Page 1 of 2 12 LastLast
Results 1 to 10 of 12

Thread: collimating diodes - methods?

  1. #1
    Join Date
    Sep 2010
    Posts
    4

    Default collimating diodes - methods?

    Hi all,

    i´m looking for another way to collimate my diodes. Normally i do it via a bouncing mirror 15m away, shining back to a sheet of white paper right in front of me. So, cranking the lens barrel in until I reach the sweet spot. Works all fine but its wobbly and not very refined imho.
    But I heard of a method using some kind of telescope lens and a digital camera with a filter (to save the ccd sensor), all positioned right in front of the diode. so , when you reach collimation you can watch the image of the emitter comfortably on a monitor. Anybody heard of such a method? This seems to be an elegant way without always having to point at this damn small mirror and trying to get the wobbling image clear while adjusting the collimator lens... I just can´t quite think of a setup that can do this in a reliable way.
    Any thoughts about this from you experts?

    Greetz from Bavaria

    Florian

  2. #2
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,169

    Default

    Ah, your describing a Spiricon LBA>..
    ~
    A 1/2" or 3/4" or 1" B&W CCD camera without a lens.. Basically a 12 bit BLACK AND WHITE video capture system. The CCD or CMOS sensor must be larger then the beam or this fails miserably. You need a CCD with a known chip part number, so you know the pixel size, which is how you know the beam size.
    ~
    AN OD6 ND filter or two and a rotary variable density filter are needed, as the laser diode beam profile is current dependent. Having only a fixed filter can make it difficult to define the true edge. Diodes are difficult to measure at low currents...They tend to lase around the middle only at low currents. You really want them up at 80 to 100% of full power.
    ~
    Take a picture, move the camera on a rail 100 cm, take a shot, let the software tell you the profile and the divergence and Gaussian or Top Hat Fit.
    ~
    I've been doing this for years with their hardware and a TM-6CN or TM-7EX camera with known pixel sizes and adjustable shutter speeds.
    ~
    https://land.ophiropt.com/beam_profi...-General-18-11
    ~
    YOU DO NOT PUT A LENS BETWEEN THE BEAM AND THE SENSOR! UNLESS YOU LIKE MEASURING THE LENS TRANSFORM FUNCTION INSTEAD of your BEAM PARAMETERS.
    COLOR CAMERAS ARE USELESS FOR THIS DUE TO EDGE FINDING ERRORS AND RAGGED INTENSITY PROFILES.
    ~
    The math is tough as you need to carefully determine the edge intensity curve to match the Gaussian profile definitions used internationally on what a beam is.
    ~
    Trying this with a tiny webcam sensor is fruitless when it comes to a true , accurate, measurement. It goes without saying that focusing the beam results in burned pixels!
    ~
    A Galvo, a ramp generator, a pinhole, a biased photodiode, a load resistor and a storage oscilloscope have been known to work too. (My old way). If you use a Galvo, you need a tangent error correction in some cases.
    Steve
    Last edited by mixedgas; 03-26-2019 at 14:09.
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  3. #3
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,169

    Default

    This was easy with gas lasers with stable profiles. Measuring diode lasers that jump in profiles every few seconds is not so easy.

    You need lots of expensive optical attenuators etc.

    While one of two of friends are importing lasers for resale and both of them needs these setups, I 'm not so sure this is for everyone. It gets expensive fast, and volume ND absorbers that can handle a 12 watt beam are not cheap.

    If you only want to check the edges of the beam and not the profile, a razor blade, a micrometer ranslation stage and a laser power meter plus some graph paper will get you there. Two translation stages a few meters apart will do it better.

    If your not aligning projectors on a daily or weekly basis, it is not worth the investment.

    My day job is being a optical technology professional, and it makes sense for me to have the ccd, linear translation, and rotating slit instruments.

    Steve
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  4. #4
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,169

    Default

    This was easy with gas lasers with stable profiles. Measuring diode lasers that jump in profiles or mode hop every few seconds is not so easy.

    You need lots of expensive optical attenuators etc.

    While one of two of friends are importing lasers for resale and both of them needs these setups, I 'm not so sure this is for everyone. It gets expensive fast, and volume ND absorbers that can handle a 12 watt beam are not cheap.

    If you only want to check the edges of the beam and not the profile, a razor blade, a micrometer ranslation stage and a laser power meter plus some graph paper will get you there. Two translation stages a few meters apart will do it better.

    If your not aligning projectors on a daily or weekly basis, it is not worth the investment.

    My day job is being a optical technology professional, and it makes sense for me to have the ccd, linear translation, and rotating slit instruments.

    Are you aligning beam shaping prisms on a daily basis? IF so, beam profiling is something you need. Otherwise I'm not so sure that you as should buy hardware instead of finding a long, desolate farm with a big barn to align on.

