buffo
The idea is that if the beam reflects off the aperture window back to the galvos, the red component of the beam could go all the way back up the chain into your red diode(s), causing instant diode failure due to back-reflection. (This assumes that a significant fraction of the beam power would be able to make it all the way back to the diode, and that the diode is already operating at close to it's maximum output.)
In practice, this is exceedingly rare. For one, most diodes are not running right at their failure point. And second, the amount of reflected energy (even off non-A/R coated glass) is usually less than 10%. So it's not as big of a risk as you might initially assume.
Also, tilting the glass forward at an angle like that shown above will only help you if you *never* scan below the horizontal. Otherwise, if you scan lower, eventually you'll hit the point where the beam angle (relative to the horizontal) is the same as the tilt angle of the glass (relative to the vertical), and you have the same back-reflection problem.
In fairness, with a slightly tilted window, the point where the beam can be reflected directly back to the scanner mirrors will be below the horizontal, which means that even if you did scan below the horizontal, in theory you'd only ever hit that reflection "sweet spot" while scanning. But this doesn't offer you any real protection, since facet damage due to back-reflection will kill a diode in a manner of a few nano-seconds. So even if the beam is moving and you only hit the sweet spot for a brief moment, it doesn't really matter.
On the other hand, with a window mounted straight up (vertical), it's possible that if your galvos were *perfectly* centered in their mounts, then the normal beam output position with the galvos at rest (no scanning) would be exactly at the back-reflection point, so if you were warming the lasers manually with no scanning going on, the chances of hitting that "sweet spot" for back-reflection with maximum laser power would be higher. But again, this is *very* rare...
Personally, I mounted my A/R window flush against the aperture and never looked back. Also, my A/R window has *horrible* coatings. We're talking reflection rates of 6% to 9%! So I did have a fairly decent back-reflection risk in my projector. And even so, I never killed a diode. (My projector has a pair of the original Maxyz modules, which are most definitely running right at their output limit. Yet they're still kicking just as strong after over 3 years of service.)
Recently I removed my aperture window because I was tired of losing so much power from the shitty window (which cost me $60 from OneStopLaserShop.com). Been thinking about trying some of the Edmund's A/R-coated class linked to above though...
Adam
Reply With Quote
Originally Posted by blowfly
has many good points, and certainly true - though I always did it more for the chance that, in the course of scanning, you'd get some 'retro' (rather than 'back') reflection off the glass, onto your Y galvo, and back out the window - also, *really* rare - but, if you do a lot of 'graphics' shows (which I / we do for Corp, gigs, etc) the last thing you want is *any* chance of 'ghosting' - the angle eliminates that possibility, since all of even the 1% reflected-back (in the case of 'CDAR' there, ~99% transmissive, or-better, wavelength-dependent....) would get dumped into the 'ceiling' of your projector...
PM-is cool... 
....and armed only with his trusty 21 Zorgawatt KTiOPO4...
the 'framers' are usually artists or art-sensitive, so they'll handle with more care, in my experience... 
, someone in the Netherlands who can deliver a piece of at least 6cm x 6cm and how to frame it in the projector? glue it on the inside? then i need a piece of 7cm x 7cm or so.
