Combining 4 GH04W10A2GC diodes

[Light superglue]

Here I have finished the long expected device to combine 4 beams from Sharp GH04W10AGC diodes in the old fashioned Lasertack micro-KE beam combiner!
These are not new diodes DTR is selling now but still the old ones what Hakzaw extracted for me from used Junktronix modules.

Old fashioned I mean because LT does not produce combiners any more with offset of 3 mm. The new modules are for combining like IIII. But with the old one I can do the beam geometry
II
II

and this is better for later focusing and cylindrical correction than IIII shape would be.

The story was that:

First - I have connected all diodes in series and tried to arrange the beams with mirror prisms rated for 400-480nm which I have been using all the way for NUBM44 and NUBM08 but these ones did not reflect well these 405-410nm beams (pic. 2). Probably because at this range limit the reflection efficiency was too sensitive to beam polarisation... so Alex Schatalow from Lasertack has sent me the replacement - new prisms rated especially for 390-440nm free of charge. Thank you Alex!!!

Second - the oval holes in the mirror mount bases were not long enough - I drilled them to elongate a little to be able to install prism mirrors almost at the bottom - 0.5mm high from the baseplate. In this way more of the reflected upper beams were passing over the bottom prisms. The whole was mounted on the Al box for electronic components having the size to insert 2 cubes of ice. I tried this with Ushio red diodes earlier and this trick helped to keep them at room temperature over 10 minutes working time or such.

Once mirrors adjusted the spot at 5m distance is about 1cm long. At 0.1-0.2A it looks as a line but at 1A becomes thicker. The wall paint still does not burn at this power.

 

[RedCowboy]

Have you done a LPM test to see how much loss there is ?

 

[CDBEAM777]

HMMMMMmmmmm ???

Going to Cat Show with the wife...……...…....or alignment of Four Beams

………………………………………………………./\………………………………………………...

.......................SO...Looks like they are balanced on the PITA scale !!!...…….

CDBEAM

 

[Alaskan]

How old is that knife edge combiner?

 

[Light superglue]

Hi CDBEAM,
If you look deeper into this stacking the upper left line on pic.7 looks dimmer than other lines - I may need to reposition the mirror later or pretty much of energy is still falling on the edge of the prism and lost.


Hi RC,
This will come now! My old LPM has zero line at -0.16W and here are the pics. of power test.

I tried turning it on up to 1.4A just for couple of seconds to make pictures and then added ice into the well isolated box and waited until the Quad mount temperature came down to 6ºC.
Keeping it cold gave some gain vs. room temperature: from 4.88 to 5.23W at 1A and from 5.33 to 5.9W at 1.4A.

Then image i8 was made after keeping 1 min at 1A, image i9 when current starts at 1.4A after cooing to 6ºC again, and image i10 1 min after that.
So based on last pic. about 5.4W power can be obtained from these four 2nd hand diodes together at 1.4A if they are kept at 20ºC but 5.2W at 1A and 16ºC looks more efficient (25% overall efficiency here if we count).
I would say that even at this little cooling going over 1A is not much worth it.


Hi Alaskan,
According to invoices I keep that combiner (Article No. PD-01163-E) was purchased in August 2016 for S06J diodes but never used for that.

 

[Alaskan]

Hmm.... I have a couple that old I've yet to use, thanks.

 

[kecked]

PBS two beam, knife edge result of 2 PBS cubes. Use wave rotator if you want to not have an x.
Not that hard. Oh and pay your royalty bill to arctos.....

 

[Light superglue]

Why PBS 2+2 beams? I would better PBS 4+another 4! 10W beam would be better than 5W, would not it?

Now we come to the 3rd bunch of pictures and here the featuring pictures should be those of the super bright beam making walls explode or something similar...

But unlke the previous super bright beams made from 4 combined NUBM44 or 8 parallel NUBM08 beams this 405 or more likely 410-415nm beam is not as bright itself as glass becomes where it passes through - so on all the pics my handy could do the combiner itself was brighter than the beam. Finally I made a picture of the beam with no combiner in the image and here its propagation can be seen well.
And then I felt a smoky smell - beam finally started burning the white wall paint at 5m distance. This was slower than 4 combined NUBM44 (they did it in 5 seconds) but bingo - another wall burner has been made and proven!!!

I am also thinking about another design of experiment - installing 2 bars into the ice box with sliding lens mounts aiming to add a kind of focusing beam expander here...

 

[kecked]

You can only pbs combine two beams. One s polarized and one p polarized. You then have to knife edge combine the two resulting pbs beam sets for four combined diodes. Sorry no free lunch.

Is that a tredmill in your kitchen?

 

[Light superglue]

Hi Kecked,

No, it is a bicycle. Tredmills are for those afraid of going out to the street.

I know the things you say, I have combined up to 4 laser beams by PBS, dichro and KE many times before. Now I want to go further and start combining 8 beams - so testing now different ways of doing that... And I am not after a free lunch here.

 

[Alaskan]

I'd sure enjoy looking at how you made the four beam PBS combiner, I have an idea of what you might have done.

 

[kecked]

I had not thought of this but you could do it. Do aThe equivalent of a knife edge into a pbs cube. Now you can prestack a couple beams next to each other and then combine the stacks. No mirrors if you are careful with diode placement. Interesting idea I had not considered. Use a rotator so the rest is only as thick as one stack. This means you get a beam that has less mirror loss and is half as thick.

 

[Alaskan]

That was my thought too. Another way I saw someone posting about wanting to do many years ago on PL was to drill a hole through a cube to send a third beam through it to coaxially combine with two much larger diameter beams going through the cube. I've love to try that too, but having a cube drilled through like that accurately in plane with the cube without damaging the layer in the middle would need someone with the right equipment, not cheap.

That would be a weird beam, only good for looks, a light show machine? I am sure there are probably less expensive ways to produce more power, like just knife edging more together, but they wouldn't look the same.

 

[kecked]

I don’t see how the hole really helps. It just ties you to an orientation. Passes through the cube you are free to use the entire cube. In fact....! You could use a broad cube and do rgb through a single cube and the regular dichro combining. Eliminates two cubes

 

[Alaskan]

For the moment I cannot remember the method of utilizing a hole through it to add a third beam of any polarity or wavelength. I've been trying to find the thread over at PL, not found. Can't remember what angle I was seeing with such a method now, will probably remember later.

 

[kecked]

It's just a clever way of knife edging. No reason you could not pbs combine all the sets of beams you want and then knife edge the result.
If they were slightly different wavelengths you could dichro combine the results for a single collinear beam. just have to find the right dichro to do it. Say 660 with 650 with 640 Just need really expensive dichros but yea you could get a wideband red out of it in a collinear beam.

How about PBS 660 PBS 650 PBS 640 twice and knife each pair then dichro all of it for 12 colinear beams. not sure what that would even look like. You'd have to look at relative brightness to see if it was worth including 660. Still gets you 8 beams. Straight up knife edge seems less work but a fatter beam. You MIGHT be able to tilt tune the same dichro stock....