Mitsubishi ML520G71...Red Holy Grail or Flashlight Fail ??

[CDBEAM]

Gentlemen.....stop and think about it.....all of us...in the " Laser Hobby " area....have really moved our hobby forward......many, many have contributed. And...we will continue to progress !!! Things have changed....in a fundamental way in our hobby.

No longer must we buy our blue lasers…we make our own….if we overdrive the LD...and it goes COD….put in a new diode(s)…..and dial it down some !!!The Cost/mW/Risk balance is greatly improved. Life is good.
Now….Red is very close to joining the ranks of blue !!! A huge thanx to all who have contributed ….and continue to forge ahead. When red is conquered….we will tackle green…..either with DI….or by some other method !!

One goal is to continue to drive the cost of the hobby down…..the more reasonable….the more may participative in our hobby….as I say….The Quantum Well is deep…Photons for all !!!! Another goal is to share our knowledge, experimental results….and overall enjoyment in each other’s success….and learn from others …”unsuccessful experiments “. In a way….our aggregate efforts….are a lab….and what we lack in efficiency….is compensated by enthusiasm !!

Please …continue to travel and evolve within our little corner….as like minded individuals….from ALL around the world join in !!! Thank you ! BEAM

[edison]

You can argu about overdriving them or not but one thing is for sure, if i sell a quad 637nm single mode diode module with lets say 900mw about 99% of my customers ask how many diodes is use. If i say 4 the second question is, are you overdiving them? So in my point of few you can,t over drive them unless the customer knows about it and get the garantee that the diodes will be replaced when they died. If i would take the risk of selling overdriven modules then i would loose money before i even start with 637,s. The replacement of the diodes would cost me a fortune and also i would loose good customers that rely on good quality modules. Modules are like tools, that,s why a carpenter buys and expencive makita because it never fails. That is how you should look to a projector. Its a tool that laserist need to do a show. So overdriving diodes is not really a good idea.But as a hobbyist you can get away with that if you want to stay a hobbyist. As soon you doing payed shows you need reliable equipment.

But nevertheless kiyoukan i,m very impressed how you found a way to collimate the diodes. I,m still not sure if its useable for graphics. I would like to have the cilindrical telescope to play with so when doing a groupbuy count me in. I have 4 dave lenses so hopefully there will be a groupbuy soon.

[steve-o]

CD- well said :]
Edison- I completely agree with if your selling modules to customers you want them to be as reliable as possible. Warranty repairs and company reputation is not something to take lightly with a business.
For experimenters such as some of us, though, over-driving these diodes (such as is done with the 445s) could be beneficial if a mttf (life-test) was established at different current levels. However, this data would be valid only if the manufacturer were to never change the manufacturing processes for these diodes for a specified time-period, and they sent an ECN to all of us if they were changed (which ain't gonna happen )
My point being: Overdriving is good for R&D; e.g. finding the acceptible operating limits of this LD, and if you're prepared for acceptible losses. Overdriving is not good (as Edison stated) for important shows and finished sold products.

[CDBEAM]

...This question will likely ramble on....almost like the " Tube vs.Solid State " debate that has continued for the last 60+ years in the audiophile community !!! See the attached array.....yes....it is a gross over simplification.....but....it starts to put the many variables into some order for consideration. Anyone....and there are many....with at least an order of magnitude more knowledge on LD properties than I....please make suggestions to modify this chart !!

It is a complicated issue....and the main watershed is " Hobbist " vs " Professional "...Thanx....BEAM

[Solarfire]

yes much closer.
It takes some time to setup to make sure every thing is perfect.
You can chose where to knife edge you could go for a larger part to lessen the dust problem.
yes dust is the enemy of this setup.
That why its going to be covered in an enclosure.
More details and pics to come this week or next now that i can work.

Have you actually tried this? I don’t want to sound pessimistic, but no matter where in the beam path you double the beam width via knife edging, the beam will be twice as wide. For example, if a cylinder lens magnifies a 1 micron focal point to 4mm then it will magnify a 2 micron focal point to 8mm. Not to mention Simons point, which is also a reason for using the Plano concave/Plano convex combination instead, also keeps aberrations down due to the smaller angle of incidence on the lens.

[kiyoukan]

I will start another thread with pictures and more details on the dual build.
i built it and taken it apart 3 times now.
also here is a link some of you may not want to read.
http://laserpointerforums.com/f50/de...age-71241.html
http://laserpointerforums.com/f49/cu...ses-71286.html
So hopefully monday i will have everything back together again to post.
The setup is rather large, 150mm length by about 127mm wide.

[steve-o]

here is a link some of you may not want to read.

LOL


____________

[kiyoukan]

Okay so a pain in the ass moment.
The diode i de canned is now near death.
I did some inspection of the diode and it seems i damaged about 60% of the right side of the emitter.
I noticed the beam had was lacking wings and the even power it had before so i took a look and sure enough its damaged.
I have on on its way from DTR he said he had them in stock so it should be here by Friday.
But i have all the lenses in place and am loving it so far.
Cant wait to sell all my old pointer stuff and so i can order some cool new things.

