NUBM44 6W+ 450nm Laser Diode

[planters]

I think the main thing that the pointer only people need to remember at all times is the safety of others

Agreed. Also, I've enjoyed simple audio amp driven mirrors, speaker mounted optics and lumia effects as much as the high end scanner driven graphics systems.

Its nice to have a laser working in the background while building something or while I'm on the computer and that is where the pointer is unable to deliver. But, pointers are OK, projectors are OK and so are laser tools. No problem.


The design of a heat sink also needs to take into account the fin density. It is much more difficult to machine sub mm fins, but the increase in surface area and hence the contact with the air is proportional to the fin count times the surface area.

 

[crazyspaz]

I have just completed a 180min. continuous run with the diode DTR sent to me. The diode was driven @ 4.5 A and the output power remained stable @ 6.34 watts.

The beam size and divergence were also better than the first diode I tested.

Now my question for the members here who focus on laser pointers. I have no issue with these. I am not one of those " you only want to point it at airpl###s" or cut electrical tape. But, what about the thrill in building an incredibly powerful, desktop sized module that without 3 phase power or a water source will produce 40W of blue light? You can generate patterns with a simple scanner or effects with lumia. With this kind of power you could scan an image and burn it into a 4x8 sheet of plywood. Only a small number of the laser-show professionals around the world deploy 40 W projectors. That would hold my interest for some time.

Trust me I wish I could, but that's where my chronic lack of money comes into play...

 

[Alaskan]

I agree, I didn't mean to infer radiation could be anywhere near as significant as convection. Many heat sink manufacturers make their heat sinks black because they believe the extra heat radiation from doing so is worth it.

 

[Nhfoley]

I'm working out the details of building one of these diodes into a CNC and trying to understand the beam characteristics, specifically whether I need to use corrective optics beyond the g2 or 3-element lens to achieve a focused spot size in the ≤0.1mm range.

My understanding of these diodes is that without any lens at all, near the exit of the diode, the beam profile is roughly a long ellipse, with a small amount of divergence in the plane of the major axis and a moderate amount of divergence in the plane of the minor axis.

Assuming that is correct, has anyone here measured the aspect ratio of the initial elliptical beam and the ratio of the horizontal and vertical divergences?

If they are significant enough that they need to be corrected to achieve a ≤0.1mm spot, is there a recommendation for cylindrical lenses to start correcting the beam? (Seems like I need a pair, at minimum, and some measurements of the beam to figure out their ideal focal lengths.) ...Or optical design software that would let me start poking around at solutions of my own?



Also, somewhat off topic, but what is the approx. Minimum working focal distance with a g2 lens with a diode like this? Having a hard time finding any specs for them.

 

[RedCowboy]

Yes it's a matter of surface area multiplication. Every fin is a heat vehicle and it's total surface area is much greater than the area of attachment. I often use the sinks scavenged from pc processors and other various electronics because the ratio of fin thickness to total surface area has already been calculated for optimal use of space and/or material.
I have homemade tabletops that will run all day without TEC thanks to lots of finned heat sinks and a small fan.
Pointers though have aesthetics as a top concern and although thicker mounts buy time before the duty cycle is up there are limits as output power and waste energy increases. Eventually active cooling may become necessary for HH but for now we could probably get by with a handle attached to a long finned body.

 

[RedCowboy]

I'm working out the details of building one of these diodes into a CNC and trying to understand the beam characteristics, specifically whether I need to use corrective optics beyond the g2 or 3-element lens to achieve a focused spot size in the ≤0.1mm range.

My understanding of these diodes is that without any lens at all, near the exit of the diode, the beam profile is roughly a long ellipse, with a small amount of divergence in the plane of the major axis and a moderate amount of divergence in the plane of the minor axis.

Assuming that is correct, has anyone here measured the aspect ratio of the initial elliptical beam and the ratio of the horizontal and vertical divergences?

If they are significant enough that they need to be corrected to achieve a ≤0.1mm spot, is there a recommendation for cylindrical lenses to start correcting the beam? (Seems like I need a pair, at minimum, and some measurements of the beam to figure out their ideal focal lengths.) ...Or optical design software that would let me start poking around at solutions of my own?



Also, somewhat off topic, but what is the approx. Minimum working focal distance with a g2 lens with a diode like this? Having a hard time finding any specs for them.

The one I had would not focus up close to anything but a small line with the 3 element or G9.
I get the idea of cylindrical lenses for the x and y axis but I have been hearing that not all of these diodes have the same divergence. Are there just the few odd balls or are some NUBM44's different than others?

 

[planters]

You are correct about the raw beam profile. The emitter size is not yet posted, but the raw divergence should be similar to the better known, lower power 445nm diodes.

It would actually be better to correct the two axis separately because they have different divergence and different, original source dimensions. The problem is that with such rapid divergence the two cylinder pairs would each have to be aspherical cylinders and these are very expensive. Otherwise, the cylinders will introduce spherical aberration.

