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

Thread: Miniature flexure mirror mounts

  1. #1
    Join Date
    Oct 2005
    Location
    Melbourne, Australia
    Posts
    678

    Default Miniature flexure mirror mounts

    Any one have any idea where I could find some miniature flexure type mirror mounts, similar to those used in arctos lasers ?

    I've seen the surplus ones that mi-lasers sell, but they only appear to be adjustable in the Y axis......
    Now proudly stocking and offering the best deals on laser-wave

    www.lasershowparts.com
    http://stores.ebay.com.au/Lasershow-Parts

  2. #2
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Make them. Works for me. Full X/Y adjust with two small nuts and bolts, a bit of brass angle, a bit of brass plate, a brass pin, and a spring. Once you get your design right, pay for a bulk order to be made by a machine shop that can do really small stuff.

  3. #3
    Join Date
    Oct 2005
    Location
    Melbourne, Australia
    Posts
    678

    Default

    got any pics ?

    We'd definetly consider shelling out hte cash on doing up a run of these if people were interested.. as we want them ourselves...
    Now proudly stocking and offering the best deals on laser-wave

    www.lasershowparts.com
    http://stores.ebay.com.au/Lasershow-Parts

  4. #4
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Not yet, still waiting for screws. Only done the small brass angle and plate so far, 4 plates drilled with 1mm in each corner ready to adapt to whatever screws I can get to best suit purpose.

    Got a partial SketchUp model though, of the whole mount (two minor variants on it), so you can look in that to see the scale and method. Bear in mind, I use one metre for one millimetre. Was important in old versions, and I never lost the habit. A demo version of SketchUp Pro can export to DXF for other programs if you want to adapt it to a machinist's tooling that way.

    These will need adapting, mine are made for a specific beam cube size and design context, but I'll be using a very similar method for small dichros or mirrors or anything else. The mount still lacks one detail, one corner is empty, but will have two recesses on opposing surfaces for the spring. To be sure that the mount stays stable against shocks, this spring might not best be spring steel, a small silicone rubber bead might be better. Experimentation is still needed, this is as far as I've got so far until I have the screws I need. Small recesses for the nuts will be wanted too, to get them out of the way. This seems crude, but it's better than threading the holes. Easier, faster, cheaper, and it won't put stress on things when adjustments are made.

    It will be easy to get made up in bulk though, and probably as cheap as getting a few small printed circuit boards made, if you can translate that model (modified as needed) to whatever format the machinist wants.

    Not the neat answer it could have been, but it will work. I can make these by hand, so a good machine shop can turn them out easily to high accuracy.
    Last edited by The_Doctor; 05-05-2007 at 16:27.

  5. #5
    Join Date
    Oct 2005
    Location
    Melbourne, Australia
    Posts
    678

    Default

    Nice... but how are they flexure mounts. They look/sound like kinematics...
    Now proudly stocking and offering the best deals on laser-wave

    www.lasershowparts.com
    http://stores.ebay.com.au/Lasershow-Parts

  6. #6
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    I left that point as second, as you wanted X/Y (as I do). If flexure means a one piece construction with the screws forcing the opening on a kind of fixed hinge formed by the shape of the one-part mount, then the adaptation needs to be made with two saw cuts, so the 'hinge' is reduced to a region in one corner. By the time that is formed, there isn't much difference between the type of mount, so doing it this way makes sense, with separate parts. I considered the one-part method, but it involves threading two VERY small blind holes, and if you screw that up, you have to remake the entire mount. Failure rates in manufacturing could be higher, and the machining methods are already far more demanding than doing it as I describe it. Mine also gives you far better choices for controlling the degree of spring tension, and of renewing damaged threads if a wrong choice is made, and it's field repairable, too.
    Last edited by The_Doctor; 05-06-2007 at 04:22.

  7. #7
    Join Date
    Dec 2006
    Location
    denver,co
    Posts
    1,078

    Default

    I was thinking something along these lines. The flex mount itself is 1/2 x 1/2. The beam splitter is 1cm3. The adjust screws are 6/32 and there is a 4/40 screw on the bottom of the back to mount.


