New! Sharp Diodes 1W 520nm & 5W 455nm & +2W 638nm

[CDBEAM777]

Yea...I think Trinh does not have the FAC'ed diodes....SOoooo...we wait.....Drum Roll please !!!

CDBeam

 

[Alaskan]

Laser Tree sent a photo of the FAC they use, I’ve seen round fibers used like this before. They state is AR coated fiber but is 5% loss typical for such with AR coating?



Edit: I contacted Laser Tree and she fixed the shipping display so the default shown first is free shipping. Researching the loss of FAC lenses, they should produce far less than 5% loss, because of this I don't believe this seller is using the lowest loss material for FAC correction. However, since I don't see any other sellers FAC correcting these diodes, for now, I will be glad these are available, but for what I'd like to see in the future is for them to use a optimized FAC lens, although I'm not sure how much that would increase the price, maybe more than I think. Also, the D form FAC tends to be difficult to align due to the hyperbolic surface compared to a simple cylinder fiber which are very inexpensive to produce.

See specs here for a PCX FAC @ 1% loss: https://www.limo.de/fileadmin/user_upload/Downloads/Fast-Axis-Kollimator (FAC)/FAC590.pdf

Photo below, D form PCX FAC:



A simple cylinder fiber can indeed be used as a FAC micro lens, just that I wonder how easy they are to source with an AR coating. Here's an article about both PCX and simple glass cylinder fibers used for FAC: https://www.laserfocusworld.com/opt...ted-microlenses-correct-laserbeam-astigmatism

Here's the same kind of cylinder FAC on another diode, this is what I have seen in the past myself:

 

[RedCowboy]

I have heard people say that a FAC actually expands the slow axis but this never made sense to me for anything but possibly CNC use but not for producing a low divergence beam.

If the typical plano aspheric FAC acts upon the fast axis as the beam coming out of the P/N junction reaches square there by reducing the fast axis divergence and producing a square beam that continues to diverge evenly ( remaining square ) it will be diverging evenly on each axis at the original rate of the slow axis, this when collimated through say a 2E lens will produce a square spot without the long side, but rather 2 short sides, a much smaller far field spot............however if the FAC acts upon the slow axis some place after the natural square was reached and expands the slow axis then we will have a square with 2 long sides, basically the long line/bar we typically see squared.

Sure we could use wider longer FL lenses and beam expanders, but we can do that anyway, if the FAC is not correcting the fast axis by reducing it's divergence but rather expanding the slow axis divergence only in order to square the beam then we still have a very highly divergent beam only worse I would say.

NOW for say a home CNC where a wider lens ( example 30mm ) is not an issue as it could be for pocket carry, a square beam is preferable for uniform cutting in all directions, but for producing a tight stargazing beam with these MM diodes that have one very rapidly diverging axis we would want to reduce the fast axis to match the slow axis.

This is using a single plano aspheric FAC but with the fiber FAC Chris linked the article about we see it acts upon both the fast and slow axis so our far field spot after a typical coli lens could be improved, squared and while some amount longer than the normal bar's short side also shorter than the normal bars long side, this is what the advertisement could have meant by ( FAC fiber ) which is interesting and promising, fingers crossed as we wait to see.

 

[Coonie]

I can't wait to get mine. I bought the 1.2w 638 linear fac diode and picked up a flex drive from dtr. I had to get a 9mm press tool as I only have a 5.6mm one. I bought a tube cutter that redcowboy suggested I get. For smaller builds I find myself reducing the aixiz module length to fit it in and I used my rotary tool for it but it's going to be such a better cut with the tube cutter.

 

[RedCowboy]

expanding the slow axis so the spot is more square is reducing the divergence of that axis even more than it was before which was already the lower of the two

I would think that expanding the slow axis with a single lens would mean increasing it's divergence.

 

[Alaskan]

Oops, I did write it just that way, I meant collimated after expansion too; expanded like beam expander which also collimates, not just a single negative FL lens or uncollimated output.

 

[RedCowboy]

The way I see it using FAC type optics is you can either increase the divergence of the slow axis to match the fast or decrease the divergence of the fast axis to match the slow which would yield a much more desirable effect, but this FAC FIBER that you posted the info about appears to do some of both so in theory with a given coli lens....say a 2E you would get a far field spot lesser than the usual long part of the bar you would see, but larger than the usual small part of the bar that you would see.
Also how well the beam retains a square shape is to be seen.

