Detector Watch S6 V1

Burnt out elements that have cooked off with the constant heating from the laser. It's a semiconductor lattice. It counts photon fluxes, doing so by catching each one of them and causing movement of an electron(s) - hence a current is generated. Much of the photon energies become lattice vibrations ie. heat. If that occurs in specific areas then spot temperatures will go up and wrecks the local lattice structure. So as time goes on the device as a whole converts a lesser fraction of the incoming photons and it's function degrades.

Cheers, Mike.

This entry from Hanford is worth copying entirely :

Note that one frequency is twice the other ie. harmonic. VME is a type/brand of mainframe operating system, which I believe is particularly appropriate for realtime use. One wonders what particular behaviour of that board produces such a specific frequency. That is, what is happening some 54.496 times per second? Task switching? Hardware interrupt? Something evidently worth doing over and over .....

Electromagnetism is a long range force or, if you like, is hard/expensive to shield a component completely. There are Faraday cages which are conducting shells around some volume of interest. Any incoming signal, up to some maximum level determined by the geometry of the cage, will be masked. The electrons in the conducting surround dynamically arrange themselves to form an 'mirror image' of the externally applied field ( a perfect conductor has no electric field component tangential to it's surface ). Thus it soaks up the energy of the incoming radiation, in practice almost perfectly. The film 'Enemy of The State' had Gene Hackman's character operating his electronic snooping within one.

For their operators Livingston has a new graphical interface for the Thermal Compensation System ( keeps the mirrors in good shape ) :

I can imagine myself playing with that! Fire up the lasers, use the slider to select power, choose a mask to heat the centre or the periphery, aim it and cross the fingers. :-)

Cheers, Mike.

( edit ) I've had a glance at the specs for the photodiodes for AdLIGO. They're quite elaborate beasts compared to when I first studied such things. Predominantly Gallium-Arsenide but with Indium, Tin, Aluminium and Nitride in the mix. Not with PN junctions but P-I-N junctions, about 8 layers all up. Gold and/or Platinum contacts. I'll have to look that up.

( edit ) Ah, I is for intrinsic ( semiconductor ), lightly doped equal numbers of n's and p's. Makes a PN junction less of a cliff and more like a ramp. The junction width is wider and thus has a lower capacitance on reverse bias, ie. faster. A photon hitting the I region frees up a carrier -> current. Small dark current. Here's a nifty energy band applet.

Here's another good example of tracking down noise, from Livingston :

BNC is a type of coaxial cable. From your aerial to the TV set, say. ( One can think of it like a loop/circuit of wire but with the return leg wire coming back inside the outward bound one - so it has some characteristics of a Faraday cage. Twisted pair is a cheaper attempt to emulate this ). I call it the fiddly type ( but aren't they all? ) as one needs to be quite careful when fashioning terminations. If the shielding mesh isn't continued carefully to the connector then stuff will leak in - also inadvertent grounding or floating. In this case the 60Hz influence would leak into the circuitry that controls the laser that shapes the end mirror on the X arm.

I recall not having a proper understanding of BNC termination with an office network some years ago. To reduce the performance depending upon the wiring of the building and other factors, then in addition to an impedance load at a T-piece on one end of the line of computers ( to deal with signal reflection ) one has to ground to the case on the other end. Computer power supplies only give you a virtual/AC ground for the network card outlets via the mobo. So a DC offset can arise - you may detect this when you touch around the T-piece! :-)

Cheers, Mike.