Page 4 of 6 FirstFirst 123456 LastLast
Results 31 to 40 of 54

Thread: Hello All, Laser Question/Help

  1. #31
    Join Date
    Nov 2019
    Location
    West Chester, PA
    Posts
    61

    Default

    I just want to confirm, the seller tech support said that I should take out the current galvos from their mount and put them on the other mount. Do I need to align after doing so? I guess I'm wondering if the galvos know their position after the assembly is rotated.

  2. #32
    Join Date
    Mar 2013
    Location
    Langhus - Norway
    Posts
    349

    Default

    If the galvos have the same diameter, they should fit the holder just fine. Just be careful not touching mirrors, or brake them when mounting.


    Espen
    __________________________________________________ __________

    More projects than time available.
    More projects started than finished.
    More money spent than earned.
    More failure than success.
    Just got to love lasers!

  3. #33
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Default

    Quote Originally Posted by jdavis7765 View Post
    I just want to confirm, the seller tech support said that I should take out the current galvos from their mount and put them on the other mount.
    It's likely that they suggested this for two reasons. First, it's easier - since the galvo mount often screws in from the bottom and it's hard to get at those screws without pulling more of the projector apart. And second, it's possible that the mount already had a shim between it and the baseplate to get the galvo mirrors to the proper height. If you removed the mount you could misplace the shim.

    Do I need to align after doing so?
    As a general rule, no you should not need to re-align the beam. However, you may need to tweak the rotation of the galvos in their mounts slightly.

    If the X galvo is turned too far back (too far clockwise as viewed from the rear of the projector), the mirror will not be at a 45 degree angle when the galvo is in the neutral or home position. This can cause part of the incident beam to miss the Y galvo mirror when it bounces off the X mirror. It will also shift the center position of the scanned image to the right on the wall. (If the X galvo is turned even further back the beam will begin to miss the edge of the X-galvo as well.)

    You can have the same problem with the Y galvo too; if it's rotated too far counter-clockwise it will be closer to vertical than the ideal 45 degree angle, and again the beam coming off the X galvo will "spill" or miss the edge of the Y mirror. (This also has the effect of moving the middle of the scanned image up on the wall.)

    Try loosening the mount for the X galvo and then gently turning the galvo counter-clockwise while the unit is powered on but not scanning anything. Try to get the X mirror to be as close as possible to a 45 degree angle when it is at the home position. Then snug up the mount and start projecting something. See if that doesn't fix the problem. (Be careful; we're talking about small movements here. Also, you may need to shift the galvo in or out - closer to or further away from the aperture - in order to keep the beam in the middle of the X mirror.)

    Then do the same thing with the Y galvo. Again, 45 degrees is what you are looking for. And the in-out movement of the scanner body in the mount is to adjust the mirror position so the beam is in the very center when everything is at home or neutral position.

    Once you've adjusted both scanners, run a few tests and verify that you don't have any beams spilling off the mirrors. If so, then great! Good job - you're done.

    If you still have some beam spill, then it's possible that the incident beam isn't perfectly centered on the X mirror. To correct that you would either need to shim the galvo mount or adjust the alignment of the incident beam. (And of the two, shimming is easier.) But we'll cross that bridge when we come to it. Chances are good that you won't need to worry about this.

    I guess I'm wondering if the galvos know their position after the assembly is rotated.
    Scanners only know their position relative to the body of the scanner. They don't know the orientation of the scanner body in the mount, nor do they have any way of knowing the orientation of the incident beam. So, yeah, this is something you need to correct for.

    Note that by rotating the entire body in the mount, you are basically adjusting where the middle of the pattern will land on the wall. Either up and down or left and right depending on which galvo you're messing with. Ideally both scanners should be positioned with the mirrors at a 45 degree angle. Being off by a few degrees isn't a big deal, and some people intentionally position the Y galvo so that the middle is shifted up a bit (helps to keep beams out of the audience), but if you get too far off you will start to see the beam spilling off the mirrors.

    Remember that the scanners do not always come to rest at the neutral or home position when power is removed. So you can get them close by positioning them with the power off, but you'll almost always need to tweak the alignment one more time with the power on but no pattern being displayed. (When you first power on the galvos they will snap to the home position and stay there until you start scanning something.) Once you have them where you want them, make sure the adjustment screw on the mount is good and snug and you won't ever need to mess with it again.

    Adam

    EDIT PS: Who plays the Ukulele?
    Last edited by buffo; 01-12-2020 at 02:52.

