Christie D4K25, Руководство по обслуживанию

Страница: 43 / 138

Section 3: Service Setups
D4K25 Service Manual 3-13
020-100880-01 Rev. 1 (06-2012)
3.4.4 Final Convergence Assessment
When should you be satisfied with your adjustments? Read the following section before coming to a
conclusion.
1. On the M25 Prisms, the Red image is slightly larger than the Blue and Green. The larger size is caused by
the natural properties of Red light in that it is a “slow” wavelength compared to Blue and Green. The angle
that light is traveling is affected by every medium it travels through, and in this case there are several
pieces of glass between Red and the screen – the Red image is affected most noticeably as it ends up
slightly larger.
a. This means that the best possible convergence would be described as follows: The center area
highlighted in Figure 3-13 should be perfectly aligned showing solid white pixels, as shown in the
simplified diagram. ( Figure 3-13 )
b. The Red should be evenly 1/4 to 1/2 a pixel out all the way around the outer area not highlighted in
Figure 3-13 , showing cyan (green+blue) towards the center of the screen, as shown in the simplified
diagram. ( Figure 3-13 )
c. Green and Blue should always be perfectly aligned to each-other (to within ¼ pixel total in one or a
combination of directions.)
Figure 3-13 Best Possible Convergence
2. In certain circumstances, there may be further tolerance deficiencies in the prism assembly in combination
with the projection lens that will cause the normal Red over-sizing to become slightly non-uniform.
a. In practice, this extra increase in size has generally appeared on the lower-right of the image, causing
an apparent clockwise twist in the lower-right corner only – the upper left, right and lower-left all look
OK.
b. The only way to deal with this will be to “Split the difference” pushing the rest of the image slightly
twisted counter-clockwise to minimize the effect on the bottom of the screen, keeping the subtitle zone
as converged as possible.
c. If the problem appears to be more than an extra 1/2 of a pixel out of uniformity, a new lens/prism
assembly may be the only solution.