Unified Optical Scanning Technology

Chapter 4.6.2 - Scanner Devices and Techniques: Double-Pass Architecture

4.6.2 Double-Pass Architecture

Double-pass operation allows reduction of the pupil relief distance and elimination of some beam expander optics. It functions similarly with over- or underfilled facets. One basic form is introduced in Section, shown schematically in Figure 1.4, and illustrated realistically in Figure 4.3. This method is one of two generic forms of double-pass operation. The second form, illustrated in Figure 4.21, is discussed

Fig. 4.21 Alternate double-pass configuration—edge input. Compare with Fig. 4.3 Input beam enters from point P, just outside the scan field of length L. All beams reside in plane that is oriented perpendicular to the rotating axis A. Numeric values represent a typical design case. From R.E. Hopkins and M.J. Buzowa, "Optics for laser scanning," in Laser Scanning and Recording, SPIE Milestone Series, Vol. 378 (1985) Reproduced by permission of the publisher.

below. In the first form, utilizing 'central' illumination, the input beam (Figs. 1.4 and 4.3) is directed in the vertical meridional plane of the lens assembly to arrive ideally centered in the scan direction. However, it is displaced angularly in the y-z plane (Fig. 1.4), skewed by an amount β/2, which is to be minimized.

There are two principal effects in this first illumination technique. By propagating nominally centered in the scan direction of the lens system, the beam arrives at the facet suffering essentially no off-axis aberration in that direction. Consequently, the remaining significant concerns are the cross-scan aberration, which is compounded by the skewed (fixed) input and scanned output beams, and the magnitude of the residual bow (cross-scan sag) resulting from the skewed traversal of the scanning beam through the lens. Here are challenging tasks for computational exercise to determine and minimize the aberration and bow [H&S].

A useful indication of the functional variations of scan bow is attained by starting with the relation

in which e is the bow displacement (sag) in units of the lens focal length f, α is the deviation angle of the skewed beam with respect to the axis (β/2 in Fig. 1.4), and Θ½ is half the full-field scan angle Θ (per Fig. 4.2).

For a rough estimate of the skew angle α‚ consider that the off-axis focal point in the image plane (po or pi in Fig. 1.4) is displaced from the axis by an amount equal to the beam height (in the y-direction) at the lens. Figure 1.4 reveals this as a conservative value for the illustrated f-number beam; approximately f/10. This allows easy insertion of a folding mirror per Figure 4.3. It only becomes challenging when the beam height (hence the displacement allotted to the focal point) is reduced to a few millimeters, burdening the shape and fit of the folding mirror.

When considered in this manner, and assigning the beam height H, then the deviation angle α may be represented as α sin-1 (H/f).This reduces Equation 4-24 to

eliminating the sinα factor and independent of the focal length. Evaluation at the typical beam height H = 5 mm and Θ½ = 30° (a 'large' scan angle into a flat -field lens) reveals a bow sag of e = 0.775 mm. At the 'small' scan angle of ±10° (with the same H), the sag reduces to e = 0.077 mm—1/10th of that at the ±30° scan angle! And, of course, e varies in proportion to H in Equation 4-24a.

If the residual bow is considered to be excessive, one may compensate this with a complementary bow generated by other means, for example, adding a refractive prism in the beam path between the scanner and lens [H&S,Kra] or altering the facet angles of the (initially prismatic) deflector to provide small pyramidal angles with respect to the rotating axis [Sch,H&S].

The alternate method for forming double pass, noted above in this section, generates no bow. However, it imposes off-axis aberration in the scan direction. Observation of Figure 4.21 reveals that double pass is achieved in this method by injecting the input beam toward the flat-field lens from a point P just outside the nominal scan field (of length L).The system is almost symmetric about the centerline and free of cross-scan error (in y-direction of, e.g., Fig. 3.7). This is valid so long as the input beam and the scanned output beam propagate in the same plane (e.g., the plane of the paper in Fig. 4.21) and be perpendicular to the rotating axis A of the prismatic polygon (or single mirror deflector of Fig. 1.8).

The off-axis aberration in the scan direction is initiated by the input beam (injected at point P) as it propagates through the most marginal portion of the lens. Because this beam is aberrated initially with some astigmatism and coma and remains invariant during subsequent reflection from the deflector, it injects this aberration function throughout the entire output scan process. This initial aberration is then com-pounded by (and not necessarily complemented by) any new variable off-axis aberration created during scan of the output beam through the lens. Here, again, is an opportunity for computational exercise, likely of comparable effort to that required for the first-described skew beam method.


Already a GlobalSpec user? Log in.

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.

Customize Your GlobalSpec Experience

Category: Dimensional and Profile Scanners
Privacy Policy

This is embarrasing...

An error occurred while processing the form. Please try again in a few minutes.