 

[Alaskan]

CDBEAM/Bob or perhaps another member who bought one will need to show us, the ones I bought are all headed to him for a project. I just think once we make the beam have a certain shape through optics that it will remain that shape, except for the slow expansion due to inevitable divergence, but having unequal divergences for each polarity causes the ratio of size between the two to widen in the distance, unfortunately.

 

[RedCowboy]

Yes it's the aggressive divergence of the fast axis that's the big issue, that's why it makes no sense to expand/increase the divergence of the slow axis except for making a square beam that diverges rapidly all around, it would be better to reduce the divergence of the fast axis and have a square beam that uniformly diverges at the rate of the slow axis.

 

[RA_pierce]

As suggested by the name of the optic, the FAC lens reduces the divergence of the fast axis to obtain similar divergence angles in both horizontal and vertical axes.
The issue with broad emitter diodes is not necessarily that divergence of the raw output is high, but that the axes have very different divergence angles.

I have experimented with high power IR diodes in the past, with FAC. Although the divergence does not match the best gas, solid state, or even single-mode diodes we have, FAC does produce a much more symmetric beam.

I have 638 and 525 FAC diodes on order.

Addendum:
If you take a look at the values for the parallel and perpendicular divergence angles on a data sheet, the ratio of the two will give you a good idea of how "bad" the beam will be. For example, the high power 405 nm diodes have a ratio much closer to 1 than do the 638 nm diodes. The importance of this difference is apparent when you compare the width of the beam stripe after collimation with a single aspheric lens.

 

[RedCowboy]

Yes, should be able to do the math based on the data sheet, if it's accurate but at a glance the square is showing both the same which is a good sign.

 

[Alaskan]

So, if I have a handle on this now, what is happening with a FAC lens placed uber close to the diode is the fast axis alone is collimated early (or partially collimated) to reduce the amount of fast axis beam spreading so it more closely matches the slow axis output?

For now, it appears to me what we do with a cylinder pair to square up a rectangle shaped beam to more closely approximate a rounded output is the opposite of what is happening when using a FAC lens close to the diode.

 

[RedCowboy]

I wouldn't say opposite.
The FAC is a cylindrical lens, it's used to reduce the divergence of ( the fast axis only ) before a coli lens acts on the entire beam.
A c-lens pair is used to reduce the divergence of ( the fast axis only ) after a coli lens acts on the entire beam.
A c-lens pair is different in that you 1st over expand ( increase the divergence of the fast axis then reduce it with the 2nd c lens to gain advantage ( I call it optical leverage ) but that's not a technical term.

 

[Alaskan]

I'm going to be doing some reading on this for awhile, one thing though, when I think of the term divergence, I'm usually thinking of collimated divergence, not the uncollimated divergence. Maybe my thinking of it that way is causing some confusion when I write about divergence. Sure, the raw uncollimated output of the diode has a divergence; the fast and slow, just that the uncollimated output is not the kind of divergence I am normally thinking of.

 

[CDBEAM777]

OK...As I understand the operations of a Cylindrical Lens Pair...The Slow Axis is expanded/multiplied....This is why we identify C-Lenses as 2X, 3X...6X...With the Slow axis expanded....With this multiplication.... we then have a almost have X=Y expansion....Unfortunately...I have not done study of the FAC lens operations...SO...OK...I suppose....what ever optical magic we do...I am really only concerned....how close to X=Y can we achieve. How close we get translated into how square the beam is !!
We are in a race really....so...as the beam exits the diode.....we have X=Y....almost perfect....but...not completely perfect....at some point....at some distance....X does NOT equal Y....The further the distance to the Beam target....the greater X does not Y.....So what....at such great distances...Parallax view takes over....and ALL is good !!! CDBeam

Added....No matter what optics we employ...the die cavity of the LD will never deliver perfect X=Y....so....any and all of our optical manipulations will never be perfect enough to deliver a perfect X=Y at some distance....but then....we just do not need it.....So....we can endlessly discuss " How many Angles are on the Head of a Pin " Who cares...not me...Show me the "Money Beam Shot "....and I will be a happy Mad Scientist !!! HAHAHAHA

 

[Coonie]

"Money Beam Shot "

Haha