How fast is electricity?

[ShawnD]

I was talking to a friend and he said "wireless internet should have a lower ping because it uses radio and nothing goes faster than light". While I agree with his statement about light being pretty fast, could it be possible for electricity to be even faster than light, in a non-conventional sense? Read on, I'll explain.

Electricity flowing through a wire is not similar to a bullet traveling down the barrel of a gun. It's more like those http://www.angelfire.com/home/sesquiq/joe.html you see in gift shops. 5 balls are hanging from the frame, all touching. You let one ball swing and hit the other 4, and the only ball that moves is the one at the end. Electricity is the same deal. When you apply a voltage across a wire, the electrons added to the wire on the negative side do not need to travel the entire length of the wire before lighting a bulb on the other side.

With this concept in mind, what if the cable was 1 light-year in distance? Would the wire have a measurable amperage in less than 1 year after pulsing a voltage? If so, wouldn't this show that electricity is faster than light, even though no single electron is traveling that speed?

edit: maybe an example would help
Suppose the ultimate speed limit of the universe was 10 per second (units don't matter).
I have a tube of length 20.
Inside this tube is a series of elastic balls which fill up a distance of 15.
If I hit the balls on one end, the kinetic energy moves to the last ball which starts to move at a speed of 5, and travels a distance of 5.
The overall time from when I hit the balls to when the last ball reaches the end of the tube is 1 second.
The speed of the actual ball was 5 per second, speed limit of the universe is 10 per second, the speed of the signal I just sent is 20 per second (20/1 = 20).

 

[disregardthat]

I read this question in a science magazine, only here they asked if information could be sent faster than light through a dense solid tube. If you put a long tube up to the moon and knocked at the bottom, would the vibration reach the moon faster than c? The answer was no, because the atoms did not react to each other as fast. And of course they quoted the sentence we all hear that nothing exceeds the speed of c...

I don't believe that this is very different, at least not in concept. An electron would not react fast enough to a neighbour electron bashing into it to exceed the speed of light. Well, this is only what i think.

 

[ShawnD]

I guess that makes sense. Does this mean wireless has lower pings than wired?

 

[disregardthat]

After your edit in the first post: The equations to calculate the time the balls use to react to each other would drastically increase as the speed of light decreases. So the same principle holds.

 

[DaveC426913]

Re: wireless vs. wired router: The actual travel time of the signal from source to destination is utterly negligible compared the electronic hanky-panky that's involved in manipulating the signal before it goes out and after its received.

 

[disregardthat]

I thought the wireless internet traded information with electromagnetic waves, and not electrons.

 

[cesiumfrog]

"wireless internet should have a lower ping because it uses radio and nothing goes faster than light"

I tried parsing that a few times, the best I can get is: you and your friend think EM radiation is not a "thing" and therefore is not limited to traveling at.. the speed of EM radiation?

the kinetic energy moves to the last ball

When does it do so?

 

[disregardthat]

Uhm, who can deny that EM doesn't move at the speed of EM? :S

 

[cesiumfrog]

oh, so it's with reference to wired internet (rather than a complaint about wireless). ne'ermind then

 

[chroot]

Oh, wow. There are so many misconceptions in this thread that it's hard to know where to begin..

It's more like those balls on strings you see in gift shops. 5 balls are hanging from the frame, all touching. You let one ball swing and hit the other 4, and the only ball that moves is the one at the end. Electricity is the same deal.

For starters, this is absolutely, 100% wrong. If an electric field is applied between the ends of a conductor, the electrons in that conductor have a very slight motivation to move in the direction of the field. The electrons' motion is still actually dominated by thermal energy. The free electrons (i.e. those in the conduction band) act very much like a free electron gas. Their mean thermal velocities are something like a 100,000 meters per second or so, and the motion is random.

They do not advance from one end of the wire to the other at 100,000 meters per second. Instead, the drift only very slowly from one end of the wire to the other. In common situations, the drift is only some tens of centimeters per hour.

So, now I must deal with the misconception that the electric field travels instantaneously. It does not. Disturbances in the electric field travel at roughly the speed of light. Disturbances travel more slowly through materials than they do in free space. For example, changes in the electric field propagate through interior traces of FR4 fiberglass printed circuit boards at about six inches per nanosecond, which is only about half the speed of light. The dielectric material surrounding the trace is what slows the propagation down, because it acts against changes in the magnetic field outside the wire, which must necessarily accompany any changes in the electric field inside the wire.

Signals (and information) cannot propagate through wires at c, much less faster than c. It is possible for a signal's phase velocity to exceed c, but not its group velocity -- but this is only a minor quibble, and information can only propagate at the group velocity.

I should also reiterate Dave's point, which is that the time that a packet spends in a router's memory far exceeds the time it takes to propagate from one router to another. Wired networks are not, in any meaningful sense, faster than wired networks.

- Warren

 

[rbj]

i was beat to the punch by Warren, but i'll say two things: the speed of propagation ("wavespeed" or "group velocity") of a signal down a pair of wires (otherwise known as a "transmission line") is often about half the speed of light in a vacuum. but the speed of the electrons ("drift velocity") in such a conductor is far slower (which means that we don't have to consider the relativistic effects on electron mass). these two speeds need to be differentiated from each other.

