40-50W blue laser from tightly packed parallel 08 beams

[RedCowboy]

How hard are you driving your 08 block ? That is what is the output from your power supply going into the 8 diodes ?
I am curious how many watts are you putting through the PBS cube from techhood because I did not think they would take tens of watts ?

Are you still planning to do this with a nubm44 block from a V1 using G2's ? If so I would be interested to know how much a PBS can take.


--edit--

I see 1.99a on your power supply display, so if you are driving two strings of 4 in series parallel @ 15v and 2a then each diode would be getting 1a making 1.1w at each Gball for a total of 8.8w raw for the block and maybe 7w after losses in your final beam ? or are you driving them all in series at 32v and 2a ?

I have under driven GBalls before to get a smaller/tighter beam but I have not attempted to knife edge from GBall intact diodes, it's cool to see you have made it work but I am curious about your final beams wattage, can you share a result please ?

 

[brianpe]

Great work. I didn’t know you could get knife edging that tight. It looks like your cylindrical lenses are about 1-2 cm from the collimating lenses. Was that where you got the tightest beam? When I’ve used cylindrical pairs I‘ve had to butt them as close to the collimator as possible to keep the divergence tight.

 

[Light superglue]

Brianpe, yes, they are 1 cm away, this is needed to keep space for temperature sensors on the block. I do not think that a 1cm separation would make a huge rise in divergence.

Red, did you look at the original thread? I always connect all LDs in series, never in parrallel, because less current means less losses in the curcuit. 32V at 2A make 64W load. But with 12V, 15A max. rated power supply at the moment I cannot make more than 3.3A current at 35V.
I am replacing it by 24V supply in order to make the block glow at full power at 4.5A current.

About v1 light engine with 7 G-2 on top I did proof of concept and it seems to work but I would post this in another thread due to different design and many pictures to follow.

 

[Light superglue]

Now I have replaced 12V AC-DC adapter by 24V one and can reach 4.5A current through all 8 LDs in the block.
Since my walls already have many marks, I wanted to protect the wall by a thin PTFE sheet (thickness might be 0.2mm) but it did not help much - after half a minute the sheet and wall behind were burning again. Then I replaced it by a thicker one (0.5mm) and this protection has been doing well.
The beam spot at 4.5A is still the same 2cm square as before at 2A but the beam itself is very bright and well visible even at day. The spot is very bright even if shot through OD6 googles and my camera can hardly see the ruler under such a brightness...

About PBS:
I mesured its temperature by a tiny IR-thermometer holding it very close to the top of the cube.
Another NTC thermal sensor is attached to the N. block.

So if current of 3A runs for 5min, the T on the surface of the block stabilises at 50ºC with HS fan working and PBS surface reaches 65ºC (very hot to touch).
Then I wait 2-3 min untill block is down to 30ºC and PBS to 35ºC (there is no fan cooling it) and start 4.5A current.
In 1 min block is 50ºC, PBS 53ºC.
In 2 min block is 60ºC, PBS 62ºC. Here I stop and wait again.

I have repeated it twice, and do not notice any visible damage to PBS cube. Maybe they have to be of special quality for hours of constant working but for 1-2 min the Techhood one seems OK.

 

[RedCowboy]

Great, I am glad the cube is holding up, when I talked to OPT in Poland they said I would need a special made cube for over 10w but they may have been thinking about continuous duty.

 

[Light superglue]

I wonder what could be the temperature inside of the cube. What I can mesure by IR thermometer is on the surface.

Question:
When PBS suppliers say about damage threshold what do you think they mean: damage of the beamsplitting layer inside, glass material itself (from unequal heat distribution because glass is not a metal to evacuate heat to surface quickly) or AR coating on the surfaces?

Anyway yes, for continuous duty it might be better to think about PBS cooling as well - something like putting it steadily on a HS with small fan and adjusting the 2nd beam direction by moving the side mirror.

Thorlabs also has polarizing BS plates (which have bigger surface compared to cubes) but very expensive and not for 450nm.

 

[Light superglue]

Well after 1 year when many of us were playing with new toys called NUBM31/34T it is about time to revive this project. 31T is strong but does not have focused beams!

Likevvii wrote that it was made for pumping phosphor at 10-20 cm distance and it looks to be true. More a source of blue light than a laser. Laser for me is something from Star Wars movie, must be a thin beam hundreds of kilometers long... or at least hundreds of meters...OK, OK at very very least - hundreds of centimeters!!!

With NUBM08 block I have almost done a build like this after all beams were corrected with cylindricals and knige-edged+PBS combined...but the issue still was each beam being focused differently by its G-ball lens and once well and badly focused beams were combined, the result was not perfect.