  4. #34
    Join Date
    Nov 2019
    Location
    West Chester, PA
    Posts
    61

    Default

    Quote Originally Posted by buffo View Post

    EDIT PS: Who plays the Ukulele?
    That would be my son, I have no musical ability whatsoever

    I just marked the current position of the galvos so I could keep them at the same relative location. It appears everything is operational 100 percent now. Pictures below are 8 degree scan angle @ 40kpps (according to laser screen). Not sure I am not sure if the ILDA test pattern that I would see the circle pulling away from the sides of the square is the one pictured. If not, let me know and I will try to post with correct test pattern. It appears the laser is back to its former glory but now I have more insight as to what I should not do with a laser
    Attached Thumbnails Attached Thumbnails 20200111_121619.jpg  

    20200111_121814.jpg  


  5. #35
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Default

    Quote Originally Posted by jdavis7765 View Post
    That would be my son, I have no musical ability whatsoever
    HAHA! Cool.

    I just marked the current position of the galvos so I could keep them at the same relative location. It appears everything is operational 100 percent now.
    Very good!

    Pictures below are 8 degree scan angle @ 40kpps (according to laser screen).
    The first picture is the Laser Media Test Pattern. And it looks very good. The second picture is not the ILDA test pattern, however.

    The ILDA test pattern looks like this:




    If you're curious about the test pattern and what's going on with that center circle, you may want to read my scanner tuning tutorial, as it explains both the ILDA test pattern and the Laser Media Test Pattern.
    ILDA also has a document that explains the ILDA test pattern fairly well.

    I'm very curious to see what the ILDA test pattern looks like at 40K and 8 degrees on these new scanners. If it looks even close to correct, that would be quite the feat considering how affordable these scanners are.

    Adam

  6. #36
    Join Date
    Nov 2019
    Location
    West Chester, PA
    Posts
    61

    Default

    Quote Originally Posted by buffo View Post
    HAHA! Cool.



    Very good!



    The first picture is the Laser Media Test Pattern. And it looks very good. The second picture is not the ILDA test pattern, however.

    The ILDA test pattern looks like this:




    If you're curious about the test pattern and what's going on with that center circle, you may want to read my scanner tuning tutorial, as it explains both the ILDA test pattern and the Laser Media Test Pattern.
    ILDA also has a document that explains the ILDA test pattern fairly well.

    I'm very curious to see what the ILDA test pattern looks like at 40K and 8 degrees on these new scanners. If it looks even close to correct, that would be quite the feat considering how affordable these scanners are.

    Adam
    Ahh, I do have that one I think, just no picture. I will post back later tonight with that pattern but I am pretty sure it looks okay, but I did not examine for all the tells the pattern is used for. I have looked through the scanner tutorial before, I think I got the .ild patterns from the forum post it is under.

  7. #37
    Join Date
    Nov 2019
    Location
    West Chester, PA
    Posts
    61

    Default

    One pic is at 40k and the other at 20k. I think the pic that ends in 52 is the 20k pic.
    Attached Thumbnails Attached Thumbnails 20200113_181846.jpg  

    20200113_182052.jpg  


  8. #38
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Default

    Quote Originally Posted by jdavis7765 View Post
    One pic is at 40k and the other at 20k. I think the pic that ends in 52 is the 20k pic.
    If you are sure that one of these pictures was taken at 40K, then there is a problem with the way Laserjuice is displaying the test pattern. Some software includes optimization code that will add points to a frame if it decides that the jump between a pair of points is too far apart. If Laserjuice includes code like this, it must be disabled in order to properly display the ILDA test pattern. This is because the ILDA test pattern has an element that is designed to be too difficult for the scanners to reproduce 100% accurately. If the software adjusts for this, it screws up the whole point of the test pattern.

    If you look closely at the frame file for the test pattern you can see that the center circle is actually made up of 12 points that are positioned completely outside the center square. The idea is that the galvos will be under constant, maximum acceleration while trying to get to those points, but since they're so far apart the scanners never even get close to the first point before they head off towards the next one. The result is the 12-sided polygon (dodecagon) gets "rounded off" to a much smaller circle that should land just inside the center square.

    The galvos actually race around those 12 points 3 times in a row, and the result should be a perfectly round circle that just barely touches the middle of the sides of the center square. This represents roughly 3dB distortion compared to the input signal (which is those points that lie outside the square in the original frame file). So yes, that part of the image is distorted, but it should be distorted in a predictable and measurable way.