The advantage with the first non-cylindrical aspherical collimator is it is a pretty good approximation and moves the beam away from the diode where more optics can be added. If spot size is the driver then a cylinder pair that is in the 5-10 x range is probably a pretty good starting point. Expand the resulting beam, which will be an asymmetrical stripe in the near field, with a PCV lens followed by a large, low FL ratio PCX. The bigger this second lens the greater the working distance.

The G-2 could be moved farther to the diode than collimating distance to produce a near field focus, but the FL ratio would be high, generating a large spot. Continuing to move it farther would reduce the FL ratio and decrease the spot size, but then this asphere will be working further and further from its manufactured FL and will introduce increasing spherical aberration.

As an extremely rough first approximation based on my initial set up, at 14M I could bring the diode down to 15mm in the problematic axis(slow). So I suspect that with the same basic aberrations in the system, you might produce a 0.1mm spot in this fatter axis at a 10cm FL.

 

[Minamoto Kobayashi]

Hi.
In my laser collection the best cooling ever made was in my Hercules 600:
a combination of a well made finned heatsink and a variable speed fan (aka active cooling system) assure real 100% duty cycle. I tested the Hercules for 20 minutes, and after the classic DPSS instabilities the diode become very very stable near the end of the 20 minutes test.
Another well made finned heatsink was created for my two custom Zaser XL (bigger head and more fins).

For the new NUBM44 diode I ordered the new Cypreus IIIb with three elements glass lenses and JL beam expander: in that way I will obtain a nice beam and a very collimated spot, despite a noticeable loss of power.

I tested the burning power with my Frankenstein build with NDB7875 @ 3W diode Vs my Zaser XL with NDB7A75 @ 5W: both the build mount a LG modified beam expander, but while the Frankenstein mount the three elements glass lenses, the Zaser build mount the G2 lenses.
The results is: the Frankenstein build has less burning power at close distance but it is capable of smoking bark at about twenty meters, while the Zaser build has a tremendous burning power at a close distance but at twenty meters the bark do not smoke.
This show how important is to have the dot line as short as possible at the output, because much power wasted over a a long dot line is an unuseful thing, at least for me.
I prefer to have a laser, not a flashlight :na:

 

[RedCowboy]

Could we not retune the 3 element design to better reflect the new emitter dimensions for a best compromise beam?

 

[Minamoto Kobayashi]

It is "handcrafted" possible this?

 

[Bionic-Badger]

Could we not retune the 3 element design to better reflect the new emitter dimensions for a best compromise beam?

The 3-element lenses are just what we get from those Chinese manufacturers. You'd probably need to buy/source some lenses to tune anything, which would be expensive and require some manufacturing of rings, etc. to space them out nicely. You'd probably be better off using a G-series lens and buying some cylindrical lenses to reduce the beam size in the wider axis.

 

[RedCowboy]

The 3-element lenses are just what we get from those Chinese manufacturers. You'd probably need to buy/source some lenses to tune anything, which would be expensive and require some manufacturing of rings, etc. to space them out nicely. You'd probably be better off using a G-series lens and buying some cylindrical lenses to reduce the beam size in the wider axis.

Yea I have several boxes of various lenses many mounted so I can play with distance on my crude light bench, some are coated some not and several beam expanders plus tons of mirrors, dicros and even some cubes, prisms and gratings. Now I am not an expert on optics however I am learning and one day I will be able to order by the numbers just what I need, but for now I'm basically trial and error. I could probably get a hot focal point at an given range with just a g2 and a sliding beam expander pair. But the formula for a perfect correction I'm not proficient at calculating yet. I bet as the new diodes become the norm some industrious Chinese manufacturer or a hobbyist here will calculate and produce or have produced an updated lens set that's simple and affordable for the masses.
p.s. Check out this simple beam effect I was goofing around with this morning and see if you can guess what it is.
https://youtu.be/c5_qiMK6BMU

 

[planters]

What is the problem with the aspheric lenses? I do not see why a three element, spherical lens would be superior. A cylinder pair or as an alternative, a prism pair is almost certainly needed because of the extremely asymmetrical beam profile. It is very unlikely that a radially symmetrical lens will be able to correct both axises simultaneously.

 

[Atomicrox]

The 3 elements lenses crop the beam, like a sort of spatial filter. That's why I (and a few other members) prefer them to the G lenses. You lose about 20-30% power but to the naked eye your beam still looks better and that's what counts in handhelds unless you're using the laser to burn stuff.

Of course proper correction would be better, but that is costly and makes handhelds too big.

Edit - Sorry to derail the thread, but you might be interested in this. You have a lot of equipment and expertise, would be cool if you could analyse that phenomenon.

 

[planters]

The problem with cropping the beam anywhere but a focus is that diffraction effects will deteriorate the beam quality.

If I was trying to do the same I would use the standard G2 or G9 and then a small PCX/pin hole/PCX. There is no optical design software involved. It is all coaxial and could be mounted in a long threaded tube with ring spacers. The pin hole might take a few trial and error experiments.

I'll look at that link. It seems interesting.

 

[RedCowboy]

Shouldn't there be a way to semi square up the raw output very near the diode face via a tiny cylindrical convex rectangular lens then focus that semi squared output from a reasonable diameter?