[steve-o]

.. from another (hi-jacked) thread

Badger wrote:
hey steve-o a dual mitsu 300mw 638 diodes cranked up to 500mw each ,and they been at 2 days cw test, and 36hrs scanning
and all good ,roll on the 50mw diodes

Is this the 520G71 badger? And you have had some luck taming it.. using lava or cyl lenses or not .. pix? ..

[kiyoukan]

So i have gotten 1W of 635nm.
Or well very close.
But i have ran into a road block.

The problem is aligning that beam on such small surface is freaking driving me crazy.
Even after hours of adjustments the best beam i got was 6mm wide
Far from the 4mm i was willing to accept.
I cant even begin to tell you the feather adjustments i was making were to large sometimes.
So i have kind given up and knife edging inside the correction optics.
But i still have a solution.
Stack vertically before the correction optics. this will get you the beam of 4mm by 4mm.
So the question now is how do we do a quad.
Well 2 dual setups with a pbs cube and waveplate.
Now you can try to save money by using the cube and plate inside the correction optics so they share the same large c lens.
and the fact that they will overlap will make the beam just the same.
But if you do this make sure to do it with the beams crossing at a bigger point.
in my setups you see them crossing at 10mm away from the first lens i would set it up about 30mm that way it would be much easier to align to the cube with a beam of a decent size.
So a dual build is still possible with a very similar price tag, a quad is still possible with the added waveplate and cube.
A 8 build would not work.
Now you could try to cut cost even more in the quad via stacking then using the wave plate and sending the other stack threw the same 2 c lenses however there is a limited amount of space between the first c lens and the diode.
If you go beyond 30mm the beam will come out larger than 4.5mm so you would need to be able to fit the waveplate and cube and knife stack all before that c lens.
It might not be worth the cost saving for the as it will make it even harder to align.

[badger1666]

hey steve-o
mysetup is as earlier in this thread just tweeked abit to 1.14w for the pair, just over 700ma each i think , will check tomorrow

.. from another (hi-jacked) thread



Is this the 520G71 badger? And you have had some luck taming it.. using lava or cyl lenses or not .. pix? ..

[dilbert]

I spent a day at the last uklem rebuilding my red with different combinations of lenses etc and finally got the perfect 1mm x 4mm beam to match my blue.. Fantastic divergence too, perfect match, biggest and most important change was switching from 405g2's to brass inserts holding the 2mm lenses. Easy to align and PBS tiny beam too. You could knife edge very easy with this lens.

[steve-o]

I've been fiddling w/ the lavalens setup today, getting 5x5mm at the ap (which is ok for me) and a very reasonable 430mw output at just over 800mA with no issues as of yet but the dot on the tree looks huge (see pic). I'll have to go measure the size of it and figure the divergence in a bit.
Thanks badger, I'll try to go find the page. This thread (40 pages) is absolutely out-of-control lol.
Dilbert, sounds like youve got a good handle on it .. care to share some pix?
638 w lavalenses.jpg

[kiyoukan]

steve what is your power before optics?
you should be getting alot more power at 800ma.
Unless you lenses are dirty or you got a low % diode.
I know mine at 749ma are getting just around 500mw raw.
I tested mine again after cleaning the mirror off and got just about 1.08w.
So that mirror can knock off a good bit of power but that was known and now with my new setup of stacking prior to correction optics i should be able to get the same power as im not changing anything but the placement.
And with the mirror going to be knife edging about 10mm from the diode lens the mirror should not be as effected by dust.
I hope to get a stable 1W red at current of 750ma with it TE cooled to 15C/ 60f.

[steve-o]

Yeah all my optics seem to be dust-magnets :] I don't have a mirror (?) I have only the laser diode, colimation lens and the Lava lenses in this test-setup. I'll measure the raw output right after pizza.. (it's done) ..

[taggalucci]

with my current 3.3fl lens i got a beam of just under 4mm at aperture and a divergence around 1.1

Was this uncorrected? I'd be interested to know what the uncorrected beam specs are for both axes in the near and far field.

I have new lenses from dave and 2 other company's on the way with FL from 2 to 2.25 to 2.5 and my current setup was done with a 3.3FL lens

So i will have enough lenses to do alot of testing.

Would you be happy to post your findings on these lenses? It would be good to know the near and far field beam specs for both axes of each lens (with power readings too if possible).

Stack vertically before the correction optics. this will get you the beam of 4mm by 4mm.
So the question now is how do we do a quad.

What was the stacking height difference you've used, and was it with the 3.3FL lenses?

A 8 build would not work.

I'd be interested to see how tight 8 diodes could be knife-edged/stacked (depending on diode orientation) and how bad the divergence would be without correction but with some de-focussing of the collimators to square up the beam. Might be OK for large scanners, aerial laser projectors...

[steve-o]

steve what is your power before optics?

I'm getting 661mW raw output from the diode w/my (diy) meter. I'm sure that the lens and barrel that I'm using are chopping the edges off for a significant power loss and making for a slightly better beam profile.

[kiyoukan]

i will try to answers those above questions when i have the time sry.
So you are getting 661mw raw and what about after diode lens?
You should not be dropping below 500mw unless one of your lenses is very bad.

