My NUBM08 Beater.

[Pman]

So out of all the lasers you have built a 44 is probably the most output you can get right now as far as I can tell. I still haven't bought one yet although I do have 2 07E and a third diode awaiting a build. Bought and built an 08 because it was a different wavelength. To be honest I was hoping a stronger diode would have come out by now that could be pushed into the 10W range so I've put off the 44 even though it would output double what my 3.5/3.6W 7875 lasers would. May have to give in and get one as I do have (2) MX900 with sinks ready to go plus another spike type unit with a copper sink although I wouldn't use that one as it's only good for (2) 18350. If I get the paypal $ 3 people owe me I might bend and buy.

 

[RedCowboy]

YES the NUBM44 is King of the Hill at the moment. Running at 4.5a they last a long time, actually I don't think I have killed one at 4.5a yet and I work them hard. The 44 blows the 07 and the oldies away, it's just highly divergent, but just a 3X expander from sanwu makes it a BOSS burner.

Now I had a NUBM44 running at 5.0 amps in a heavy heat sink and I don't know what the output was but it was noticeably stronger, I got junk on the window sadly.

I have heard numbers over 8 watts and I don't doubt it, maybe one of DTR's larger copper mounts and the SXD set at 5.0 would be a good pair as heat is the enemy, heat build up nibbles away at output while a chilled diode can output some higher than normal numbers.


I theorize that future 455nm diodes will get stronger but I hope we get some with the can/window, the Gball presents some challenges, if you don't have a NUBM44 running at 4.5 - 4.75 amps I would get one and test it in a good heat sink, I bet it's the strongest you will see right now.

If future 455's are all Gballs the nubm44 may become a collectable gem.

It's almost always easier to lens down a larger more divergent but stronger single emitter/diode than combine or converge many smaller but less divergent diodes.
For instance a NUBM44 with a G2 and 6X cly fast axis correction makes a bar/spot about the same size as a NDB7875 with a 3 element at 15 feet. But it has 250% the power.

 

[Exerd]

I've always been on the same "deal" you are on with expansion leading to higher burning potential at further distances with the higher diverging multi-modes, but that thought brings up an interesting question I have RedCowboy.

Instead of expanding, which takes a lot more room and mounting components, why not just scale up the first laser lens? You can easily "expand" a beam 3X just by using a laser lens with 3X the FL, and same NA value as the G2. Albeit it must be a good laser lens so that spherical aberration is gone, but Edmund and Thor have a ton of these. As a matter of fact I am ordering one tomorrow for a knife edge array of NUBM07+07+08+NUGM01. Yes a bit funky throwing a green in there, but the color should look nice if I can get them tuned very tight.

That's just something I always wondered through, why use an expander if you could simply use a good 25mm lens to provide the same output minus the weight?

 

[RedCowboy]

I have experimented with wider lenses at longer focal lengths, but because the fast axis is so much more divergent, beam shaping is a more efficient method and keeps the optics smaller.

I actually have used a 2 element 25mm lens in the past, but it did clip some.

Also we are dealing with a multi mode beam, the center dominate wavelength is the strongest, but the side bars are not going to focus the same and this really shows up at longer distances, that's the shadow around the beam you see at night around the distant beam.

In order to get all the output into a small point at distance takes a ever widening lens, although an anamorphic lens of the right grind as Alaskan has suggested could work. But round only lenses will get big and the slow axis will cross over, the output needs to be squared up, but a wider lens at a longer focal length can be used, that's what the 3 element and 3X expander are doing, the spot is an ever widening line and always will be if not squared up.

Here is my 25mm lens on a M-140. There is no other lens, just the raw diode's output into the 25mm and it's a slide focus.

I scavenged them from surplus 52 inch projection TV optics. But the 06 did start to burn the AR coating after a while.

 

[Exerd]

I have experimented with wider lenses at longer focal lengths, but because the fast axis is so much more divergent, beam shaping is a more efficient method and keeps the optics smaller.

I actually have used a 2 element 25mm lens in the past, but it did clip some.

Also we are dealing with a multi mode beam, the center dominate wavelength is the strongest, but the side bars are not going to focus the same and this really shows up at longer distances, that's the shadow around the beam you see at night around the distant beam.

In order to get all the output into a small point at distance takes a ever widening lens, although an anamorphic lens of the right grind as Alaskan has suggested could work. But round only lenses will get big and the slow axis will cross over, the output needs to be squared up, but a wider lens at a longer focal length can be used, that's what the 3 element and 3X expander are doing, the spot is an ever widening line and always will be if not squared up.

