncommon for your signal to travel only through wire.

There are many ways signals travel:

  • Wireless (satellite mostly): very expensive to run wire for thousands of miles under oceans
  • Fiber optic: when over long distances and speed important with lots of money this is often used, especially under the ocean. If you have money to deploy wire in water, a little upgrade isn't going to hurt, plus, since it's light, if there is a little leakage, it's not going to affect the signal.
  • Wires: often over short distances. Reliable and cheap, but not the fastest.

That being said, it is quite likely that the route between the US and Europe is either fiber optic or wireless, so it's going to be transferred into another type of energy.

The internet is composed of many different networks. It is possible to go through a wire all the way, but the traffic is directed the fastest route, often fiber optic if you can. If fiber optic is being used 100%, yes you may be able to get all wire. You may even be routed through Australia if they and Europe have a fiber optic connection and one with the US, but the US and Europe doesn't have fiber optic.

Also, some router/switch/booster/repeater may have a transistor or similar to boost the signal, "disposing" of your signal and imitating it with another one.

Even if you could get one wire, no transistor or anything, and no fiber optic or wire, when a electron is added to the wire, the farthest one pops out, almost like the first one in is the first one out, or a line.

A slightly better analogy is the milk carts at a store. When it's full and you try to shove one in the back, with enough force, the front one pops off to be processed (by the janitor) and they all shift down. If you controlled the length between each one you pushed off the shelf, you could create a code. (Kinda.) It is similar with electrons. A jug of milk is a electron, and each "slot" (space for a carton: four slots in a shelf that can hold four jugs of milk) is an atom/particle/whatnot.

Anonymous Penguin
  • 3,750
  • 8
  • 35
  • 57