Page 1 of 4 1234 LastLast
Results 1 to 10 of 34

Thread: Rgb spatial filter. One blade pinhole method

  1. #1
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default Rgb spatial filter. One blade pinhole method

    Hi all!!

    After reading LOTS of posts about spatial filtering I decided to buy some achromatics and 100µm pinhole to play with.
    The purpose: Filtering RGB alltogether…yes, I know it is not the best way to go, but anyway wanna try.

    I disposed the lens and started to positioning pinhole… but because my RGB was delivering to much power (3w), my RGB beam drilled a new hole near the 100µm!! by error, I ended with a good customised pinhole!

    So, in order to establish a best-affordable method (not buying 100µm pinhole), I removed it and started with a “new” razor-blade method for RGB (Thanks Mr Planters!!!)

    The method:

    I extracted razor blades from ordinary gillette-machine. These are really thin blades, so easy to drill by focused laser beam (just 2,5sec @4W). I suppose these are stainless-steel, sorry, but I don’t have tried any other material yet, have to try.
    Click image for larger version. 

Name:	00c blades.JPG 
Views:	85 
Size:	31.4 KB 
ID:	44836
    Click image for larger version. 

Name:	08 razor blade.JPG 
Views:	106 
Size:	60.4 KB 
ID:	44844

    So, when beams are perfectly FF aligned, I put the achromat telescope and focused to infinite.

    Low Down RGB power (to avoid drills when adjusting blade). Then disposed the blade on diagonal (could be vertical too), and started to find the blade-edge-on-focus far-field.

    Far-field: When the blade edge is near focus and near beam core it appears as a shadow. So the trick, is to adjust to have the blade-edge shadow as focused as possible (sharper blade-edge image)
    Then…move the blade until it covers the beam core, and re-adjust holder until it is also focused on the other side.
    Click image for larger version. 

Name:	00B razor blade on left.JPG 
Views:	74 
Size:	130.3 KB 
ID:	44837

    This is important because, if not perfectly perpendicular (so focused along all focal plane), we obtain interference circles and so on when the blade is drilled by the beam.

    The idea is to have the blade-edge focused along the focal plane, that is, the blade-plane is perfectly perpendicular to the lens beam axis.

    So gently moving blade along the focal plane to ensure the blade-edge is focused on the left (beam core at right) and to the right (beam core covered by blade).
    Then, fix the holder and test the above again, to check if blade is still well on focus.

    Move the blade covering the beam-core, glue it with thermal bi component adhesive and let it cure. Put the power to the max!! After 2-3 seconds the beam drills de blade, making a customized pinhole!

    no filtering:
    Click image for larger version. 

Name:	01 nofiltering.JPG 
Views:	84 
Size:	51.6 KB 
ID:	44838
    FILTEREDthe wing is due the objective cam, the spot is really nice)
    Click image for larger version. 

Name:	03 filtered.JPG 
Views:	86 
Size:	37.8 KB 
ID:	44839

    Yes, this is a ‘dirty’ system, but the results are astonishing and we get a quasi-round-gaussian pinhole perfectly matched to the our beam core geometry, cause only the core have the power to drill the blade.

    However, the cost is -20% power, but I think it worths cause we have a near-perfect gaussian RGB beam, an all three beams inside same geometry.
    POWER WIHOUT SPATIAL:
    Click image for larger version. 

Name:	06 power wihout spatial.JPG 
Views:	100 
Size:	108.1 KB 
ID:	44840
    POWER WITH SPATIAL:
    Click image for larger version. 

Name:	07 power with spatial.JPG 
Views:	97 
Size:	105.6 KB 
ID:	44841

    Click image for larger version. 

Name:	04 setup.JPG 
Views:	110 
Size:	61.0 KB 
ID:	44842
    Click image for larger version. 

Name:	05 setup.JPG 
Views:	89 
Size:	82.4 KB 
ID:	44843

    Jordi (SORRY MY ENGLISH)
    Last edited by jors; 01-04-2017 at 23:56.

  2. #2
    Join Date
    Nov 2008
    Location
    Cleveland Ohio
    Posts
    2,599

    Default

    can you list the lens you used.

