Announcement

Collapse
No announcement yet.

New interferometer setup

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Yeah, very very low power, the most powerful one is 330 microwatts which is below the threshold to cause eye damage. A supermarket scanner has a laser three times as powerful as this.

    I have to take a yearly laser safety test at work because thats the kind of machines we build. Some of our machines use lasers up to 600W. Though I dont work with lasers as much as I used to now that I design and build test fixtures for the boards and modules we make for our machines. I used to do laser installations, alignments, and stuff like that. Im currently in the middle of designing tests for this: https://esi.com/products/pcb-processing/high-density-interconnect/geode/

    Comment


    • #32
      Got the renishaw head working tonight. Set up a 29ls32 line receiver and dumped that into one of these: https://www.usdigital.com/products/a...interfaces/QSB

      Seems to work fine and verified the pin that I thought was resolution select, is. So with basic linear interferometer optics it gives you a resolution of 80nm in low res mode, 10nm in high res mode. With a plane mirror setup that drops to 40/5nm. A plane mirror setup double the beam path which doubles the resolution.

      There is a piece of software a couple guys wrote called uMD1, hopefully I will be able to get this to work with that, otherwise I will have to write an interface in labview.

      Comment


      • #33
        Took everything over to a friend house where he has a granite slab with a linear motor stage on it and set up both lasers side by side to calibrate the Renishaw. Still need to do the math and see if I got useful numbers out of it.

        Untitled by Jerry Biehler, on Flickr

        Comment


        • #34
          if i think about it i dont understand how resolution under 1/2 wavelenght is achieved.

          Comment


          • #35
            You get x4 just by the interferometer setup using a 2 pass setup and a plane mirror. From there you do it electronically with interpolation. The signal that comes from the sensor is analog so you can easily digitize it and get higher resolution. Same exact method as used when dealing with sine-cosine encoders.

            LIGO is a variation of this basic setup and can measure distance changes 1/10000th of the diameter of a proton. It’s also 4km long so it might be hard to carry around though.

            Comment


            • #36
              Laser light is electromagnetic radiation, just like radio waves. And just like radio waves, they are sinusoidal. When two sine waves of the same frequency are allowed to interfere with each other they can combine either directly in phase (0 degrees), directly out of phase (180 degrees) OR at ANY phase angle between. If the same light source (laser) is used by splitting it's beam and sending the two parts of it's light over different paths, then any changes in the length of either of those paths will show up as a phase change.

              A zero degree relationship between them gives you a maximum or a bright fringe while 180 degrees gives a null or no light. And each phase angle between them gives a different brightness value which can be measured and converted to the actual phase angle. So brightness = phase angle and phase angle = distance as measured in a percentage of a half wavelength. For longer distances you have to count the peaks and nulls.

              This is the electronic equivalent of counting the interference bands when a machined and polished surface is compared to an optical flat with monochromatic light. You can not only see full (bright) or half (dark) bands, but all points between them (shades of gray).

              It's analog and digital together.



              Originally posted by dian View Post
              if i think about it i dont understand how resolution under 1/2 wavelenght is achieved.
              Paul A.
              SE Texas

              Make it fit.
              You can't win and there IS a penalty for trying!

              Comment


              • #37
                Yeah thats about it, this describes a homodyne laser where it just outputs a single frequency. With the Agilent/HP/Keysight laser it is a heterodyne and they have a tube that emits two separate wavelengths slightly apart from each other, in the case of the one I have, about 3.7Mhz. You get a reference signal from measuring the beat frequency of the outgoing beam of the laser and then you measure the beat frequency of the received beam and you compare, its the doppler effect. This way you can measure distance and velocity. Zygo lasers used a much higher frequency, 20MHz.

                And it turns out the data I collected yesterday was all garbage. I didn't realize the USB encoder interface has a 6MHz limit and I went past that and lost counts. So I am rigging up a 4x36" piece of aluminum with the interferometer optics at either end and a linear guide in between. A double sided mirror will go on the truck and I can put the lasers on either end and compare that way, I should be able to just set up on the garage floor. Once I make space. Its been just too hot out there to do anything. If not once the universities open up Ill bug my professor friend to let me use one of his optical tables for a bit.

                Comment


                • #38
                  Is this project just for fun, or do you need these systems to perform at published specs?

                  Comment


                  • #39
                    Mostly for fun, also its nice to do things like check ball screws. One thing I want to do is check the actual accuracy of the harmonic drive on my telescope and see if it is actually any good.

                    The Agilent unit works fine, the lifetime wavelength accuracy of the laser is supposed to be .1PPM. The Renishaw, not so much. The PC adapter is available in various flavors, PCMCIA, USB, GPIB, but they are all unobtanium. Thats why I am trying to get the quadrature output working. Its kind of a backup for the HP unit but its something I have to do now because if for some reason I break the HP I have no way to calibrate the Renishaw.

                    Here is a paper on all the ways I can screw up the measurements: http://literature.cdn.keysight.com/l.../5952-7973.pdf

                    Comment


                    • #40
                      Originally posted by macona View Post
                      Mostly for fun, also its nice to do things like check ball screws. One thing I want to do is check the actual accuracy of the harmonic drive on my telescope and see if it is actually any good.

                      The Agilent unit works fine, the lifetime wavelength accuracy of the laser is supposed to be .1PPM. The Renishaw, not so much. The PC adapter is available in various flavors, PCMCIA, USB, GPIB, but they are all unobtanium. Thats why I am trying to get the quadrature output working. Its kind of a backup for the HP unit but its something I have to do now because if for some reason I break the HP I have no way to calibrate the Renishaw.
                      It likely won't matter for your intended use, but you should take the 0.1ppm wavelength accuracy with a grain of salt.

                      Comment


                      • #41
                        Yeah, thats about a .02um possible error in 200mm over the life of the laser. And thats not even guaranteed, thats worst case. Not worried about it.

                        Comment

                        Working...
                        X