    Steve
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  5. #5
    Join Date
    Oct 2012
    Location
    Germany
    Posts
    1,049

    Default

    ... or two big mirrors and "multi-zig-zag-folding"?

    Viktor
    Aufruf zum Projekt "Müll-freie Meere" - https://reprap.org/forum/list.php?426
    Call for the project "garbage-free seas" - https://reprap.org/forum/list.php?425

  6. #6
    Join Date
    Sep 2010
    Posts
    4

    Default

    Hi Steve, thanks for the quick reply.

    I may have expressed myself a little vague. I am not after beam profiling or measuring intensities. I´m working with stacks of single emitter red diodes (mainly Ushio 170mW, Sharp 180mW in 638nm and Mitsu 836 in 658nm if mixed red is desired) as well as the 1W 520nm and 2W 445nm TO9 diodes. Each one in its own diode cup with a lens barrel in front. When I have for example 24 of them in a single plate (in 2 groups of 12 diodes) I go near the collimation point with each lens, pointing on a black wall. Then I switch to my mirror 15m away that bounces the beam back to a sheet of paper right in front of me. Then I finally collimate each lens of the array to the best image of the emitter. After all diodes are collimated it goes on with knife edging and so forth.
    As you can imagine, the beams do not exit the array all parallel, but they give a "shotgun pattern" 15m away, lets say 80x30cm. As my bouncing mirror is only 10x10cm, there´s a lot of pointing and its hard to keep the diode i`m just collimating onto the mirror. It works OK, but there should be a more elegant (and thus faster) way to accomplish perfect collimation. And yes, i do projector aligning and module building on a daily basis and for a living
    My Idea was to kind of simulate those 20 or 25m of space to be able to collimate the beam to infinity. And via some attenuators and a monitor, so I won´t get blind some day. Those green 1W diodes are some nasty bright bitches, even with my welder´s goggles on...

    Hallo Viktor,

    I actually tried the zig zag method some years ago, but my surface mirrors are not good (=flat) enough obviously. Every reflection deforms the beam just a little bit, and that adds up to an unusable, distorted image.

    To give you guys an idea what I do: This is a red module (never mind the shot diodes) I got back for repair. You see the arrangement of the diodes in the block. That´s only the half of a 6W 638nm module, but equipped only to 5 as per request of the customer. So, when i have to build a couple of these babys some time saving (and even better beam quality?) would be much appreciated.
    Attached Thumbnails Attached Thumbnails rot.jpg  


  7. #7
    Join Date
    Dec 2007
    Location
    Denmark
    Posts
    106

    Default

    Hi Florian.

    Don't know what is you max. distance. But we are so lucky i have 75 meter from our shop to a back wall at the other end, when we need to do alignment. :-)
    About the zig zag way, i have used also, and it works fine with 1 mirror, but as you say, with more mirrors is not so good.

    Interesting red builds. I'm also working on getting 10W Red for some 30W+ RGB laser we need red for.
    I got a Laser Source from one for a EU Manufactor, a 6W Red single mode red there was Pol so you are able to stack 2 modules to 12W with an external PBS + WP, and still keep the 6x6mm beam with 1x1mrad Pack. It was done with a a pattern of 6x6 diodes with 200mW single mode red diodes (problerly HL63263DG), but the divergence was not what they promised me. Instead of 1 mrad it was more then the double.
    So expect they had used a FL2 to be able to have 36 diodes (6x6 pcs) on the 6x6mm pack. They have a small beam at nearfield, but on the other hand the divergence is higher at farfield (not 1mrad). The source is shipped back for refund because of the spec was not ok as they promised.
    So it just to find a balance on divergence vs beam size, i have EMS8000 for 6.5mm aperture so a 6x6mm would be the best match, but it still have 1mrad is not so easy to get. Most manufactors can do and 8W singlemode module with 1mrad max. what a 6x6mm beam. But higher is a problem to get.
    Else plan B could be to do a 638nm and a deep red 658nm side by side. 658 is nice but 658 is not still 2-2,5 less visible Vs. the 638nm diodes.

    What have you achieved in in beam specs. ?
    Can see many use FL4 lens for there builds, but was thinking about using a FL4.5 - 4.6 to have some marken to keep 1 mrad. My plan was to use the Sharp 185mw, because it's cheap vs. power.

    Greeting
    Poul.


    Quote Originally Posted by messerfloh View Post
    Hi Steve, thanks for the quick reply.