Here is my 25mm lens on a M-140.

Ahh ok, I see what you are saying, I think.

You're saying that while technically a 3X longer focal length than a G2 with same NA as a G2 will produce a beam 3X larger with 3X less divergence than a G2 alone, it's the fact that the corrective optics could not be as small since the beam would be expanded already pre-correction.

Is that what I'm reading?

 

[RedCowboy]

The raw diode's output is already expanding, I just catch it further out with a wider lens. Rather than 2.39mm more like 25-30 mm and the beam is 25mm wide, I did have some clipping.

The input lens in sanwu's 3X re expands the 3 elements output, we can cut those 2 steps, but correction still works better for longer distances otherwise the line length grows with distance in proportion to it's height.

It worked pretty good on my NUBM06, this video is 14 feet in distance, just that 25mm lens set up as a sliding zoom and no other lenses and the factory Gball was removed.

 

[Exerd]

OK here's what I'm trying to allude to:

Scale.

I don't offhand know the specs of the G2 lens (it's simply the only small lens I use regularly). But let me make up some phony specs to convey my idea; say that the G2 has a 4mmFL, and it's diameter is 6.5mm. Taking a step back, the raw output of the diode is the same divergence at 1mm from the diode as it is at ten feet based on an inverse square rule. IE at 10mm from diode the raw projected size will be ten times larger than it was at only 1mm, however the fast axis and slow axis still remain in the exact same ratio regardless of distance--without focusing optics used.

Going back to my phony specs assuming a G2 is 4mmFL with 6.5mm diameter (let's assume an NA of 0.6), scaling those values up to a bigger lens would mean that no clipping would occur since the acceptance angle (NA) remains the same and the raw output keeps the same fast and slow axis angles of emission regardless of distance.

Taking that idea directly to use would then mean this:

A lens of (4mm*3)12mmFL and diameter of (6.5mm*3) 19mm with NA of 0.6 means you have a copy of a G2 that is scaled up 3X. This means if no clipping occurred on the G2, then likewise no clipping would occur on the 19.5mm lens. Since the raw output expands at a constant angle across both axis, when it leaves the 19.5mm lens the exit beam will be scaled up 3X in size, but divergence will be 1/3 of the G2. This is the very principle of expansion; expansion is really just the use of a longer system focal lens with diameter that will not clip.

In conclusion, a good laser lens will in fact duplicate the expander output solely based on focal length being longer per scale, and a scaled up aperture so clipping won't occur.

Each diode has a critical lens NA value required not to clip based on raw output. If the fast axis expands at a rate of -20°/+20° from optical center, that means that any lens with a minimal NA of 0.36--which has a max acceptance angle of ~42°, will never clip. Again this is just assuming the diode has a full angle of ~40° on fast axis, though it maybe be less or more than that.

So to wrap it all up, the 12mmFL x 19.5mm diameter lens will "see" the same input of the smaller lens, but the beam will exit at 3X the size, with 1/3 the divergence of the smaller lens, also with same exact efficiency as the smaller lens. In essence you have just expanded the beam 3x, without using any more lenses than one in the whole system.

The lens must mimic the smaller lens in aspheric coefficient to truly be a perfect scale. Using any spherical lens would produce rings when focused to a point, so it is crucial to use a laser grade aspheric contour.

But! This whole idea assumes no correction involved. Corrected or not though, the beam is still expanded at 3X over the G2, so even the fast axis will have an angular deduction to 1/3 of the smaller G2 lens. IE, the line size at 90ft will be the same as it would be at 30ft from the smaller lens.

If correction still needs to be used, this changes everything since it would need to accommodate a beam size starting 3x larger.

But take all this with a grain of salt. I'm fairly certain this is how it will always work, but spherical aberration can still plague good lenses, and the G2 has extremely little to no spherical aberration. Might take some trial and error to find a truly good lens output, but the specs when scaling should still always hold true when the lens performs well.

Please do enlighten me if I'm missing the boat somewhere here.

 

[RedCowboy]

The rear FL of the DTR G2 is 2.39mm just FYI and I must admit that I am not a lens expert, in theory you should be able to scale up and your starting beam width would be wider and your ability to focus tightly at distance extended, but because of the diodes uneven divergence you will always produce a line that with distance is getting proportionally longer as it goes, even if it's a tiny line it will be a lengthening line.

Both axis diverge at a different rate and the longer part of the initial output is actually the slower diverging.

Shaping the aggressive axis is the easiest way, because with round lenses only, even aspheric you will always have a lengthening line no matter how small you shrink it.