445nm diode with beam correction optics (anamorphic prism pair)

[mccarrot]

Is there any particular reason why cylindrical lenses should work better than a prism pair?
I read that a few times already but I would be interested in knowing a little more background.

Prism pairs gives a much higher loss compared to cylindrical lenses.

[RedlumX]

Prism pairs gives a much higher loss compared to cylindrical lenses.

How come? In principle there shouldn't be any difference. Perhaps in practice because practically all anamorphic prisms are made of SF11 (*) which has high loss below 450nm (thanks Andy who pointed this out). That's why I am looking into custom ones out made of BK7 plus coatings. A group buy seems worthwhile but I am not tempted to get into this mess, perhaps somebody else?

Note added: I just played with prisms coated for 532nm, and at 532nm they have 3% loss, but like 30% loss at 445nm.

Note added2: (*) SF11 because of the high index of refraction. Recall that the magnification is given by M=2n (where n=index of refraction), when one chooses incidence under the Brewster angle. For SF11, n=1.8 and M=3.6 roughly. For other angles (ie, different magnifications), the sloped sides need to be coated for the wavelength in question, in order to avoid considerable loss. This means that for a magnification suitable for the blue diodes, which is like 4 or 5, in connection with low-loss glass like BK7, one necessarily needs proper coatings.

[Xytrell]

With a prism pair, it's two optics. With a cylindrical lens, it's one optic. Two optics generally have higher losses than one optic

[mccarrot]

With a prism pair, it's two optics. With a cylindrical lens, it's one optic. Two optics generally have higher losses than one optic

you need 2 cylindrical lenses, you use them like telescope.

to make a prism setup you have to shoot the beam at a angle at the prism, this will give you bigger loss, and the glass of a prism pair is thicker.

[shrad]

glass itself should not be a loss, for what I know it's the interface between two optical mediums which gives losses

[Laserman532]

glass itself should not be a loss, for what I know it's the interface between two optical mediums which gives losses

you can expect a 4% per surface loss on all un coated glass at normal incidence on average across the visible spectrum (most people know, but i wanted to put the number there)

[RedlumX]

glass itself should not be a loss, for what I know it's the interface between two optical mediums which gives losses

Right, so it is a matter of the coating, be it prisms or cylinder lenses, and in each case you have two times two interfaces.

Well everyone must come to his/her conclusions about what to do. Good luck setting up a crossed pair of cylinder lenses ;-)

[shrad]

I find mounted prism pairs to be much easier, as no alignment is needed expect putting the pair right in front of the beam

[ElektroFreak]

Some losses can be minimized by ensuring the first prism face is at the brewster angle for the wavelength in question.. You can find this by trial and error if you're using pre-mounted prisms.

[gogu]

Nice setup shrad,
i dont have one of these 445 diodes, so my observation of whats happening might not be relevant.
But what i've noticed when placing a 473nm dpss (closest wavelength i have to 445) through the prism pair is loss due to reflection caused by the angles involved. not necessarily losses through the glass itself.
getting away from these angles is impossible, someone might be able to tweak them for best performance, but there will always be losses due to the angles of the prisms.