LOC reflection damage

[FlightLevel350]

I have heard about how the long open can diodes are susceptible to reflected light damaging the thin wires in the diode. I have a LOC driven at 412 ma, using an aixiz acrylic lens. My question is, will this ultimately damage the diode because the acrylic reflects light back onto the diode? Should I consider getting a glass lens?

 

[daguin]

Reflected photons are a problem with ALL diode lasers. All other factors being equal, the lower loss the lens, the greater chance for long life of the diode.

ALWAYS use the best lens you can afford.

Peace,
dave

 

[ElektroFreak]

Acrylic lenses won't necessarily lead to diode death from reflection, but I have killed several LOCs with back reflections so I do know that LOCs are particularly sensitive to them. I only use LCCs now. I believe that the reason for this is that there is no window or can to help block the returning light (although I could be wrong)...

 

[FlightLevel350]

I guess i should think about getting a lens from jayrob, I've been putting it off, but maybe it's finally time to get one. Plus the increase in output should bring my laser over 300mw

 

[Benm]

The window will not make a difference really - if its transmittive in one direction, it will transmit equally well in the other.

What does matter is the shape of the side of the lens facing the diode. If this side of the lens is convex, it will reflect some of the light but spread it out so little of it will be reflected back into the laser diode. If it is plain (plano), it will reflect a fair amount back, the closer it is to the lens, the worse this becomes a problem. This might happen when you use a plano-convex lens 'the wrong way around'.

If the surface would be concave it could even focus the reflected light to a small area and damage a bond wire or something, but there would be no use for a concave lens at this position. A meniscus lens also has a concave surface, and could be used in this application, though i dont think they often are.

 

[ElektroFreak]

^As I said, I could be wrong about the window, but the shape of the lens doesn't explain why LOC diodes are much more sensitive to back reflection as LCC in the exact same optical configuration. My own experimentation consisted of two LOCs and two LCCs each in a z-bolt heatsink with an aixiz glass lens. I then took an uncoated piece of glass and held it directly in front of the aperture. The LOCs each died instantly, while one of the LCCs died after several tries, and I still have the other.

 

[Benm]

Oh, i misread things there - so you reflected light back into the output focus lens.

Seems like an interesting difference in that case. I'm still not convinced the window is what makes the difference though, i suppose the diodes have internal differences as well.

The first thing to determine is if the uncoated glass is hit at exactly 90 degrees, which should send the light right back into the lasing area, or at slight angle where it would reflect onto some other part of the LD.

Have you examined the dead LOC diodes, and found things like broken bond wires or perhaps other damage? It would also be interesting to open the damaged LCC for further inspection, though i suppose it would be difficult to exclude any damage was caused by removing its case.

The best experiment would be to remove the case/window from a working LCC and see if it becomes as sensitive to the problem as LOC's are normally.

 

[Flaminpyro]

Good for you going the extra mile for the answer ! and good for the rest of us, Kudos to you my friend ! have you inspected the diodes under magnifacation ? that would be interesting to see what happened and maby de-can the LCC and repeate the test to see if it dies right off. +1 for you guy

Peace All Pyro...

^As I said, I could be wrong about the window, but the shape of the lens doesn't explain why LOC diodes are much more sensitive to back reflection as LCC in the exact same optical configuration. My own experimentation consisted of two LOCs and two LCCs each in a z-bolt heatsink with an aixiz glass lens. I then took an uncoated piece of glass and held it directly in front of the aperture. The LOCs each died instantly, while one of the LCCs died after several tries, and I still have the other.

 

[ElektroFreak]

Unfortunately, my shop is limited as I live in a small house.. I don't have the ability to conduct a microscopic examination, but I still have one of the LOCs from my experimentation that I would gladly send off to anyone who might have that capability.

I also didn't try to de-can an LCC and retry the experiment, in fact that didn't even occur to me.

After killing a few LOCs while trying to set up various optical combinations, I came to the conclusion that reflections were killing them. I tested this outright and sure enough, instant death. I had used several LCC diodes in the same combinations with no trouble, so I was immediately curious. That's when I did the above experiment and discovered that for some reason the open-can diodes are more susceptible to death from reflection.

During this experiment, I tried to make sure the glass reflected the light directly back into the lens at least for an instant by wiggling the glass slightly. Sure enough, the instant the angle was right, the LOCs died.

 

[LarryDFW]

FlightLevel;

Throw the 50 cent plastic lens away.

Use at least an Aixiz glass lens for your Red (under $10 shipped).

I use only glass lenses, and I have never damaged a Red diode.

The best glass lenses are about $60, but they are AR coated for Red and collimate to a better burning beam.

Larry

 

[Benm]

Its a curious phenomenon for sure. I use aixiz glass lenses mostly too, so far no probems with the LOC's.

Perhaps it has to do with how hard you drive them too - the few percent of light reflected back from a pane of glass could send it over the limit for COD, but reducing output power by 10% or so should completely solve that problem.

I would suspect something else is going on here though, and the death occurs at other than perfect 90 degree angles. Power might then be reflected somewhere near the emitting area where it causes the defect,

 

[ElektroFreak]

Its a curious phenomenon for sure. I use aixiz glass lenses mostly too, so far no probems with the LOC's.

Perhaps it has to do with how hard you drive them too - the few percent of light reflected back from a pane of glass could send it over the limit for COD, but reducing output power by 10% or so should completely solve that problem.

I would suspect something else is going on here though, and the death occurs at other than perfect 90 degree angles. Power might then be reflected somewhere near the emitting area where it causes the defect,

You may be right about the damage being dependent on how hard the diode is driven. I run them at 350mA, which is actually slightly less than most people.. still, that's over the rated current for these diodes.

 

[HIMNL9]

Well, thinking about geometries ..... if it was a reflection from the lens, must be a concave lens, but afaik, there's no concave lenses in front of the diodes in any focusing units, used in lasers .....

So, probably, is just a direct reflection from the focused (infinite) beam, that return all the power to the emitting face , that cause these damages, probably .... or in some cases, maybe, on the wires, also i personally think that is a bit difficult to melt a gold wire with a reflection, also if focused .....

..... ehm ..... except, maybe, if the gold wire is already carrying too much current and is overheated, so also a little increase of temperature can damage it ..... but you are not using your lasers at overspec and more high currents than tag ones, right ? ..... right ? .....

 

[Benm]

I think you'd be able to melt the gold bond wires IF you were to hit it with the full blast of the laser focussed exacty on one of them. This focussing could happen in practice if the reflected beam is slightly off. The focal distance on the way back will be the middle of the active area, but slightly shifted up/down/sideways, exactly where you'd expect a bond wire to be.

But would this work with the reflection from only a glass pane? Thats just a few % of light coming back in, yet seems to be guaranteed to kill them.

If you reflect all of the light back in, you'd expect the diode to die right then and there (of COD), but there must be something else going on here.

 

[HIMNL9]

Let me made a supposition (just for the fun )

We're basically using these diodes for, what ? ..... 200 % or more of the tag rated power ? ..... probably more, right ?

Now, basically, when a direct reflection happens, we can imagine something similar (the beams are ofcourse the same color, i just drawed them different for differentiate the paths) ..... the coming out light, focused from the lens (with still a little divergence), become reflected from an object (in the example, a mirror), that bounce it back through the lens, that focus it approximatively in the middle of the chip, or in any case, inside it (cause the beam is still having some divergence also if bounced, so the focusing point is a little bit more far than the emitting face of the chip)

Now, remembering that we're using these diodes already at 200% or 250% of their rated powers, i can imagine at least 2 good reasons for which the chip die, in this conditions:

1) The resonant face of the chip, just cannot handle the additional power, and fry (remember, if the diode is rated at 100mW continuous use, and you use it at 250 mW, the face is already working at 250% of the rated value, managing 150% more power of the one for which is made, and if the reflex add also just other 150 mW, it become 400% of the rated value ..... and i guess that almost anything, handling an overload of 300% more of the rated value, first or after die (and laser chips are just extremely fast and sensitive )

2) also in the case that the resonant face resist, the chip itself in this conditions is managing the same overload, that can rise the structural temperature inside the chip over the limits where the chip can work , so it just collapse, going in short circuit in fractions of secons, and melting all the rest(remember that the reflected power is focalized inside the chip ..... where is possible that the focus point of it, for short chips, is rear them, causing less damages, and this can also justify the more "resistance" of the short chips).

And yes, also the wires can be melted for an off-axis focalization of the beam, as Benm said ..... after all, those wires are already working in overload, and being overheated from the high power you're using, so also if the gold is usually a good reflective material and heat conductor, in these conditions is easy to damage it, too.

 

[Krutz]

backreflections from plain glass is max 8% (4% from each surface). funny that 8% make such a difference? sure, the reflection will probably enter the cavity, and maybe enhance the lasing.. but I would have thought you would have to be within 8% of COD to begin with?

thanks for the destructive test, electrofreak!

manuel