Re: Novice question, conservation of energy

Neil Bates wrote:
"Phil Hobbs" <pcdh@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message news:46C92699.4090800@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Neil Bates wrote:
"Phil Hobbs" <pcdh@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
The impossibilities start when you're talking about thermal light. In that case the polarization trick only works once, and you lose half the light in each beam making it polarized in the first place. We talked about this just recently.

Hi, I suppose by "recently" you mean the thread, "A question about the phase difference of the beam splitter"?
No, I meant earlier discussions about the second law of thermodynamics and the conservation of radiance. I haven't been following the beamsplitter thread.

Cheers,

Phil Hobbs

OK, so what do you think about the problem of using symmetrical combiners, and any transmitted-reflected phase difference other than 90° causing problems? I mean, adding the amplitudes naively and squaring gives the wrong total power output (no conservation of radiance.) To conserve energy, we have to imagine an adjustment - for example, impedance in the half-reflecting film just forces the output to be the input rate. Otherwise, we would naively be adding two sets of around 0.707 times entering amplitudes, for an output total of twice the input power. Or with entering phases 180° from that, the output would "naively" be zero. Do you think that's how it works?

Regards,

Neil Bates

http://tyrannogenius.blogspot.com



I wouldn't come at it that way. What I like to do is to stay close to the physics, which in this case means the math (i.e. the Fresnel formulae). That way I get less confused and waste less time. I'm also much more confident in the first and second laws of thermodynamics than I am in any sort of arm waving, mine or anyone else's. Overturning either of those would require very strong evidence indeed.

The constraints on the phases of reflection and transmission at boundaries come from patching conditions--from Maxwell's equations, one can easily show that tangential E and H, and perpendicular D and B, are continuous at the boundary, and that's where the Fresnel formulae come from. That being the case, unless you find some violation of the Fresnel formulae, there's no room whatsoever for any process that would violate energy conservation.

Cheers,

Phil Hobbs
.