Projector: "Ampro 2000"

Hi, I want to update my stuff to HD, I tried 720P with a PS3, It came out but with the convergence all mixed up, and I find my picture with regular 480p a bit blur :(

When I try to do the convergence... it's never perfect, it never fits EXACTLY, so does someone already passed by there and know how to do a perfect job. The problem seems to be that every time I stretch one side... the value the other side cannot go more than the max :/ to resume, the overal picture is a bit blur.

You guys have experience, do you have any tips ?
And does anyone know how to put the value into a memory and switch between them ? Regular <=> HD

thx for your time.

Whenever you use a new source you need to redo convergence completely, starting from scratch. I'm not really familiar with the Ampro's, so I'll give you a general idea how you should do convergence. It's best to use an external test pattern from a DVD or HTPC.


First, you must make sure that the mechanical setup is good. This guide was written for a NEC PG, but much of it should apply. http://www.curtpalme.com/NECPG_MechSetup1.shtm. You basically want to make sure that the raster is centered and that the projector is dead center and level with the screen. You will need to somehow mark the screen to achieve this. If the mechanical setup is done improperly, you'll never get good convergence.

Then you can do convergence. First you should display a crosshair pattern and use vertical raster position to center the blue and red onto the green (the 3 colors should matchup horizontally since you physically moved the guns). Use as little static convergence as possible. Next you can do keystone to get a square image. Then use the linear and amplitude controls to align everything at the edges with a crosshatch pattern. You will have to go back and forth between the controls until you have a good balance. Then you can do any keystone for the red and blue specifically and any other geometry controls you need. This should get you pretty close. The main idea is that you shouldn't have to use excessive adjustment on any control...if you do then your mechanical setup is likely wrong.

If the source you are converging can be used to generate a test pattern (such as the DVD playing a test disc) then you use it. If your ultimate source is a video processor with your other sources plugged into it, use the test patterns built into the processor (use them at the output resolution you watch). Absent all this, use the internal test patterns of the projector at the frequency closest to what you will watch.

First, depending upon the age of the 2000, the video input board may or may not handle tri-level sync that HD requires. I had two 4000G's and one of them would do HD and the other would not. They are the same as 2000G's but with bigger tubes.
Second, what internal patterns are you using? If you are just pressing test, that is wrong. You need to press 4 then test to bring up a genlocked test pattern.
However, AmPro's internal test patterns can be crap and be inconsistant so, Marks suggestion of inputting test patterns from the source is the best option for you.
Third, there are many things generic to all projectors that can lead to a soft picture. Correct mounting distance, optical focus and electronic focus to name a few. Some things in your case which are projector/tube spacific are old tubes that have lost their ability to focus or mis-adjusted static focus pots and tube fungus.
The first thing you should do at this point is check the tubes. Put up any crosshatch test pattern (the one you use for convergence), turn down the contrast to a level that allows you to look directly into the lenses without hurting your eyes. Examine each tube face. Look at the patterns. If it's sharp on all three, you have mechanical focus problems. If the pattern is fuzzy overall on the tube face, chances are the static focus pots need to be adjusted or the tubes have lost their ability to focus. If you notice while examining the tubes that parts of the pattern are sharp and others are fuzzy, you have tube fungus.

Depending on the age of your projector, the focus pots are mounted on the high voltage power supply or on a splitter block mounted on the belly of the projector in the center near the green tube. Caution: Thousands of volts go through that area of the projector. never use two hands inside of a running projector. So adjust static focus at your own risk.

Thats about all the information I should give you without knowing your level of electronics savy.

Chip

Whenever you use a new source you need to redo convergence completely, starting from scratch. I'm not really familiar with the Ampro's, so I'll give you a general idea how you should do convergence. It's best to use an external test pattern from a DVD or HTPC.

Really? Crap I didn't know that. So that means anytime I hook up a different game console, PC, or DVD player I'll have to redo convergence? Is there some way to get around this? Preferably relatively cheap.

Really? Crap I didn't know that. So that means anytime I hook up a different game console, PC, or DVD player I'll have to redo convergence? Is there some way to get around this? Preferably relatively cheap.
Well you only need to do convergence the initial time you set up each source. You should be saving each setup under a different memory block that way you can switch between sources using the input list menu.

