Hacker News new | ask | show | jobs
by ISL 4712 days ago
The short answer is yes.

Aside from tiny losses like tunnelling, scattering from vacuum excitations and thermally-generated fields, etc., a perfect optical cavity will store light indefinitely. No laser required, only light of the appropriate wavelength and phase.

The combination of Fabry-Perot cavities and power-recycling in the optics of the LIGO gravitational wave observatory "gains" a ~100 W laser up to a few megawatts. If a cavity is formed in free space, like LIGO, then careful attention to vacuum is required to prevent loss.

ULE clock reference cavities have quality factors that are much higher, and need no vacuum, as the light propagates entirely within a glass substrate.

Edit: I should add that the "tiny losses" mentioned at the outset are precisely what prevent you from making a "perfect cavity". As the quality of a cavity/oscillator increases, the number and deviousness of loss mechanisms does too. This is especially the case at frequencies that are low compared to those at which an experimenter can iterate.

For HN, consider building a host that can run uninterrupted for 10^12 seconds (30,000 years).

1 comments

...but the effective storage time of the power recycling cavity in LIGO is about 1 second. This is a rather long time by optical standards, but I'd say it's a long way from "indefinitely".
Agreed. My post may have been more tautological than I'd hoped (in the absence of loss, a cavity is lossless). Thank you for the important reminder that in nature, there is no DC.

In the sense in which I'd understood the question "Cavity losses are often dominated by reflectivity losses at the mirrors: does an improved mirror make a big difference?", I think my reply is relevant.