Convergence problem in RGB CRT-based projection TV

Hi a friend has a Toshiba RGB CRT-based (3-crts) projection TV which has a convergence problem. There are some trimmers hidden below the shell of the TV to adjust the red and the blue, but they adjust these only horizontally. He does not want to spend money in ordering new convergence chips, so he asked me if I could do something to fix this.
The convergence problem is mostly at the sides of the screen, at the center it is ok.

Also there is an annoying shaking of the screen at fast rate, that is not good for the eyes to watch.

Any ideas if any of these can be fixed somehow?

The convergence signals come directly from the timebases so they are synchronized to the horizontal and vertical scans. That means they can't make the screen 'shake'. The underlying problem is probably ripple from the PSU.

Brian.

Convergence is adjusted on Deflection Yoke of the CRT ( I talk about standard CRT TV Tubes ) and this is fixed at the factory but still adjustable..

Convergence is adjusted on Deflection Yoke of the CRT ( I talk about standard CRT TV Tubes ) and this is fixed at the factory but still adjustable..

Click to expand...

This would be a three CRT system. There are optical adjustments to converge the images depending on the distance to the projection screen, basically rotating the CRTs so they triangulate at the target. The shape distortion due to different launch angles still has to be corrected electronically so the whole of the three rasters converge over the whole area.

My point was that the shape correction is derived from the timebases so it has to be at a fixed position relative to the scan lines. That means any 'wobble' or 'shaking' must be unlocked from the timebases and almost certainly comes from ripple in the PSU output. The usual cause is dried reservoir electrolytics at the PSU input, they allow some of the 50Hz/60Hz line signal to modulate the timebases and 'strobe' with the video source which probably has a similar but locked vertical scan rate.

Brian.

betwixt said:

This would be a three CRT system. There are optical adjustments to converge the images depending on the distance to the projection screen, basically rotating the CRTs so they triangulate at the target. The shape distortion due to different launch angles still has to be corrected electronically so the whole of the three rasters converge over the whole area.

My point was that the shape correction is derived from the timebases so it has to be at a fixed position relative to the scan lines. That means any 'wobble' or 'shaking' must be unlocked from the timebases and almost certainly comes from ripple in the PSU output. The usual cause is dried reservoir electrolytics at the PSU input, they allow some of the 50Hz/60Hz line signal to modulate the timebases and 'strobe' with the video source which probably has a similar but locked vertical scan rate.

Brian.

Click to expand...

Ok I will check out the PSU filter capacitors thank you.

There are focus lens on each CRT and crt brightness and electronic focus potentiometers. I tried my best to set all these. The green and red focused ok but the blue was not so crisp. I tried to focus the lens of the blue, but still could not get a very crisp blue.

About the convergence problem, I did not know I could rotate the CRTs, do you think I should really try this? It seems the red is a bit out on one side of the screen and less on the center.

That kind of projector was never perfect, the technology was replaced several years ago with single light source and LCD filters.

The rotation isn't the CRTs along their axis, it is to rotate them left and right so the beams converge at different projection distances. If you think about it, the RGB images are spaced apart at the projector but must overlap at the screen so the outer ones need to aim slightly inwards. How much depends on the distance between the projector and screen so it has to be adjustable to each installation.

The life of the CRTs is quite short because they run at high brightness, if they don't focus probably it means the tube is due for replacement. They are expensive and probably hard to get hold of now.

The best way to set convergence is to display a light grid on a dark background so you get parallel horizontal and vertical lines. About 10 lines, equally spaced is optimal. If you can arrange that and photograph the image it would help to diagnose the problem.

Brian.