JLasers - High Power LD Divergence

[loreadarkshade]

Comparing High Power LD Divergences
Click the images or this link for more details

 

[loreadarkshade]

The empty data will be added in the coming days

 

[CurtisOliver]

Comparing High Power LD Divergences
Click the images or this link for more details

View attachment 77536View attachment 77537

Your work regarding these blue diodes hasn’t gone unnoticed. Great work and much appreciated. The power graphs with wavelengths and now divergence comparisons are incredibly useful. It’s going to be very useful now to be able to choose the right diode out of the wealth of diodes available based on certain characteristics.

 

[Gryphon]

This is gold! Amazing data, thank you!

 

[loreadarkshade]

My pleasure folks! I do this for you guys
The graph is to relative scale by the way
I've added the 7875 data... wow. What a crazy tight diode!

 

[paul1598419]

For the longest time the NDB7875 was my highest power blue diode. The divergence on them is so much better than the others I've used.

 

[RedCowboy]

The NUBM42 and NUBM49 have worse divergence ( cast a wider raw bar width at a given distance ) than NUBM44/47/08/0A/0E/0F
You have to measure the raw output without a lens because of edge clipping, even the copper modules we use will clip the edges of the output of the higher divergence units.

That said, even in a copper module and with a lens the 42/49 had a wider bar than the usual suspects, in my limited testing anyway, that and the 42/49 had lower max over driven output than the NUBM44/47/08/0A/0E/0F and I measure them all in a heavy heat sink but without active cooling, because most builds here don't employ active cooling, simply a large enough heat sink to accommodate the desired duty cycle.

Edit/Note: I have harvested many diodes, however my NUBM42/NUBM49 test subjects were purchased, so what a Chinese reseller presented as 42's and 49's may or may not have been correct and variations may exist between batches and production dates.

 

[loreadarkshade]

I measured power with stock Gball, decanned and then tested with G2. they were identical to the mW.

Divergence without lens /= divergence with lens.
This is because measuring divergence with a lens projects the laser crystal, and focuses it.
Think of a diverging point source. Huge divergence without a lens, but a pin point with a lens.

in other words, Smaller die crystals will have better divergence with a lens, no matter their raw divergence without a lens.

if you look at the datasheet, the 7875 has WORST raw divergence than the NUBM44, but with a lens, this is a totally different story. Like I said, divergence with lens is based on crystal size and not raw divergence.

What use would it be testing power and divergence without a lens, most builds will use a copper module and lens

I used active cooling to maintain 21°c to keep tests consistent. The diode is at a constant temp throughout.
Letting my module get warm even by a degree would throw all my tests out of whack.

All divergence tests were done with the same exact 2E lens, at same exact distances from the wall (about 13ft), and then focused as tight as the beam would go, and then measured with my calipers. Not sure how you got worst divergence from the 42/49 !

 

[paul1598419]

Actually, the divergence depends on the emitter size alone.

 

[loreadarkshade]

Actually, the divergence depends on the emitter size alone.

Exactly

 

[RedCowboy]

I measured power with stock Gball, decanned and then tested with G2. they were identical to the mW.

Divergence without lens /= divergence with lens.
This is because measuring divergence with a lens projects the laser crystal, and focuses it.
Think of a diverging point source. Huge divergence without a lens, but a pin point with a lens.

in other words, Smaller die crystals will have better divergence with a lens, no matter their raw divergence without a lens.

if you look at the datasheet, the 7875 has WORST raw divergence than the NUBM44, but with a lens, this is a totally different story. Like I said, divergence with lens is based on crystal size and not raw divergence.

What use would it be testing power and divergence without a lens, most builds will use a copper module and lens

I used active cooling to maintain 21°c to keep tests consistent. The diode is at a constant temp throughout.
Letting my module get warm even by a degree would throw all my tests out of whack.

All divergence tests were done with the same exact 2E lens, at same exact distances from the wall (about 13ft), and then focused as tight as the beam would go, and then measured with my calipers. Not sure how you got worst divergence from the 42/49 !

Here's some information for your information.