    The problem, of course, is when the software tries to add extra points so that the galvos can actually make it all the way out to those points that are positioned well beyond the center square. That totally messes up the test. (Basically it's cheating the test by making it easier for the scanners to get all the way out to those points.) It looks like this is what is happening in your case.

    Normally if the scanners are properly tuned for 40K the circle will be inside the square like the picture I posted above when you have the scan speed set to 40K in software. If you then reduce the scan speed in software, you'll see that center circle start to grow in size while the rest of the pattern stays as it is. This is because points are being sent to the scanners at a slower rate, but the scanners are still running at the speed they were tuned for. This gives the scanners a little more time to get to the point they're headed for. As you reduce the scan speed even more, at some point the scanners will finally be able to arrive at those points, at which time the circle will stop getting bigger and will start to look like a dodecagon.

    However, in both of your pictures the circle already looks like a dodecagon. (You can see the corners.) That either means that the scanners are really operating at a much faster speed (like 60K or better, which is highly unlikely), or the software is adding extra points to the image to allow the scanners to actually make it all the way to the corners of the dodecagon.

    At this point I think you need to meet up with Brad and have him test your projector with a different controller and software combination. That way he can verify that any "frame optimization" is turned off.

    Adam

  9. #39
    Join Date
    Nov 2019
    Location
    West Chester, PA
    Posts
    61

    Default

    Quote Originally Posted by buffo View Post
    If you are sure that one of these pictures was taken at 40K, then there is a problem with the way Laserjuice is displaying the test pattern. Some software includes optimization code that will add points to a frame if it decides that the jump between a pair of points is too far apart. If Laserjuice includes code like this, it must be disabled in order to properly display the ILDA test pattern. This is because the ILDA test pattern has an element that is designed to be too difficult for the scanners to reproduce 100% accurately. If the software adjusts for this, it screws up the whole point of the test pattern.

    If you look closely at the frame file for the test pattern you can see that the center circle is actually made up of 12 points that are positioned completely outside the center square. The idea is that the galvos will be under constant, maximum acceleration while trying to get to those points, but since they're so far apart the scanners never even get close to the first point before they head off towards the next one. The result is the 12-sided polygon (dodecagon) gets "rounded off" to a much smaller circle that should land just inside the center square.

    The galvos actually race around those 12 points 3 times in a row, and the result should be a perfectly round circle that just barely touches the middle of the sides of the center square. This represents roughly 3dB distortion compared to the input signal (which is those points that lie outside the square in the original frame file). So yes, that part of the image is distorted, but it should be distorted in a predictable and measurable way.

    The problem, of course, is when the software tries to add extra points so that the galvos can actually make it all the way out to those points that are positioned well beyond the center square. That totally messes up the test. (Basically it's cheating the test by making it easier for the scanners to get all the way out to those points.) It looks like this is what is happening in your case.

    Normally if the scanners are properly tuned for 40K the circle will be inside the square like the picture I posted above when you have the scan speed set to 40K in software. If you then reduce the scan speed in software, you'll see that center circle start to grow in size while the rest of the pattern stays as it is. This is because points are being sent to the scanners at a slower rate, but the scanners are still running at the speed they were tuned for. This gives the scanners a little more time to get to the point they're headed for. As you reduce the scan speed even more, at some point the scanners will finally be able to arrive at those points, at which time the circle will stop getting bigger and will start to look like a dodecagon.

    However, in both of your pictures the circle already looks like a dodecagon. (You can see the corners.) That either means that the scanners are really operating at a much faster speed (like 60K or better, which is highly unlikely), or the software is adding extra points to the image to allow the scanners to actually make it all the way to the corners of the dodecagon.

    At this point I think you need to meet up with Brad and have him test your projector with a different controller and software combination. That way he can verify that any "frame optimization" is turned off.

    Adam
    I think that makes sense. For this test I just used the SD card and wrote the ildatest to it and displayed through there. I do have helios DAC and lasershowgen and could display through there disabling optimization, would that do the trick for this test to work properly?

  10. #40
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Default

    Quote Originally Posted by jdavis7765 View Post
    I do have helios DAC and lasershowgen and could display through there disabling optimization, would that do the trick for this test to work properly?
    I'm not familiar with the Lasershowgen software, but if you can disable optimization then yeah - that should work fine.

    Adam

Posting Permissions

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