    Chad
    Attached Thumbnails Attached Thumbnails flex mount asm.jpg  

    Stress-Plot1.jpg  

    Last edited by chad; 05-06-2007 at 14:04.


    When the going gets weird, the weird turn pro.


  8. #8
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    Too high tech for me. Actual data. Wow!

  9. #9
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by chad View Post
    I was thinking something along these lines. The flex mount itself is 1/2 x 1/2. The beam splitter is 1cm3. The adjust screws are 6/32 and there is a 4/40 screw on the bottom of the back to mount.
    That's the one-part method I was thinking of, though in a small brass or aluminium block that had two saw cuts to form the flexing panels. Like I said, it has two blind holes to tap, and to get a decent count of turns per inch, you'll need either a very small size tap, resulting in a fragile thread, maybe too fragile to maintain the force that will be constantly applied to get the shift you want, or you'll need a special thread that costs more to get. Either way, tapping it will be harder because it must go right through one layer before being stopped by the next. By the time you have enough separation to make this easy, you run into another, more serious problem...

    That flexure mount (still a form of kinematic, btw, I looked that up) has three axes of significant translational (offset) error. The only way to reduce that is to make the milled separation between each layer so narrow that threading the blind holes will be a seriously tough challenge. In that mount, milled from a single block, there is no way to adjust the laser position, or that of the base of the mount, so when you flex the mount, the cube will shift because it's not rotating around its own centre. That means that it will never be at the true intersection, so alignment will never be very close to ideal. Also, it could oscillate if physically shocked, though that might not be much of a problem.

    Mine's cruder on the face of it, but it gets two axes of movement through a point that is a very small distance from the cube centre, less than half, so the offset error is reduced in all three axes by more than half the error in the one-piece flexure. If the pin is replaced by a threaded hole (the only need for threading, in a standard small size, no blind holes to worry about), the cross-beam offset can be reduced to zero with a careful use of three adjustment screws. The apparent lack of stiffness in my method isn't that much of a problem, the damping and stability of position is entirely regulated between the pin (or third screw's tip) and the bead of compressible material in the opposite corner. The one-piece mount allows no choice of materials, so it's not possible to optimise it for a given situation once it's formed.

    Last but not least, I didn't need SolidWorks to tell me any of that.

  10. #10
    Join Date
    Dec 2006
    Location
    denver,co
    Posts
    1,078

    Default

    All very true points. I am not going to argue them. I just am simply going to explain why I think it is a good choice for what I am working on and why the off sets will not be much of a problem for my application.

    OK first I am not expecting to get 8 degrees worth of adjustment out of it. In my application I am expecting a very small amount of adjustment. Probably less than one degree, more than likely in the arc-minutes range. I can simply offset the cube to make up for that other axis of translation before I UV cure the epoxy.

    Second, I am relying on machining and the rest of the housing to do 99% of the alignment, the flex mount is just to provide for the very small fine tuning.

    Third, My goal was to make a ~400 MW red in as small of a package as possible, with as little adjustment as possible, and leave room for the driver. I also wanted to keep it as stable as I can. If you have seen the post on the duel red you can see what I am after. I am not saying it is the best approach, or that it will even work. I just have been looking at the way people have been combining reds and was trying to come up with a more compact, stable, and elegant approach.

    I have the ability to machine all of this (if I can find the time ) and was trying to keep the pieces down to a minimum. My logic was a monolithic block of 6061 with the collimator holders and the flexed adjustment and heat sink and driver enclosure in a 2"x2"x1" space seemed like a good idea.



    Btw I am not trying to get into a pissing contest about whose "mount" was more ridgid I simply had an idea and thought I would try it out.

    Chad

    P.s. I didn't need SolidWorks to tell me that either, But I own it, use it daily and find it to be a great tool to work out ideas. After that I can open the SolidWorks file in my cam program generate nc code then cut it on one of the cnc machines.
    Last edited by chad; 05-09-2007 at 11:42.


    When the going gets weird, the weird turn pro.


Posting Permissions

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