    I may have expressed myself a little vague. I am not after beam profiling or measuring intensities. I´m working with stacks of single emitter red diodes (mainly Ushio 170mW, Sharp 180mW in 638nm and Mitsu 836 in 658nm if mixed red is desired) as well as the 1W 520nm and 2W 445nm TO9 diodes. Each one in its own diode cup with a lens barrel in front. When I have for example 24 of them in a single plate (in 2 groups of 12 diodes) I go near the collimation point with each lens, pointing on a black wall. Then I switch to my mirror 15m away that bounces the beam back to a sheet of paper right in front of me. Then I finally collimate each lens of the array to the best image of the emitter. After all diodes are collimated it goes on with knife edging and so forth.
    As you can imagine, the beams do not exit the array all parallel, but they give a "shotgun pattern" 15m away, lets say 80x30cm. As my bouncing mirror is only 10x10cm, there´s a lot of pointing and its hard to keep the diode i`m just collimating onto the mirror. It works OK, but there should be a more elegant (and thus faster) way to accomplish perfect collimation. And yes, i do projector aligning and module building on a daily basis and for a living
    My Idea was to kind of simulate those 20 or 25m of space to be able to collimate the beam to infinity. And via some attenuators and a monitor, so I won´t get blind some day. Those green 1W diodes are some nasty bright bitches, even with my welder´s goggles on...

    Hallo Viktor,

    I actually tried the zig zag method some years ago, but my surface mirrors are not good (=flat) enough obviously. Every reflection deforms the beam just a little bit, and that adds up to an unusable, distorted image.

    To give you guys an idea what I do: This is a red module (never mind the shot diodes) I got back for repair. You see the arrangement of the diodes in the block. That´s only the half of a 6W 638nm module, but equipped only to 5 as per request of the customer. So, when i have to build a couple of these babys some time saving (and even better beam quality?) would be much appreciated.

  8. #8
    Join Date
    Sep 2010
    Posts
    4

    Default

    Hello Poul,

    here I have some beam pics of a 6W / 36 diode red module like the one in the photo above, but fully equipped. 2x 18 diodes doubled by cube.
    I use the Sharp diodes with a 3mm FL lens.
    The beam is around 6x6mm, so it fits perfectly to the EMS 8000 with 6,5mm aperture
    Second pic is taken after 12m distance with full power, the beam has about 20x16mm here. This would calculate 1,16 / 0,83 mrad full angle. Hard to get a picture, on the mirror it gets best, even on a matte black surface it blooms up like hell
    Attached Thumbnails Attached Thumbnails 6W nach ATF.jpg  

    6W 12m.jpg  


  9. #9
    Join Date
    Oct 2012
    Location
    Germany
    Posts
    1,049

    Default

    ... some high-power diodes with diode-bars use interesting collimating methodes -- the unfocussed "collimated" spot on the ceiling was after roughly 1,5 Meters even smaller i one axis, than on the V-card nearby

    Viktor
    Attached Thumbnails Attached Thumbnails Rofin-200W-Dioden - 3A.jpg  

    Rofin-200W-Dioden - 3A-Karte.jpg  

    Rofin-200W-Dioden - 3A-Decke.jpg  

    Rofin-200W-Dioden 1 Offen.jpg  

    Q-Strahl Kollimiert 8A-Pappe.jpg  

    1 Spot.jpg  

    Aufruf zum Projekt "Müll-freie Meere" - https://reprap.org/forum/list.php?426
    Call for the project "garbage-free seas" - https://reprap.org/forum/list.php?425

  10. #10
    Join Date
    Dec 2007
    Location
    Nottingham, UK
    Posts
    2,793

    Default

    For collimation I'm a big fan of shear plate interferometers (sadly £££, available from Thorlabs).
    If you have low qty of diodes to align, it's not worth the money; stick to getting the smallest spot at the longest distance. However we had several hundred to align for a project, so was well worth it.

    I made a little test jig to hold the uncollimated modules, hooked up to an ITC4001, [you don't need anything that fancy] just something you can twiddle the current fractions of mA - this part is important.

    You're basically aiming for threshold + a few mA's. This is generally a good brightness of fringes and almost every single diode I've collimated has sufficient coherence length at this current region. Due to manufacturing differences, the optimal current for collimation fringes will not be the same exact figure every diode.

    In reality: clamp a diode in, connect to test jig, spin the lens to get parallel fringes, blob of UV curing adhesive + cure [or low strength threadlock] to secure the lens, and unclamp easily takes no more than 60 seconds once you get the hang of it.
    Be very critical on getting the fringes perfect. "S" shaped fringes mean your lens is off-axis / non-centred. Not always a problem for >=4mm FL's but significantly more of an issues on super short ~2mm FL lenses [I hate those!]
    Also watch out for glue shrinkage if securing the lenses, as that can change the collimation a noticeable amount as it cures.

    This setup takes about 1/4 of my small optical breadboard. We'd check collimation was done right by "going long". But this was only used as a sanity check, and not for actual collimation / to infinity!

    All the best,
    Dan


    Attachment 55258
    Last edited by danielbriggs; 03-28-2019 at 05:33.
    - There is no such word as "can't" -
    - 60% of the time it works every time -

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •