CurtPalme.com Home Theater Forum

[rotoscoped]

A few days ago, I removed my red tube and changed the badly yellowed glycol. All went well. (I'd changed the blue tube glycol with no issues a couple years back.)

When I turned the projector on, I noticed two strange things. First, the convergence was reset, everything misaligned. Second, and stranger, the blue tube had a pronounced pincushion at the top and bottom of the image, quite bad at the top. This is especially odd since I didn't touch the blue tube or any of its connections throughout this process.

I then attempted to converge everything. I roughed in red convergence, and there are no apparent problems. The blue image, however, will not allow any "up/down" adjustments. I can adjust "left/right" convergence, but other than the central adjustment zone, which allows you to globally move the blue image, I can't adjust "up/down" anywhere. The bizarre pincushion shape of the image remains in place. I can't "pull" it back towards the green grid. This is true, by the way, for guided, interpolated, and random access convergence. It should be said that, prior to this, my convergence was quite good - the blue in the top corners didn't perfectly match up, but the discrepancy was small, a quarter inch, just in the corners, on an 82" screen.

I rechecked all the connections and everything seems as it should be. And, indeed, nothing related to the blue tube was touched in the first place.

Any thoughts?

[Curt Palme]

You've lost all V convergence on the blue tube. Check the ribbon cables going to the convergence board, check to make sure that you're not out by 1 pin on the convergence yoke. There's a number of fusible resistors on the convergence board that can go open without reason, and assuming that you don't have the ribbon cables out by 1 pin, then it's likely that one of those resistors has opened up.

[Curt Palme]

You've lost all V convergence on the blue tube. Check the ribbon cables going to the convergence board, check to make sure that you're not out by 1 pin on the convergence yoke. There's a number of fusible resistors on the convergence board that can go open without reason, and assuming that you don't have the ribbon cables out by 1 pin, then it's likely that one of those resistors has opened up.

[rotoscoped]

Thanks for replying, Curt. I double checked, and the ribbon cables are all seated correctly. I looked over the convergence board and didn't notice any damaged resistors -- but I'm a novice in this area and may have missed something.

I've posted photos if anyone would like to inspect: https://imgur.com/a/ycN0EFb. (I was reluctant to remove ribbon cables, so you'll have to browse multiple images to get a look at all the resistors.)

I read in the service manual re convergence issues:
"Coil failure is a possibility but unlikely. To check for this you can swap yoke plugs on the CVA to see if the problem follows."

Should I swap blue and red convergence yoke connectors (ie. the orange/brown/violet/black wires going into the convergence board), and see what's adjustable? If I still can't adjust blue, if the problem follows the ribbon cable, this would suggest the ribbon cable is faulty? And if I can adjust blue but not red after swapping, this would suggest a problem at the convergence board?

[Tim in Phoenix]

Guys

"And if I can adjust blue but not red after swapping, this would suggest a problem at the convergence board?"

It helps to understand the circuitry. Convergence waveforms are generated from digital values on the control board, sent as analogue waveforms and amplified on the convergence board. Your blue north/south signal is lost somewhere in there. Yoke failures are almost unheard of.

I have boards here if you need some to try. I would need your control board firmware version to match what you have now.


Tim Martin, E-Tech Systems Scottsdale

480 253 2781 please text me first\0

[rotoscoped]

I will keep that in mind, Tim, re the boards. Glad to know that's an option, though I hope it won't come to it...

After reseating the blue tube ribbon cable on the convergence board last night (removing it and replacing it to be sure it was well connected), I find today that I have blue vertical convergence again -- EXCEPT in the top right corner. What remains of the pincushion distortion is still there. It is only the farthest two blue convergence zones at the top right that won't allow vertical convergence.

This, coupled with the appearance of the pincushion distortion in the first place, suggests something is amiss. The effect on the image is tolerable, but by no means ideal.

I welcome any troubleshooting ideas or tips.

[Curt Palme]

It may just be the image, but it looks like fungus is also starting in the blue tube, in patches.

For the convergence, do a full reset, and start over. Then use the 'interpolated' mode, rather than the 'guided' and see if the blue problem goes away.

[Curt Palme]

It may just be the image, but it looks like fungus is also starting in the blue tube, in patches.

For the convergence, do a full reset, and start over. Then use the 'interpolated' mode, rather than the 'guided' and see if the blue problem goes away.