[steve-o]

I'm having the best luck with a aixiz triple element red ar coated. It's 471mW after the lens. I'm losing 190 mW. Yes, it's definitely not the best lens in the world. At this point, I don't care. I'm ready to start building a red to go w/ the green/ blue. I need to see an RGB SOON or elst me start looosing me mind ............... ........................... ..AAhhhh-hahahahahah ..

[kiyoukan]

So you are losing 28.7% on the diode lens.
you are losing about 8.8% on correction optics.
you are losing 37.5% total. err 35%
You need to find a better diode lens.
Also how clean are your correction optics?
That loss on the diode lens could be due to back reflections to the diode and that is bad, not just power loss but will shorten the life of the diode.

[steve-o]

Ok, I got some sleep. Re-editing post.
I'll take your word for the percentages k.
I do need a better collimation lens, I have experimented with all the ones that I have, and am holding my breath in anticipation for new findings here. I have posted my findings a few pages back.
My optics are clean. The lava lenses are brand new, fresh out of the box. What you are seeing in my pic are a few specs of dust, highly amplified by the camera and the laser light.
Back reflections. They can kill a diode, not just shorten the life (I speak from experience ;] I will check the aixis' AR coating .. I doubt that they are good quality, don't know for sure.
To sum it up, I believe that the losses are coming from the long FL lens and the small diameter of the barrel, chopping off part of the beam.
Thanks for your concerns and I appreciate the advise

[steve-o]

My reason for this approach is simple. Price and expedience. I currently have (3) G71 diodes, (3) col. lenses (3) retainers /housings, etc.
If I were to start construction right now I would be able to have 1.3 Watts of red in a couple weeks. The other option is to wait .. hmmm .. I think I'll take the losses and start building now, rather than keep experimenting. It's good enough. No doubt you guys will find a setup with >90% throughput, eagerly awaiting that, but I gotta build something .. now .. lol ..

[Solarfire]

For what it’s worth, I’m going with Dave’s lens a PBS, wave plate and lava cylinder correction. As Simon suggested a while back, vertical stacking is the way to go, imo this is probably the best way to go with minimal optics for a good quad. In the setup below all beam paths are equal length and 4mm x 4mm @ the aperture and 1-1.2mrad is doable.

[kiyoukan]

While i think lava optics are kinda nice they are a bit small, and hard to work with.
The larger lenses are more forgiving. You can be a little bit off and not get any power loss.
If you were to be 1mm of with lava's correction lenses you would get a good amount of loss.
The fact that you got 8% on fresh lenses might be the sign they are not in position perfectly.
The last set of lava optics was cd beams and i think he tested the power loss of around 4% but he stated that took several hours to get that low.
So tonight im going to have my setup redone with the vertical stacking.
Solarfire, i know for my correction optics at least that distance from the diode lenses is very important.
I found around 30mm to be the sweet spot.
Is this the same with lavas?
If so what distance is that in your setup?
As much as i hate the aixiz housings they were nice precision forward and back movement, so it allowed me to quickly adjust the distance to make sure the spot size was the same.
Once my setup is done im going to see if any one wants to get in on getting more of these correction optics made.
This is the price quote from the manufacturer.

EFL(mm)_______L(mm)____________H (mm)______unit price (USD)
60_____________32_______________30__________25
10_____________20_______________10__________15
So the price is about the same as drlavas however with much more forgiveness as far as positioning.
I am looking to order some and i know a few other members want them.
They are very simple and have a working distance of 62mm between the 2 lenses, so for a beam under 1 mrad you would want about from the diode lens 30mm then each lens is around 6mm thick so an extra 12mm plus the 62mm between them gives you a total of 104mm.
Its slightly large but it does have it benefits as far as use and less power loss.
Right now the coating they have gets a tiny bit over 1% for 635nm but is less than 1% for most other colors.
Seeing as we only need these for red i could change the coating to have around 0.48% loss for 635nm

[logsquared]

I have been working with these diodes a lot over the past few months. Although I think I may have been the first to suggest moving the telescope away from the collimator to get different magnifications I don't think its the best sollution. Most of my builds and testing have been with 4mm lenses. So, I haven't tested much with Daves 2mm lens. Here is what I've found through practice with 4mm lenses..... Moving the telescope further from the collimator gives the ability to focus the spot through a larger range of sizes than if the first lens is very close to the collimator. However, there is only one point where the spot is truly in focus. Trying to reduce the spot size by slightly de-focusing the telescope works to make the spot smaller but doesn't change the aperture much. So, you can get lower divergence but not the smallest aperture possible. I have found that putting the first tele lens as close the the collimator as possible and making the tele ratio what the math says is should be gives the best quality spot and smallest aperture size. So, if you have 8mRad with your collimator and want 1mRad corrected use a 8X tele and put the lenses as close as possible.

Because the divergence is so great with a 2mm lens the actual results will probably be slightly different than the math would indicate. Unfortunately, I have not had the chance to test with Dave's lenses much. But my guess is the will be the best for quads with minimum optics, just needing a really long telescope.