During last months I was trying to solve this issue by inserting an extra block between cylindricals unit and KE units, I would call it refocusing block. The design is a 20 mm thick Al plate with M9x0.5 treads and in each thread I put 2 Sanwu G-2 collimators head to head.
This positioning makes a galilean expander (I already mentioned this in another thread) with expansion ratio 1.0 (so no expansion) but allows to refocus the beam, in words make a well focused from a badly focused one.

Long story short I had to dismantle all thing, repair some newly found issues with bad wire soldering and thermal sensor... and started to assemble it back.
And yes, refocusing works. You see the spots before I add this unit and after that, the collimation is greatly improved. What is lost, however, is the parallel orientation of the beams. That's why I could not put this block before the blocks with cylindricals.
Yes my block with 16 Sanwu lenses dos not have precision of the Japan made LD array, but this challenge I think to counter at KE stage to come soon...

 

[gazer101]

Well after 1 year when many of us were playing with new toys called NUBM31/34T it is about time to revive this project. 31T is strong but does not have focused beams!

Likevvii wrote that it was made for pumping phosphor at 10-20 cm distance and it looks to be true. More a source of blue light than a laser. Laser for me is something from Star Wars movie, must be a thin beam hundreds of kilometers long... or at least hundreds of meters...OK, OK at very very least - hundreds of centimeters!!!

With NUBM08 block I have almost done a build like this after all beams were corrected with cylindricals and knige-edged+PBS combined...but the issue still was each beam being focused differently by its G-ball lens and once well and badly focused beams were combined, the result was not perfect.

During last months I was trying to solve this issue by inserting an extra block between cylindricals unit and KE units, I would call it refocusing block. The design is a 20 mm thick Al plate with M9x0.5 treads and in each thread I put 2 Sanwu G-2 collimators head to head.
This positioning makes a galilean expander (I already mentioned this in another thread) with expansion ratio 1.0 (so no expansion) but allows to refocus the beam, in words make a well focused from a badly focused one.

Long story short I had to dismantle all thing, repair some newly found issues with bad wire soldering and thermal sensor... and started to assemble it back.
And yes, refocusing works. You see the spots before I add this unit and after that, the collimation is greatly improved. What is lost, however, is the parallel orientation of the beams. That's why I could not put this block before the blocks with cylindricals.
Yes my block with 16 Sanwu lenses dos not have precision of the Japan made LD array, but this challenge I think to counter at KE stage to come soon...

Could you use some sort of fiber optic coupling system to combine the individual beams into one-ish beam?

 

[Alaskan]

Complex build as it is, individually steered beams to either parallel, or to combine to a single spot would be the ideal, I think. He could build something like a FAP to produce what you are referring to, but not so strait forward as a FAP uses without having the emitters placed side by side in a strip without the relatively large mounts they use now, at least, for a single fiber output.

 

[paul1598419]

Trying to add fiber optics to a block is going to be much more difficult than you might imagine. Pretty expensive too. You'd be better off buying something that already has the cables attached.

 

[gazer101]

oof, if only there were a cheap and easy way to combine these arrays into a single beam. Imagine 1kW for <$3500

 

[paul1598419]

There really isn't a need for fiber optically coupled 445nm/450nm beams from individual diodes, but there has been for 808nm diodes. That's why you can buy those relatively inexpensively.

 

[RedCowboy]

There really isn't a need for fiber optically coupled 445nm/450nm beams from individual diodes, but there has been for 808nm diodes. That's why you can buy those relatively inexpensively.

That's incorrect.
Copper absorbs 450nm better and is now used for material processing, in time these will be available on the 2nd hand market.

1000 W blue fiber-coupled diode-laser emitting at 450 nm

The availability of high-power blue diode-lasers established a new class of laser sources for materials processing recently. With the significantly shorter wavelength compared to conventional laser sources for materials processing new applications are moving into the range of the feasible. There...

www.spiedigitallibrary.org

Fiber-Coupled Laser Diodes at FLC: High-Efficiency & Precision

FLC's fiber-coupled laser diodes deliver high efficiency and precision, perfect for demanding applications.

www.frlaserco.com

LASER 2019: Laserline and Nuburu demo kilowatt blue power

Panel session talks highlight systematic advantages of welding with 450 nm diode sources.

optics.org

 

[paul1598419]

I should have said there hadn't been a need for these. Eventually there may be a secondary market for these used units. Only time will tell as it isn't clear yet how many of these have been sold. If it turns out that as many as the 808nm fiber coupled units were sold, this could be true.

 

[RedCowboy]

-----

StackPath

-----

 

[paul1598419]

The vast majority of 808nm fiber optic units were never used for welding, but as pump packages for solid state lasers. That is why so many were sold. It remains to be seen if the demand for blue optical fiber units will match that of the pump diode demands. I kind of doubt it as there was great demand for pump packages for all sorts of applications.