Transmitting Information Faster than C

[doG]

But you don't have to speculate. This is science, and there are measurements we can make. The orbital decay of binary pulsars has been observed, and is consistent with gravity travelling at c, as well as other experiments.

That doesn't sound like a measurement of the speed of gravity. I really don't see how you could measure it's speed, as opposed to the force, without being able to suddenly make a mass appear or disappear. If the sun went poof we wouldn't see it for about 8 minutes. How long would it be before we felt it?

[timo]

If a charge a distance d away were to suddenly cease to exist, why wouldn't it take you d/c to find out? If you start the charge oscillating, you don't see a magnetic field instantly, you have to wait d/c for it to arrive.

See the beginning of my previous post #50 for an explanation of why I think the change might be instantaneous. In the case of a charge starting to oscillate around there´s two major differences between the scenario I descibed in #50, namely that 1) there is no violation of any basic law of physics and 2) that you cannot apply electrostatics to a moving charge.

[swansont]

See the beginning of my previous post #50 for an explanation of why I think the change might be instantaneous. In the case of a charge starting to oscillate around there´s two major differences between the scenario I descibed in #50, namely that 1) there is no violation of any basic law of physics and 2) that you cannot apply electrostatics to a moving charge.

 

No, you couldn't use electrostatics, you'd have to use Maxwell's equations, which are the more general form. But charge conservation isn't part of that; as long as the charge change isn't instantaneous (which would give you a nasty infinity) I think you could solve it. Or do the next best thing: start with a dipole and then quickly have the charges meet and cancel. No violations at all.

[timo]

No, you couldn't use electrostatics, you'd have to use Maxwell's equations, which are the more general form.

I am not convinced that electrostatics shouldn´t suffice, but either way: Where exactly would you see a problem in post #50? I don´t think you doubt that the field before the vanishing of the charge is ~q/r². So as far as I see there´s only two possibilities left for being errornous:

1) Gauss Law does not apply or does not apply the way I sketched that I want to use it. If you see any error there, please tell me what exactly the problem is.

2) Related to 1) but in my eyes slightly different: The assumption that the charge-free space is rotational symmetric around at least the former position of the charge. Now that would really suprise me if that wasn´t true.

 

But charge conservation isn't part of that; as long as the charge change isn't instantaneous (which would give you a nasty infinity) I think you could solve it.

My approach in #50 remains the same if you let the charge drop to zero in a finite time interval with any function (linearly, for example), as long as that time interval is sufficiently small (smaller than 8 minutes in this case). So any caclulation footing on a certain q(t) function are very welcome for analyzing what went wrong.

 

Or do the next best thing: start with a dipole and then quickly have the charges meet and cancel. No violations at all.

True. And it also is a completely different scenario for several reasons, the most important being that my assumptions in #50 don´t hold true there. Neither the use of electrostatics for the q/r² nor the spherical symmetry of the system.

[swansont]

The problem is that electrostatics assumes no time changes, by it's very nature. It's static, and eliminating a charge is decidedly non-static. You can apply Gauss's law before and after, when you have steady-state, but not during the interval in the region (dictated by r=ct) during which you have a transient.