 

[disregardthat]

Could you please explain phase velocity or give me a link to the explanation? which is understandable, Wikipedia tend to tackle me with words.

 

[DaveC426913]

All that being said, wired networks tend to be more reliable in terms of connection and thus throughput, the upshot of this is that, in general, a wired network will allow you to download more data in the same length of time than a wireless.

One could say this is faster, in the same sense that, of two cars that drive across the country at 60mph, the one that doesn't have to stop for gas or repairs will get there faster.

 

[ranger]

I've heard many explanations on how electrons "flow" and give us current. I read many of them on PF and even heard them from upper division students at college. I've definitely heard ShawnD's explanations many times (which makes sense) and also chroot's [among others]. Now I'm confused. Which one do I follow? Because it seems that everyone has something different to say.

 

[Gokul43201]

I've definitely heard ShawnD's explanations many times (which makes sense) and also chroot's [among others]. Now I'm confused. Which one do I follow?

Shawn's opinion, while it might make sense to you, is most definitely incorrect (nothing personal here).

Pick up any text on electromagnetism (e.g., Griffith's or Jackson) and you will see that Chroot's explanation here is the correct one, and there's ABSOLUTELY no doubt about this.

If you're reading incorrect explanations "many times" then you should perhaps take more care with how you choose your sources of information.

 

[rbj]

Could you please explain phase velocity or give me a link to the explanation? which is understandable, Wikipedia tend to tackle me with words.

phase velocity or group velocity? they're not quite the same (except in a phase linear system).

do you want it with math or no math? (dunno how to do the latter.)

if the former, do you know how to do basic AC circuits with sinusoids and complex exponentials?

this is very closely related to the concepts of "phase delay" and "group delay" in LTI (linear, time-invariant) system theory. are you comfortable with that stuff?

 

[Mindscrape]

Er, what does phase velocity have to do with current? A phaser is different from phase velocity. Am I missing something in the conversation?

 

[russ_watters]

I didn't see this mentioned, but it is important: if the cross-country communication lines were traditional wires, that would have an impact on ping. Since they are fiber optic, the ~3000 miles is traversed in about 0.015 seconds. That is a noteworth fraction of your ping, but nothing compared to if the backbone was a regular wire.

 

[silver-rose]

I used to game online, and I based on experience, no way does WAN beat LAN.

Maybe you guys need to take into aco**** this point:

Electrons in a conducting medium travel at their fermi velocity, which is about 10^7 ms-1. However, due to the prescence of other electrons and protrons etc... their free mean path is so small that they continously collide. This leads to the "overall" speed of electron movement through the conducting medium, which we all know is the "drift velocity".

 

[ZapperZ]

Electrons in a conducting medium travel at their fermi velocity, which is about 10^7 ms-1. However, due to the prescence of other electrons and protrons etc... their free mean path is so small that they continously collide. This leads to the "overall" speed of electron movement through the conducting medium, which we all know is the "drift velocity".

However, the drift velocity isn't the "speed of electricity". If it is, we would be able to detect such a delay in many situations where our cables are significantly longer than 500 meters.

Note that in many experiments in high-energy physics, the timing delay in electrical cables and electronic components is a significant issue. This is true when you are timing something of the order of picoseconds and femtoseconds. I work with something to which the time resolution must be accurate on the order of 2 to 3 picoseconds, and the length of our cables from the accelerator bunker to our control room become a significant factor to get our timing right. At this scale, estimating it to be close to c is often good enough.

Zz.

 

[silver-rose]

ah so Zapper, is it fair to conclude that electricity travels as fast as the electric field is distributed across the conducting medium?

 

[rbj]

ah so Zapper, is it fair to conclude that electricity travels as fast as the electric field is distributed across the conducting medium?

i'm not Zapper, but i might say that your hypothesis is correct if you substitute the word "propagated" in for "distributed".

 

[disregardthat]

phase velocity or group velocity? they're not quite the same (except in a phase linear system).

do you want it with math or no math? (dunno how to do the latter.)

if the former, do you know how to do basic AC circuits with sinusoids and complex exponentials?

this is very closely related to the concepts of "phase delay" and "group delay" in LTI (linear, time-invariant) system theory. are you comfortable with that stuff?

I have no eduactaion in math in physics. So i wouldn't understand it. This forum should have a definition base.

 

[ZapperZ]

ah so Zapper, is it fair to conclude that electricity travels as fast as the electric field is distributed across the conducting medium?

Not quite, because the metal has a "dielectric constant".

One of the things that people have missed also that even in vacuum, I can also make the speed of any EM wave slow down. This is one of the things my group also does, because we send both traveling wave and standing wave into a dielectric-loaded structure. Due to the geometry of the structure AND the boundary conditions at the dielectric surface (basically a hollow dielectric tube), the group velocity is slowed down below c even when the wave is traveling in vacuum (better than 10^-10 Torr pressure). We need that because since we are using those structures as accelerating structures, an EM wave progating at c will be move too fast to effectively accelerate electrons that are not quite at c (even though after 3 or 4 MeV, it is close enough to c). So those electrons will start lagging behind the accelerating wave and this will eventually defeat the purpose of having an "accelerating structure".

However, for most applications, and certainly in our case with picoseconds time resolution, estimating the signal in the wire to travel at c is sufficiently accurate.

Zz.