NUBM laser diodes do not use crystals like DPSS lasers do, they use substrate stacks with a reflective rear facet and a front facet with a partially reflective area at the center of the P/N junction, this is where the photons exit ( the entire stack glows, but that is not contributory to the beams divergence. ) and it's the length of the wave guide along with the cavity shape and size as well as the front facets partially reflective emitting region size/shape that factor into the divergence, the physical size of the substrate stack can not be used to judge divergence, nor the reflection you see of the entire front facet.

A 2E lens will not yield accurate divergence data because it clips the ends of the output on the fast axis of higher divergence diodes.
You need to use a short focal length single element lens to catch all of the output without clipping to get accurate divergence data, such as a G2 lens.

Here's a test, compare the same diodes with a G2 and see how your results compare..... will your results be proportionally comparative to your results with your 2E ? Not if you test a NDB7875 and a NUBM44 with a 2E then a G2 because the 2E will clip a lot off of the edges of the NUBM44 output and proportionally less off the NDB7875 output.

 

[loreadarkshade]

I meant crystals as in the gallium nitride semi-conductors
I see! Oh boy, I hope youre wrong so I dont have to run all those tests again
Ill try this later with 3 diodes and see what we get

 

[loreadarkshade]

You were correct in that the 2E lens clips off the NDB7875 fast axis! Remarkable. This is new to me.
The G2 lens does not since it focuses so closely

However my divergence measurements were equivalent with the G2 lens as with the 2E lens

Real measured divergence:

.................................2E............G2
NUBM47-A1 - 32mm - 76mm
......NDB7875 - 11mm - 25mm

If we use cross-multiplication to predict what the NDB7875 divergence should be with a G2 lens from our data, we get 10.53mm which is smaller than the 11mm we should get (this is error in measurement from calipers) indicating the divergence remains consistent
(32 * 25) / 76 = 10.53mm

This can be further proven when using this example:
Using a magnifying glass to focus the sun onto a wooden plank. The sun emits light in an omni-directional pattern. 99.9999% of the sun's light is being "clipped" around the glass lens, yet the lens can still focus the sun to a small point. If the sun was larger, it would cast a larger point on the ground, etc. Clipping does not affect the crystal die projection.

Try this with a bi-convex lens indoors. The lens can still focus your ceiling lights onto your hand, even when the light isn't coherent. A smaller ceiling light will have a smaller emitter size on your hand.

In fact, if your hypothesis was correct, all the LDs whose divergence clips the lens should all have the same divergence measurements, but they differ

 

[CurtisOliver]

Remember diode spec sheets state ranges and central figures for a reason. Two diodes of the same type can differ in results. We all know about centre wavelengths and how some diodes are lower or higher. But the same thing occurs with lasing thresholds, and divergence.

 

[RedCowboy]

The laser beams divergence ( starting beam waist and ending beam waist over a distance ) expressed as MRAD ( milliradians ) is different with different focal length lenses, this is why GBall diode data sheets have very different divergence numbers than sister diodes with flat window cans rather than lensed cans ( GBalls )

If the diameter ( beam waist ) of a laser beam increases for example by 1 mm per 1 meter of beam path/travel then the divergence is 1 mrad, so you need to measure the beam waist at 2 points along it's path, which will be different with different focal length lenses, but if a lens clips part of the output, then you won't even get an accurate measurement.

Are you using the artifact to try and determine divergence ?

488nm GH04850B2G

I've picked up a 488nm GH04850B2G build which is great, but I think it has the G2 lens, and definitely has the large rectangle artifact on one side of the spot. Would swapping the lens for the G8 totally remove that artifact? Also is 275ma still the agreed upon safe current?

laserpointerforums.com

Explaining "box reflection" and how to get rid of it!

Hello Guys, as I learned very much from the knowledge thats stored in the forum, I want to give something back. I made some experiments and can now say that I fully understand the so called box reflection. I do also have a way to eliminate it. Its not easy however. But first the basics. I took...

laserpointerforums.com

 

[loreadarkshade]

I am only measuring the beam width (the high brightness part of the focused line)