Theory of Operation
 MPE-X

The point source diffraction principles discussed for the MPE apply here also to the MPE-X.  Simply using finer scales and employing the MPE geometry is a possibility.  However, doing so results in decreased alignment tolerances.  In order to improve these tolerances it is necessary to reduce the size of the beam at the scale.  We can do this by moving the source point closer to the scale.  This is achieved by using the image of a laser diode instead of the diode itself.  This image can be placed arbitrarily close to the scale.

Our incident beam again impinges on the scale at an angle; and like the MPE-B, that central angle is out of the plane of diffraction.  The beam lies in the yz plane.  The drawing below schematically depicts the situation. It is an end view where we are looking in the -x direction.  The measurement direction is "into" and "out of" the screen.

The imaged laser spot is located directly on the Ronchi grating which insures that our required geometry is realized.  In the region of the the laser spot there are no grating lines.  As before, light which is positively and negatively diffracted from the scale interferes to form fringes on the target plane (Ronchi grating) which have exactly the same line spacing as the scale. There are of course other fringes and DC light which complicate useful photo detection.

We see that the light returning from the scale is again brought to a focus on the "Image Plane".  There we have an image of the original focused laser spot.  Diffraction from the scale, though, generates pseudo-spots which are displaced to and fro in x on the image plane.  The Ronchi grating has the effect of mixing these two spots, one with the other, providing interference.  The spatial filter located on the image plane passes only these two spots.  The beams are allowed to re-expand, but now all DC light is rejected.  Fringe contrast here is very high, unlike the MPE.  The unwanted effects of scale contaminants and the inherent surplus DC light are eliminated.

 

          Photo Detection:           

    Photo detection is accomplished using a 4-element
    linear array detector.  The "pitch" angle
    of the internal hologram is adjusted so that one fringe spans
    the array.  Of the elements A, B, C, D; A and C are
    subtracted to form one quadrature signal, and B and D are subtracted
    to form the other. This provides even more DC rejection. 





     Alignment Tolerances:     



    
This geometry is particularly insensitive
    to yaw.  Yaw effects are related to beam size at the
    scale, which have been reduced, but in the plane of diffraction
    our incidence angle is zero, and the signal response becomes
    an even function of yaw, centered on zero.


    
As for pitch, the beam size is made smaller
    by locating the source point close to the scale, and by orienting
    the laser diode with its smallest divergence in the vertical
    direction, y.  This orientation happens to also
    maximize our diffraction overlap (and signal) in the x
    direction.
 







    
The MPE-X typically incorporates a 2 micron
    scale, and therefore a 1 micron signal period.  The
    sine and cosine analog outputs very strongly reject scale contaminants.  As
    with the MPE, the path length differences between the interfering
    beams is zero so that wavelength and atmospheric effects are
    eliminated.  Also, the MPE-X has the same z
    axis indifference.  Pitch and yaw alignment tolerances
    are very high, especially given the resolution.


    
The device incorporates the same electronics
    as the MPE, so we have selectable resolutions. The finest quadrature
    resolution is 2.5 nanometer where the stability is even lower.  High
    speed outputs are also available.


    
The MPE-X is truly a superb linear measurement
    device.

          HOME