Why are diode lasers different?

[Videogamer555]

To collimate such a beam properly, you need a lens that first corrects the astigmatism, and then collimates the resulting cirular beam. So you need at least 2 optical components, right?

 

[ElektroFreak]

Correct. The most accepted way to do that is to use an anamorphic prism pair to expand the greater diverging axis only, which drops it's divergence way down since divergence is inversely proportional to initial beam diameter. Then once you've got the profile more-or-less symmetrical you use an ordinary convex lens to collimate the beam as a whole. The order of elements can also be reversed.

Typically, though, this method is used with multimode diodes rather than singlemode, since even though singlemode diodes produce an oval profile they are still easily collimated very well with a simple lens.

 

[BShanahan14rulz]

Plus, for a laser pointer, who cares about beam profile as long as it's capable of pointing ;-)

Other applications where specifications like beam diameter, shape, and gaussian profile are important to the end result mandate further beam correction

 

[Videogamer555]

Plus, for a laser pointer, who cares about beam profile as long as it's capable of pointing ;-)

Other applications where specifications like beam diameter, shape, and gaussian profile are important to the end result mandate further beam correction

Actually you do care, because an astigmatic beam can be collimated in one axis, but not the other, and therefore will spread into a line. You have to change the divergence in one or both axes so that they are diverging at exactly the same rate (which will make the beam circular) before using a regular convex lens to collimate the beam.

Also possible would be to use 2 crossed cylindrical lenses. One would collimate it in one direction and the other would collimate it in the other direction. Or you could use one cylindrical lens to correct the divergence of one axis so as to make the divergence the same in both axes, and then use a normal convex lens to collimate the beam.


However I often hear that "3 element" collimation optics are often used for a high quality beam output. I can't imagine any configuration that uses 3 lenses. Why would you ever need 2 use more than 2 lenses? What would the 3rd lens accomplish?

 

[BShanahan14rulz]

Don't tell me what I do and don't care about!

What I meant was that in a laser pointer application, a line output is "good 'nuff." Shouldn't be using a beefy high power multimode in a pointer anyways

There are many many ways to correct astigmatism. Some ways are much more cost effective than others. If every single diode had the same divergence ratio, we could just buy 10,000 of one lens, 10,000 of the other, and get a decent discount on them. But I'm not gonna be spending hundreds and hundreds of dollars to correct the beam of my $5 diode

As for the three element aixiz lenses, I don't know, all I can do is guess. They are so small I can barely even tell the difference between meniscus and planoasphere. My guess is that it is kind of like what they do with green lasers, except they use two matched opposites after a collimator, but it doesn't seem too logical in my mind. This paragraph needs a smiley at the end too, to match the rest, I reckon... Always kinda liked this one... :knight:

edit: found a diagram of what I am assuming is an aixiz
http://laserpointerforums.com/f65/new-9mm-445-diodes-73284-13.html#post1083332

 

[ElektroFreak]

^Lol.. It's definitely true that this is mostly a matter of personal preference in a pointer application, but generally speaking it's VERY easy to get an almost perfectly parallel beam (in both axes) from singlemode diodes. In fact, we routinely see divergence figures that are well below 1mrad from singlemode diodes using only a single collimating lens.

It's the multimode diodes that produce a line output, and the divergence can be quite high when using simple lenses. With these, most people prefer some fast-axis correction, but it's still a matter of personal preference and suitability for a particular application. For example, in my show projectors I'm using the current generation multimode 635nm red and 450nm blue diodes. Because of the longer wavelength, the red diodes have considerably higher fast-axis divergence than the blue diodes. Both are multimode, both produce a line output, but I only correct the fast-axis on the reds. The blues are good enough for my application as they are.