[spearce]

I spent all day yesterday getting physical alignment into what I felt was the best possible position. Spent 30 minutes today with Guy's optical grain focus technique, not to mention 30 minutes or so getting my red gun assembly mechanically aligned to the green.


Only to find out while trying to do keystone adjustment that keystone correction is not possible, because I'm not square to the screen! Arggh!!


Fortunately, I floor mount, but how do you guys do this? I measured from lower screen corner to front same side and rear opposite side corners, and made sure my measurements matched. (So left screen to right rear projector equaled right screen to left rear projector.)


I tried two methods yesterday to get them to match: first with a measuring tape, and second with identical lengths of fishing line. Apparently both methods failed me here....

 

[deronmoped]

Put up a test pattern and move the PJ to the best position for the Keystone.


Deron.

 

[spearce]

Yea, I started that one after posting. Its on a small table, so I literally (in a fit of rage) started kicking the projector around. I finally got a reasonable image on screen.


Only to find out while trying to do vertical centerline skew/bow adjustments that my screen top and screen bottom center markers were horribly skewed - by as much as 6 inches - from each other. Try setting centerline skew when you have no frame of reference.... ouch!


So I just spend the better part of the afternoon rechecking my screen markers, and have now discovered that the projector is very much in the wrong position, which is why I was having trouble. Time to kick it around some more.

 

[Lifter]

It's tough. Before anything, make sure there is nothing obstructing the lens mounting. If there's a piece of dirt in there it will throw it off.


The measurement technique you used could have problems. For one, if your measuring the corners of the screen to the projector casing, you need to realize that the tubes and the lenses are rarely set perfectly inside the case. More importantly, if the distances from the PJ to both corners are the same, like you said, and the center of the green phosphor is not perpendicular (horizontally) to the center of the screen, then it is definately not square. You need to make sure the center of the G phosphor is centered and perpendicular before you can square it up.


Here's what I would do. Since it's on the floor it should be a bit easier. First make sure the screen is flat against the wall, and that the wall itself is flat and parallel to the screen surface where the wall meets the floor. If you have baseboards or something this could throw you off, but not always. Use a plumb bob to mark the floor where the center of the screen is, and use a large squaring ruler (very large ones at home depot for cheap) to draw a straight line towards the PJ that is perpendicular to the screen surface. Slide the projector L/R to make this line run straight into the center of the G lens.


Once that's done, just make sure you center a crosshair pattern on the exact center of the green phosphor (also do this w/ R anf B at some point). You'll have to take the lenses off to do this. Peering through the lens to center the raster works when looking at the edges, but there's no way you can perfectly center the crosshair w/o taking the lenses off.


With the green crosshair centered on the phosphor, and the lens back on, you now have a straight "laser". Center the cross on the screen by rotating (not sliding) the PJ. Everything should be squared up now.


You might want to do a rough linearity adjustment while the lenses are off. Should make things easier. If your linearity is way off, and your using a large area of the phosphor, you will most likely have to do this or else your image area will be off the edge of the phosphor which will cause damage. Also, your keystone limitations could possibly be worse from having your linearity off.

 

[spearce]

Ok, this is getting fun.


I appear to be fully square to the screen. Everything is at mid position.


I expand the raster width using the eletronic controls to fill my screen. This has the effect that the machine is widening the right (onscreen anyway) side of the image area, while relatively leaving the left alone. Now my external test pattern's vertical center line is in the wrong position on screen, its too far to the left.


So I move it to the right using either the h-phase control or the raster shift controls. This causes the left side to fall short, while the right side hits the border.


This is a BD808, so I don't have a horiz. linearity control.... ideas ?

 

[spearce]

Just to add, the machine is at the calculated throw distance, maybe a tad closer, so I'm using the h-size control to increase phosphor usage. At a size that fills the screen I'm using a safe amount for the image area. I guess I can push the machine back a few inches to fill the screen...