Coherent FAP800 Diode Array Divergence Question

[Alaskan]

Hello,

I've been looking at the specifications for a Coherent FAP800 laser diode array and they state 0.16 mRad (The numerical aperture of the output beam is defined as the sine of the half-angle of the divergence cone that encircles 90% of the energy.)

What I don't understand is how this kind of laser can have .16 mRad (edit: or is it mRad?), it is just a fiber output and I've seen the circle it makes on a wall a few feet away and then another 20 feet away, I see it spreading much faster than that, so how is their divergence calculated? Am I reading this wrong?

See: http://www.coherent.com/downloads/COHR_FAP800Series_DS_0715revD_2_Final.pdf

 

[doubleone44]

Isn't it Rad instead of mRad? They're talking about the sine of the halfangle. (I could be totally wrong here)

 

[RedCowboy]

They might be talking about where the 18 or so little emitters line up with the fibers. For some reason my microscope cam won't work on my new laptop, so many things are not simply reverse compatible anymore.

These pulled emitter blocks don't have the fac lens that the FAP unit with the fiber has so I'm guessing that 0.16 mRad is where it enters the row of fibers, but I really don't know, I like the big single emitters.

Did you see those 20 watt singles with power supply and zoom lens/enclosure, looks like changing 1 lens or removing a zoom stop is all it would take and 1 bigger emitter makes it all so much easier. Especially if you want to pump a crystal, but even if not.

It also says <0.16 ( NA ) So it may have nothing to do with what comes out of the fiber bundle, rather each emitter.

 

[Alaskan]

OK, well that makes more sense, it is not mRad. I'm working on the north coast of Alaska for a couple of weeks, going to sort this out more as I can test with some lenses.

 

[Cyparagon]

0.16 numerical aperture, not 0.16mRad. It's also half-angle, not full angle.
See https://en.wikipedia.org/wiki/Numerical_aperture

NA=sinθ
θ=arcsin(.16)= ... 0.16Rad half angle, which is 320mRad full angle.

Due to small angle approximation, you can get close enough by just converting NA directly to radians in most cases.

 

[RedCowboy]

0.16 numerical aperture, not 0.16mRad. It's also half-angle, not full angle.
See https://en.wikipedia.org/wiki/Numerical_aperture

NA=sinθ
θ=arcsin(.16)= ... coincidentally 0.16Rad half angle, which is 320mRad full angle.

So that's what it means. Thanks :beer:

Hmm, It says I have to spread some rep around before I can rep the guy with the answers again, anyone want a participation trophy?

 

[doubleone44]

Sooo I was close enough. Thanks for the explanation Cyp!

 

[RedCowboy]

Sooo I was close enough. Thanks for the explanation Cyp!

Yes, it's the wiki link that helps dummies like me understand it, but you had said it, I still have a lot to learn.

 

[Alaskan]

Great, thank you for the help, what awful divergence! God, how much do I need to expand the output of this FAP to get some decent divergence, I wonder. Anyone seen the divergence for the raw output of one of our laser pointer diodes in rads to compare this to?

Edit, I found an online calculator to convert radian to degrees, so from this the output is approximately an 18 degree wide full angle, that isn't too bad at all without expansion. Many of our pointer diodes have more than that for the fast axis, do you concur with this cyp? I'm not used to working with rads, as simple as it is.

http://www.convertunits.com/from/rad/to/degrees

RcowB: I gave cyp. + R, now he is lucky with 777 points! I gave reps to everyone in this thread too, you included. Thanks for the comments and help.

 

[Cyparagon]

Here's how I remember: there are 2pi radians per circle, and 360 degrees per circle. Therefore, 360°/2piRad = 1.

0.32Rad * 1 is of course 0.32Rad, but 0.32Rad * 360°/2piRad = 18.3°. Google is an easier option if it's in front of you - search "0.32 radians to degrees"

As you know, divergence is inversely proportional to beam diameter, all else equal. The beam diameter is already 0.82mm according to the datasheet. Expanding to 1.64mm would give a theoretical 160mRad.

If you wanted 1mRad, you'd need to expand to a 260mm diameter beam (over 10 inches wide, for the US viewers), so good luck with that.

 

[Alaskan]

I have some 10, 11, 12 and 14 inch diameter PCX lenses

Thank you for the clarification awesome help.

Edit: My big lenses are glass and heavy, anyone know a source for large diameter plastic PCX lenses?

 

[Alaskan]

Here's a couple of photo's demonstrating how 18 mRad of divergence is a HUGE amount:

Here's a photo of the FAP output turned down quite a bit aiming high in the sky on a rainy night and hitting a very low cloud base only 200-300 feet high, the beam had spread so much by then you can see the individual fiber outputs breaking out separately becoming individual light cones..

This photo was taken on a clear night and better shows how the individual fiber outputs begin to diverge from one another:

Camera: Kodak M1033 HD 10 MP IR infrared converted digital camera. This camera is not too difficult to convert to IR yourself by removing the back, lifting the aluminum barrier and then unscrewing the CCD censor from the black of the plastic block to take the IR blocking window out. The laser was being powered from a 25 AH Cyclon 2 VDC lead acid cell without a constant current regulator. I've tested this FAP several times for extended periods of time and with one cell, the unit does not begin to draw more current as it warms, or the wire is counter acting the effect by providing a higher voltage drop when such might occur, not sure. The amount of wire I am using provides quite a bit of voltage drop causing the laser to output fairly low power compared to what it is capable of. If I were to shorten the wire I might have a current issue, but as it is this has not been a problem running at just above lasing voltage. Note, it is never good practice to have long wires hooked up to a laser diode as inductive kick back could destroy the laser.

 

[tsingyoung]

Any idea why coherent uses fiber boundle for coupling? Patent problem?

 

[Alaskan]

No clue here, what other way would they do it?

 

[RedCowboy]

Any idea why coherent uses fiber boundle for coupling? Patent problem?

That block has 19 emitters and they get a useable burning beam, if they knife edged and stacked those it would take more room and lens pairs.

This fiber bungle is amazing.

How would you have done it?

 

[tsingyoung]

That block has 19 emitters and they get a useable burning beam, if they knife edged and stacked those it would take more room and lens pairs.

This fiber bungle is amazing.

How would you have done it?

Thanks, maybe I misunderstood the application of this bundle, I though it was originally used as pump DPSS or fiber laser? Most LD bars are using optics to couple into a single fiber (200u, 400u,etc). For a bundle of fibers the coupling of every of the 19 LDs to the fibers with uniform power is difficut，right？Or power uniformity in each fiber is not crutial?