Learning about lasers over fiber

[VDX]

... the image on the cup was with 808nm -- the 200W-diode with the 35x-beams has 975nm.

And yes, the "stacking" of several collimated squares/lines to a quadrate is made by knife edging -- here you can see the 7 diode-blocks (with 5 single emitters each) precollimated to a "line", which the 45deg-mirrors (which are slightly scewed) "stack" into one combined beam on the most left mirror ... here it's scattered up, then (not visible) another mirror will fold it horizontally, where it's last focussed into the fiber ...

Viktor

[masterpj]

I've been doing some research overnight and know what I want.
400um to 800um fiber with SMA 905 female connector on both ends.

Found a good collimator that's adjustable.

things I haven't figured out:

- Is it necessary to get a glass fiber over a plastic fiber for a visible wavelength of around 1 watt? (or are the losses too large using plastic?)
- How would I adjust the distance of a fiber when attaching a random fiber with a pre attached sma 905 connector on it? I have for example a necsel here that has a pre attached fiber attachment (sma 905 thread) Are you able to push the fiber up and down somehow to get it into the focal point or do you have to twist the thread to distance the fiber ?

[VDX]

... I had to adjust the SMA connector or the lens in respect to the fiber-end -- have some SMA-couplings out from medical lasers, so no problem

With DIY-fibers and diameters of 0,4mm I've got up to 50 Watts of 975nm into the fiber -- but regularly burnt the output-end when near to 100 Watts

Viktor

[masterpj]

gotcha! So what im looking for is basically a 400 micron fiber with sma 905 connector on both ends (both female ends)
That will do the job for me.
Plastic or none plastic shouldn't matter too much I reckon because im not working with IR but a visible wavelength around the visible diode range around 1 to 3 watts of power.

[mixedgas]

This is awsome! Really admire this kind of stuff. I really want to learn this to perhaps learn to tame a necsel or sorts using a fiber to bring the array together.. I noticed with those 808nm (i assume they are 808nm) arrays that you are using what appears to be regullar mirrors or even dichroics where you shine one beam off one side but another beam passes through as well rather then going past it (which I think is maybe the drawing not being accurate before? I assume its all knife edging here then somehow bringing it down to a fiber end?.. (I never really delved into fiber so this is a new area for me).

~
If your projecting graphics for laser show, you need 65 to 200 uM core max. 400 is for Machining. 800 is for flood lights.
Plastic core is PITA for anything useful other then car dash lights , sidelight fiber, and decorative things. It does have it's uses in short range data links.
~
Numerical Aperture matters big time.
~
Which is why nobody wants the NECSEL except for video. Though we did look at buying ten of them for a stadium gig that needed 10 fixed large diameter beams that did not scan.
~
Your about to enter the twilight zone called "Etendue" Physics says you must lose delivered power trying to focus a high NA beam (Necsel) into a tiny fiber to try to clean up the divergence.
~
Fiber optics connections come in "Free space" and "Contact". When the fiber is polished for the two fibers to touch inside the connector, that is contact. When a lens focuses the light into the fiber, that is free space.
~
If you do not know which one you need, trouble is assured. SMA usually means "Free Space".
~
A 200 uM core usually just barely fits onto a 6-7 mm scan mirror when collimated. But the beam exiting has very high divergence.


Steve

[masterpj]

If your projecting graphics for laser show, you need 65 to 200 uM core max. 400 is for machining.

Steve

Thats super valuable info! Those are actually easier to come by. I reckon getting a focus on a fiber that small will prove to be difficult but worth a shot!
To me it seemed really small that 400 uM but I guess when you want to collimate it will expand for a fair bit before you can get the distance to correct it.
I will try to get a few fibers of 200 um to 100 um to see what is usable. I reckon too small and it becomes too misalignment sensitive + you end up spilling most of your output over the fiber and too much and you get too big of a resulting spot!

Thank you! (sorry im still digging and learning) but I've learned so far that I need sma 905 and found a suitable collimator too with sma 905 connector.

[masterpj]

~
If your projecting graphics for laser show, you need 65 to 200 uM core max. 400 is for Machining. 800 is for flood lights.
Plastic core is PITA for anything useful other then car dash lights , sidelight fiber, and decorative things. It does have it's uses in short range data links.
~
Numerical Aperture matters big time.
~
Which is why nobody wants the NECSEL except for video. Though we did look at buying ten of them for a stadium gig that needed 10 fixed large diameter beams that did not scan.
~
Your about to enter the twilight zone called "Etendue" Physics says you must lose delivered power trying to focus a high NA beam (Necsel) into a tiny fiber to try to clean up the divergence.
~
Fiber optics connections come in "Free space" and "Contact". When the fiber is polished for the two fibers to touch inside the connector, that is contact. When a lens focuses the light into the fiber, that is free space.
~
If you do not know which one you need, trouble is assured. SMA usually means "Free Space".
~
A 200 uM core usually just barely fits onto a 6-7 mm scan mirror when collimated. But the beam exiting has very high divergence.


Steve

See the edits just now! Really valuable info! So if 200 um equals 6 to 7mm then I'd be looking at 100um.
The free space thing seems to assume that normally the fiber is so large that there isn't much needed adjustment.. but Obviously I would need to adjust the fiber in X/Y Z to get into the focal point.
In the ncsel is a glass lens visable of sorts so that free space situation makes a lot of sense and is the case here!

Would be willing to lose power to get a better beam for the sake of experimentation. Might have a yellow and orange necsel coming which I scored for cheap (working) but have my green necsel still here laying for experimentation (in great condition and lasing) and know how to drive it and temperature control it and all.

I'm actually surprised about the high divergence part.. the divergence on the ncsel is said to be quite low but the beam diameter itself large so you'd assume you can trade properties between the 2: narrow the beam but worsen the divergence.

[mixedgas]

Usually to correct a diode with a fiber you have a professional outfit de-can the diode, bond on a tiny fiber as a lens, and re-can the diode.

This corrects one axis. There are two ways you can do this, fiber at right angles to the die and acting as a cylindrical lens, or butt the fiber right up to the die faucet and use it as a pseudo fiber launch. You can see method "B" used in a lot of VDX's hardware he has acquired with diode chips. Method A is only done by one or two companies in the world and they charge quite a penny for welding in the fiber holder next to the die.

I can get multiple single mode diodes into a single 100 um fiber, but how many and how is a trade secret around here. Correcting for the astigmatism takes a lot of optical bench space so for multiple diodes 2N + 3 lenses are needed and 2N extra fold mirrors are needed where N is the number of already collimated diodes. The mounts and lenses get expensive fast, the required size of optical breadboard is one meter in one direction. So basically starts approaching the size of a watercooled ion laser.

You can rest assured Necsel did the math to optimize their output to the best achievable. They are not in the business of throwing away light.

Steve

[masterpj]

Yeah decanning is certainly not an option.
There is an array of 24 tiny IR emitters and that non linear crystal at the top that converts. You can take the sma assembly off thats bonded on on top and then you have a square mirror with a very large beam.

Would there be a way to do this without getting a fiber up close? A holder that connects on the sma where I can move the fiber with adjustment screws in some directions to bring the fiber in at the focal point?

I'd be cool throwing away 50~40% of the light if that results in an usable beam. For people buying the unit at full price something like that would be unacceptible yes but at the price I bought it It would be very acceptable to get that precious 577 or 615!

Maybe there is a simpler approach to not use a fiber and correct what is there directly.

Very open to ideas or suggestions to try

Edit:

Attached a pic of internal without module housing on. Credit of picture belongs to dashapple.