  3. #3
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    yes, sorry for the missing info. These are Thorlabs lenses:

    http://www.thorlabs.de/newgrouppage9...ctgroup_id=120

    I acquired several EFLs (all positives) in order to test some magnifications, but always on a footprint relationship no more then 80mm lens to lens, cause this spatial is going to be inside micro-projector.
    In the above setup, its EFL 30 (AC254*030*A) + EFL 45 (AC254*045*A) for a 1.5X on 75mm (aprox) footprint.

    Cause the footprint is relative small, its high difficult to correctly center a pinhole, so the "drilling" method its really an advantage here, really fast to adjust, no expensive centering setups and so on. Also, the beam-core is 'always' drilling for its geometry, so if rgb is on a club, for instance, and there is slight misalignment, new geometry hole is achieved, so we assure the beam core is always perfectly shaped and going thru the pinhole.
    IF misalign is high, however, we should remove blade, re-align beams FF, and start with a new blade.
    However, cause the blue is powerful and smaller dot, the red (I) is slightly clipped by the hole, so we loose -20% overall power as I said, but we have all three beams inside the same geometry (not divergence), and this is nice for graphics. Its really important to have all 3 beams perfectly aligned FF before drill
    Last edited by jors; 09-25-2014 at 00:31.

  4. #4
    Join Date
    Dec 2007
    Location
    Stockholm, Sweden
    Posts
    551

    Default

    Thank you, nice explanation of a smart method!

  5. #5
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    thank you! just want to share crazy discoveries

  6. #6
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default


  7. #7
    Join Date
    Apr 2010
    Location
    Grand Rapids, Mi
    Posts
    2,538

    Default

    Any idea on the thickness of the average disposable blade? I tried one on my red only but it was two thick for the red only. I have spatial filters on red and blue. I want to go buy some shim stock to try but don't know what size to try. Locally I can go down to .001 inch.

    - - - Updated - - -

    Quote Originally Posted by dar303 View Post
    Thank you, nice explanation of a smart method!
    And what he said. Thanks
    leading in trailing technology

  8. #8
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    you're welcome!! thanks
    Mixedgas has very kindly offered to measure pinhole diameter and also the blade thickness with electronic microscope. One drilled blade sample is flying to him (Thanks Steve!!). So hope to get some numbers soon. I think the pinhole is about 50-70µm.
    The blade is really thin, I can bend it with just my fingers pressure, also cut it with ordinary scissors.
    Click image for larger version. 

Name:	blade01.jpg 
Views:	28 
Size:	31.5 KB 
ID:	44846
    Click image for larger version. 

Name:	blade02.JPG 
Views:	32 
Size:	33.3 KB 
ID:	44847
    Yes, sure that filtering each colour is the way to go, but could be nice to have spatial filter inside micro-projector, and IMHO this RGB filter works when not enough room inside. Also, on drilling with combined rgb beam, we have in just 3sec all colours inside same geometry! (reds clipped).
    This 'instant' all-3-beams inside the same geometry is IMO the main advantage with this technique, also we save lenses and setups with a really small footprint. Of course, each colour-divergence is another question. The combined spot is really very accurate and nice quasi-round. Because the temperature is higher in the center beam-core and gradually decreases to the edges, the drilled hole is an exact representation of this, very very close to round.

    I think 1-1,5W red is not enough to drill the hole, but this is only an assumption, not tested, so for sure multiple blades/core-shaping Planters' technique is really elegant and suitable for single colour, but also on rgb too with achromatics.
    Next step is scanning the pinholed RGB beam, to find for chromatic aberrations, hope not.
    Last edited by jors; 09-26-2014 at 01:24.

  9. #9
    Join Date
    May 2014
    Location
    Barcelona, Spain
    Posts
    439

    Default

    and by the way, I also burned my Iphone sensor while shooting beam pics!!!
    Click image for larger version. 

Name:	FullSizeRender.jpg 
Views:	42 
Size:	1.53 MB 
ID:	44848
    Last edited by jors; 09-26-2014 at 01:28.

  10. #10
    Join Date
    Dec 2010
    Location
    DC/VA metro area, USA
    Posts
    554

    Default

    This is pretty neat! Thanks for the procedure and illustrations!

    I wonder if you defocus the blue slightly if you would get a slightly larger hole for the red to pass through, and if you did, you would start to see some red junk in the FF due to the red beam geometry or if it would be OK. I guess it would be hard to change the blue focus without messing up the NF alignment?

Posting Permissions

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