Diode Quantification

Dear all,
I propose we build up here, in this very thread (and/or the Wiki!:)), a detailed knowledge base of red and blue laser diode information.
A lot of the information already exists, soaked all over PL; but a giga-post scooping it all up and standardising it, I feel would benefit everyone.

Initially I plan on putting under the microscope a variety of diodes to accurately measure: emitter dimensions, bond wire diameter, and count.
I believe a rigorous approach to this should shed light in the long term on tracking laser diode technology changes; hopefully complimenting previous "empirical" approaches with regard to collimation and the likes.
Each diode will be photographed, de-canned and then put under the microscope again for further inspection. Images of each will be attached here along with any laser diode data and a datasheet for completeness.
It is the aim to constantly update this first post; as to avoid threads of 100's of posts with the information getting burried.

Initially diodes will not be tested on their output performance; this can follow at a later stage with a standard, and rigorous testing procedure.


In short: The aim is to create a long term, one-stop comprehensive page of useful information for the community. :)


I have tested the following diodes:

ML501P73 - Datasheet
Emitter size: 98x2µm
No. of bond wires: 2

Images:
http://danielbriggs.co.uk/laser/diod...P73_Side5x.jpg http://danielbriggs.co.uk/laser/diod...73_Side20x.jpg http://danielbriggs.co.uk/laser/diod.../P73_Facet.jpg http://danielbriggs.co.uk/laser/diod...bs/P73_End.jpg



ML520G71 - Datasheet
Emitter size: 64x2µm
No. of bond wires: 1

Images:
http://danielbriggs.co.uk/laser/diod...71_5x_Side.jpg http://danielbriggs.co.uk/laser/diod...1_10x_Side.jpg http://danielbriggs.co.uk/laser/diod...1_20x_Side.jpg http://danielbriggs.co.uk/laser/diod...71_10x_Top.jpg http://danielbriggs.co.uk/laser/diod...71_50x_Top.jpg


Francesco has kindly donated the following diode:
HL63133DG - Datasheet
Emitter size: 10µm wide. I could not measure the thickness accurately optically (<1µm)

Images:
http://danielbriggs.co.uk/laser/diod...3DG-5x-Top.png http://danielbriggs.co.uk/laser/diod...DG-5x-Side.png http://danielbriggs.co.uk/laser/diod...G-10x-Side.pnghttp://danielbriggs.co.uk/laser/diod...20x-Side_2.pnghttp://danielbriggs.co.uk/laser/diod...-Side-Dims.png


Planned to test:

AndyCon has kindly donated the following diodes:

• LPC836
• LPC826
• LPC815

Simon (p1t8ull) has kindly donated the following diodes:

• HL6545MG - Datasheet
• ML101J27 - Datasheet

Edison has promised to donate the following diode:

• HL6388MG - Datasheet

Thanks to all! :)

Should anyone else have any laser diodes; red, blue or violet, to add to the list and are happy to donate them to PL Community Research^TM, having them de-canned and put under the microscope, please shout out. :)

Regards,
Dan

Another thought which I've just discussed with Briggs is we get this info on a web site so we can put the info in sections and add lens and telescopic info. Rather than having a page a mile long.

So I'll get the info on my laser-man site

Great work boys

Rob

(..in frat-initiation-like 'chant'..): Wiii-keee! Wiii-keee! Wiii-keee! :p

nice werk.. :cool: will contribute, as / when I can.. if I can just figure out this cloning-thing... :rolleyes:
j

Hi guys, I have already started this in the wiki, all it needs is for more people to add stuff to it.

go for it.

http://www.photonlexicon.com/wiki/in...de_information

chad

Quote:

---

Originally Posted by chad

Hi guys, I have already started this in the wiki, all it needs is for more people to add stuff to it.

go for it.

http://www.photonlexicon.com/wiki/in...de_information

chad

---

i didnt even know that existed

Yeah, I know. Help spread the word.

Don't be afraid to put stuff in. It is backed up and can be restored. even if it is just rough data people will fine tune it.

chad

Hi all,
I have just took deliver of some ML501P73's and threw them under the microscope.
I've sized the emitter at 100x2μm not 40μm as previously thought. I am sure my callibration is correct; a standard 100μm slide measures as 100μm... (to within 3 s.f.) ;)

Will update the 1st thread later; have to dash out; so just an image dump for now.
I have some G71's as well to test next week as well. :)

<snip> added to first post </snip>

Good work chop glad I didn't rush into buying them

nice pics dan , i am keen to find out the emitter size of the 300mw version, roll on next week
andyf97 is stuggling with the beam on these this could be why 100um eek not what i wanted to hear

Quote:

---

Originally Posted by andy_con

Good work chop glad I didn't rush into buying them

---

yep, me too... too bad though

thanks for the pics dan, they are priceless!!

Hey Dan, this is a VERY good initiative! Thank you for putting your time into this PL Community Research^TM.
I'm happy to send you one Opnext HL63133 diode for testing.
The datasheet of this diode can be found here: HL63133DG.

All is not lost yet, I am tinkering with an idea that will work even with a large emitter.

Please don't do that. Do tell about your idea its fun to think about these things. Sometimes they even work.

I love the HL63133 diodes. I've tried a bunch and these are about the best. I have 4 and it very bright and very tight. Sure it doesn't give 4 watts but it is a great beam. In fact I'm going backwards now for beam quality instead of high power. i got the PL450B 80mw 450nm blue and the new 110mw single mode Mits diode. Now if only I could get a true green I think it might be an amazing beam quality and still be at least as bright as my old hene argon combo with PCAOM.

I had some HL63133DG, gave them away, can't remember who I gave them too.

De-canned a G71 and ran similar images. See post #1.
There are significant differences between the diodes.
2 vs 1 bond wire; 100x2µm vs 64x2µm sized emitters.
The die is much shorter on the G71. I will get exact dimensions of the G71 and P73 die lengths next week.

We need more diodes in this catalogue, I can't afford to test them all. :)
It would be good to get some blues to compare, especially the new ones with the older workhorses.

this thread is getting more and more valuable!!!
(where is that "cookie" picture?)

carry on sir briggs!!

I can send you a lot of dead lpc 815 and an lpc 836 :)

Quote:

---

Originally Posted by LaNeK779

I can send you a lot of dead lpc 815 and an lpc 836 :)

---

Thank you for the kind offer Sir. Chris, but Con has posted those off to test. :)

Daniel,

Just for accuracy, interested to know how you are establishing on the end of the chip exactly where the emission comes from.

You have there a flat stripe that on the G71 measuring 64..um but is the laser emitting from all that stripe or just some central part of it.

I am asking this because I expected the G71 to be around half a P73 . Anyway you could power one to emit a tiny bit of light when you do the photos so this will be perfectly measurable?

Will be also great to get over to you some of the newer 9mm and 5.6mm blue diodes as this is going to add a huge new learning experience for everyone..

Great photos too, well done.

For education, this thread will prove to be the most valuable ever.

Andy

A (low) powered test should be doable.
I cannot think of a good reason why it would not emit from the full width of the wafer edge; my understanding is the light will propagate through the entire waveguide. However; there is only one way to find out. >:)

be very careful when you do it so you don't blow your ccd. You might want to use a filter or varibale polarizer to make sure you don't overload the camera. I did try this and it did not go well.

It is also possible that pre Iop the emitter will behave differently in terms of from where the laser emits. I have heard that the multimode diodes can run single mode up toa point before they go modal.

Quote:

---

Originally Posted by danielbriggs

A (low) powered test should be doable.
I cannot think of a good reason why it would not emit from the full width of the wafer edge; my understanding is the light will propagate through the entire waveguide. However; there is only one way to find out. >:)

---

Indeed, while your going to the effort to measure them which is very appreciated, might be a good idea to run them and measure the actual emission width at the same time, the wafer width might not be that relevant if beam only comes out of 50%, I have no idea if it does or does not so it's worth to test this too especially on the larger emitter diodes. I guess nobody knows the answer to this for sure until its been tested.

Cheers

Andy

I have just checked, and there are some fantastic diagrams in: Thompson - "Physics of Semiconductor Laser Devices".
He confirms the light emits from the full width junction, as thought.

Quote:

---

Originally Posted by danielbriggs

ML501P73 -
Emitter size: 98x2µm
ML520G71
Emitter size: 64x2µm

---

Well, F*ck. I thought that they were supposed to be the same. I thought hat was discussed in another thread earlier :(

me too, i was even told this by a mitsubishi applications engineer. So glad i did not buy some hundreds.

Just a point so far, I only have a sample size of 1...
Unlikely they would differ, but just saying these were from one supplier. :)

Maybe just a coincidence, but the emitter size ratio is consistent with the rated power output ratio.

Quote:

---

Originally Posted by planters

Maybe just a coincidence, but the emitter size ratio is consistent with the rated power output ratio.

---

Nice observation.
I like it when things follow a rule. I wonder if output power could increase by the same factor as per the extra mileage on the G71's, keeping power density the same? We have an extra bond wire to play with, so you never know. ;)

Quote:

---

Nice observation.
I like it when things follow a rule. I wonder if output power could increase by the same factor as per the extra mileage on the G71's, keeping power density the same? We have an extra bond wire to play with, so you never know.

---

I think this is a high likelihood. Same semiconductor. Same wavelength. Same manufacturer and the only reason the conductors were not increased by the NECESSARY 50% is that it was easier to add a second bond wire than to spec a 25% thicker one. As I had speculated earlier on in the G71 thread, the 73 diode is not better or worse its just bigger. Now if the net divergence is less than 50% worse then it is better, but maybe not enough to justify the substantial increase in cost. I think the uncertainty about the ultimate power handling ability of the two diodes will probably follow your speculation. The question is now more whether the optics will see the same ratio.

1st post updated with HL63133DG information; thanks Francesco :)
10µm width emitter; thickness indeterminable by this optical method (<1µm). There is a chance I could put it in the SEM; although that would mean sputter coating it and obviously no more red photons :(

Why dont you add also the rated powers, max current and x/y divergence.

Very good thread! Possibly the most useful I've seen. Only thing I don't understand is why it tailed off over a year ago. :)

For those of us (me) who haven't learnt the technicalities of diode production, may I ask why this data is useful?

Do the wafer dimensions correlate to output beam specs? Do the number of bond wires just determine the amount to current these devices can handle?

Thanks
Keith

A lot of it is beyond me, but the big thing is just seeing the detail of things datasheets rarely even mention. I saw posts citing 30µm for G71 and 50µm for P73 Mitsubishi diodes, so it's a bit of a surprise seeing proof that the sizes are double that. Originally I'd expected about 100µm for a 500mW diode, so should not have been surprised, but I'd hoped the new figures had been right. As it is, planters and other's posts indicate that using the G71 may have several advantages over the P73. The main correlation is that the wider the emitter, the less likely we are to get easily tamed light and high beam power density, regardless of what optics and money we throw into it.

I used to swear by single-mode only, for that reason, but given the costs of stacking loads of diodes, mount, mirrors, etc, it pays to at least try some modestly priced multimodes with fairly narrow emitters. Real evidence like this beats al the other data I have seen. On the other hand, just seeing the stats people get with their best efforts counts for a lot anyway.

Thanks Doc. I wonder if the size of the emitter and the amount of emitter actually emitting are the same? It is strange that they are quoting double the size - although it's only people like Briggs that are actually measuring these things, so I guess the manufacturers can usually get away with it ;)

I used to think the actual emission would be smaller width, but apparently not. :) See earlier posts before mine... That bit interests me more than strangely. Maybe there is a claim for effective width at some strength, as if Gaussian, presumably you don't get full strength at the edges no matter what you do get there. I'm glad Dan IS measuring. :) Really helpful to see new data of a type the makers rarely show. Even if we don't grasp the import at once, can't beat good detailed images. They tend to stick in memory for years, long after words fade.

"Traditional" and single mode diodes ARE much smaller, the G71's and P73's are chunky-chappies. :)

I should get back to work on this.
The new SM+MM 520nm diodes, as well as the old, new borrowed and blue 445nm's should be worth doing as well.

Quote:

---

Originally Posted by danielbriggs

"Traditional" and single mode diodes ARE much smaller, the G71's and P73's are chunky-chappies. :)

I should get back to work on this.
The new SM+MM 520nm diodes, as well as the old, new borrowed and blue 445nm's should be worth doing as well.

---

Have you already tested an M140?

Quote:

---

Originally Posted by laserlackey

Have you already tested an M140?

---

No blue or green of any sort; only the reds above.

Dan, did you decapsulate more than one G71? Also, did you see any evidence of irregularity, asymmetry or lack of concentricity in the die's mounting with respect to the main axis of the diode body?

I'm not saying I've seen any (not got the means to do it), I'm trying to establish where the variation might be coming from that caused people to choose a mount that allows looser diode placement when using very short focal lengths in a lens.

Related question... Did anyone see the centre of the beam appear to rotate around some point as they adjusted the focussing thread of the lens barrel? I'm not talking about wobble, I'm asking if this was seen by anyone even in cases where teflon tape or other means was used to eliminate the wobble that otherwise exists in the thread tolerance.

Crow,
I didn't specifically measure that; it can be done in a snap if needed. Rather poorly both G71 and P73 were samples size of 1. IIRC, one of them cost ~£80 a diode, and I didn't fancy trashing loads of them, (nor could afford to!) at a time when they were close to being hen's teeth.
This is pure speculation, but I would say the diode would not be source of the misalignment, more the variation in lens mounting to the holder, resulting in a poor concentricity between the diode+lens, exasperated by the short focal length.

That said, should anyone have any diodes they want running through a similar mangle, I'd be happy to help. :)

Eventually I will have some for you. :) I will try to be gentle to them, but any deaths, and I shall donate their tiny bodies to Science.

I agree with you about the lens. That's basically what I was trying to gauge. I bought mounts with fixed concentricity, as it's far easier to be sure a new diode will have its die where the old one was. This will make it hard for me to set ideal focus according to much advice, but that's what I'll investigate. I already have some notions about setting reliable concentricity, assuming that the holes in the mounts were initially concentric, coaxial with each other. They seem to be, they were made with better skill and means than I could manage on my own efforts. So long as they are, that leaves the thread tolerance, and the lens fit to the barrel. I think I can solve both.

All that said, if anyone can offer good description of how errors in pointing accuracy and such appear when focussing their setups, that will give me plenty to think about. Bake my noodle, I dare you. >:)

I think it is more likely the lack of perfect alignment is coming from the lens. With these very short Fl lenses sitting in mounts that have a clearance up to 10um (they do) this could explain the need ti compensate this not to mention slight out of plane operation of these same optics.