So say you use an xbox 360 and ps3. You would create a memory block for the ps3 and xbox 360 with different convergence settings on each. Raster centering/size and other service functions should remain the same, but convergence/geometry settings will be different. Just make a default block with all the service settings configured but with all convergence and focus settings zeroed. Then whenever you want to create a new memory block, copy the data from the default block and do convergence.

It sounds like you are runniing out of convergence adjustment, which seems to indicate that your physical setup is poor.

Wow! Thank you very much, I really apreciate your awnsers. :D
Thanks for your sources, now at least I know where to start!

You guys are true hackers!

As far as actually doing the convergence adjustments, I'm not familiar with Ampros but if the interface to the adjustments is digital here is the method I use.

It took me a while to stumble on this, so if its common knowledge and the way everyone does it - please forgive me. When I started trying to do convergence I would look at the whole screen and try to set the values and iterate until it looked good. But it often seemed like I would end up chasing my tail and going back and forth between various adjustments and never get it to be perfect.

What I do now is focus on each adjustment on just one part of the screen at a time. Do each one just considering the horizontal or vertical components. First right and then left or top and bottom. Look at the number that gets one edge perfect, then do the opposite side and find that number. Split the difference and then do the other adjustment that compensates for it. After a few iterations I generally find that I end up at the right values that look right for both sides without pulling my hair out.

The adjustments work in pairs: Size and Linearity. Skew and Bow. Keystone and Pincushion. Each one in a pair will compensate for the other.

Starting with size and linearity, find the Size value that is perfect for the left side. Say its 110. Then do the right and find its perfect value, say 150. Split the difference and set it to 130. Then do the Linearity finding the value that is perfect for the left say 98 and then do the right say 122. Split the difference and set it to 110. Then go back and do the Size again - usually it will be much closer to correct for both sides and the difference between the ideal number on each side gets smaller and smaller. After a couple iterations its usually the same number. After doing the left/right do the top/bottom.

If the difference between two values is an odd number I usually bias it one way or the other, and often a single click iisn'tthat big a deal.

After doing the Size/Linearity, and Skew/Bow I usually go back and make sure the center is still good since sometimes those adjustments can knock it slightly off. Then after doing the Keystone/Pincushion on one rough pass go back and check the center, then Size/Lin, and then Skew/Bow before doing a couple more iterations on Keystone/Pin since they can throw the others off if they were way out.

Hopefully this description makes sense. The idea is to just look at each side of each adjustment in isolation, find the number that makes it perfect, then do the same on the other side. Split the difference between the two values and then do the same thing with the corresponding adjustment. Even if they start off way off, usually after a couple passes they converge (heh) on the same numbers.

But like the others have said, the convergence sits on top of the foundation setup by the mechanical aim, lens flapping, optical focus and EM focus. You need to have the projector setup perfectly square and centered to the screen, have the guns aimed precisely and optically focused tight. You cant fix a rough physical setup with a fine convergence.

"The adjustments work in pairs: Size and Linearity. Skew and Bow. Keystone and Pincushion. Each one in a pair will compensate for the other."

^Keep this in mind. It took me awhile to figure out this relationship

I just really cant stress how important it is that the mechanical setup/mount be dead on. To give you a sense of why, if the projector was placed at a 0° projection angle and if it was dead center and perfectly angled, you would need hardly any convergence adjustments... just maybe a bit of geometry and amplitude and linearity (unless of course if the projector is out of whack somehow).

I just really cant stress how important it is that the mechanical setup/mount be dead on. To give you a sense of why, if the projector was placed at a 0° projection angle and if it was dead center and perfectly angled, you would need hardly any convergence adjustments... just maybe a bit of geometry and amplitude and linearity (unless of course if the projector is out of whack somehow).

Given the projector at 0 degree angle and dead center, is there a formula for virtually no keystone in relation to screen height?

Given the projector at 0 degree angle and dead center, is there a formula for virtually no keystone in relation to screen height?

Well, 0 degree and center means that line perpendicular to the screen and originating in it's center will hit center of PJ's green lens, meaning on-axis position. Since keystone adjustment compensates for off axis position of PJ, there is no keystone correction necessary in this case.

Kris