[Tim in Phoenix]

Guys


"After reseating the blue tube ribbon cable on the convergence board last night (removing it and replacing it to be sure it was well connected), I find today that I have blue vertical convergence again -- EXCEPT in the top right corner. What remains of the pincushion distortion is still there. It is only the farthest two blue convergence zones at the top right that won't allow vertical convergence."

Recommend you try matching raster heights in the Service Menu. UTIL, 9, 0901.....

[rotoscoped]

Thank you both for following up. I reset convergence to 0 and used interpolated - and the result was the same. The top right corner distortion persists. I also double checked raster heights, and everything is okay there. As soon as the issue cropped up, I went through the "Guided Mechanical Setup" again, to be sure the fundamentals were right, so I'd recently checked raster heights. It should be said, though, that the "unadjusted" blue raster has distortions worse than unadjusted red - ie. bowing on the right side, especially the right top. It requires more convergence to match blue to green.

I'll add further detail about blue convergence in the top right corner. I've realized I do have vertical convergence there, but there's no convergence "south". I can move blue "north" in that top corner, but not "south". It's as if it's "bottomed out" there. This is true for 5 convergence zones in that corner, the farthest right three on the top row, and farthest right two the next row down. They can all be further adjusted "north" but not "south". This is only true in this area. There's plenty of convergence room in the other corners.

(And re fungus in blue, the "blooming" is just distortion and reflection from the photo. To the naked eye everything is clear.)

[Tim in Phoenix]

Guys

I would have to consider a control board problem, convergence board problem, or a rail from the LVPS is slightly low. Suggest removing right side covers (blue tube side) and measure voltage on all the pins on the big Molex connector.

[rotoscoped]

Tim, thanks – I will follow up with you in a PM re further details for measuring the voltages.

Something worth noting again, I think – this unusual blue convergence distortion appeared immediately after I removed and replaced the red tube. Quite mysterious, but also surely significant. Something must have happened in the process of that tube removal to cause this issue. From my perspective, nothing out of the ordinary took place – merely the red tube cables being detached and reattached... Nothing at all involving the blue tube.

[Curt Palme]

Is there any chance that the raster is right at the edge of the tube surface? I did that a few times, and couldn't get convergence right in the corner where the raster was right at the edge of the tube. Put up the grid, look into the lenses with sunglasses on, and see if it's shifted too far to one corner or one side.

[Curt Palme]

Is there any chance that the raster is right at the edge of the tube surface? I did that a few times, and couldn't get convergence right in the corner where the raster was right at the edge of the tube. Put up the grid, look into the lenses with sunglasses on, and see if it's shifted too far to one corner or one side.

[rotoscoped]

Curt, on reading your post I was excited to explore that possibility – and was convinced for a bit that it must be right! The raster is close to the edge of the tube in the area of distortion. I reset convergence so that the blue raster was more centered. When I began to converge again, I didn't adjust the blue raster globally first, rather I first adjusted the right corner, thinking I might now have more "play" to adjust it to the correct position. Having done that, dropping the right corner as far as I could, and then adjusting the blue raster globally to match to green, I found that the distortion was just the same as before. The same "bowing" at the right corner. I had exactly the same amount of "play" in that corner's convergence when the raster was centered as when it was closer to the edge. I did have that corner perfectly converged before this recent trouble, so the closeness to the edge must not be the issue, since I had it well converged with the raster in that same position... In the pictures in my post above you can see how close the blue raster is the edge of the tube.

[Tim in Phoenix]

Say Roto......

Can loan you some boards to try, control and convergence. Would need the firmware version, press * for onscreen display.

[rotoscoped]

I do have an update and hopefully swapping boards will not be necessary... Playing with possibilities last night, I tried shrinking the vertical size of the blue raster so that the distorted blue corner, which cannot be adjusted, would meet up with the green corner. This did work. What is happening is that the blue raster is now slightly undersized overall, short of green at the top and bottom in most areas, but matching perfectly in the center to green, requiring no global adjustment, and matching perfectly in that unadjustable top right corner, requiring no adjustment. It was essentially every other zone that the convergence adjustments had to be made.

The good news – this did work. The top right blue distortion is straightened out. I'm not sure, though, if there are other reasons this workaround would be inadvisable. Is shrinking the blue vertical raster size in this way, just to 'resize it' via convergence, doing something to degrade the quality of the image? Perhaps by putting greater stress on convergence it will be more prone to convergence drift, requiring more frequent adjustment. If this is the only drawback, it seems worthwhile given that it's restored the overall convergence accuracy.

[Tim in Phoenix]

Say Roto

You have matched the raster heights and that is correct.

[rotoscoped]

Thanks for your help, Tim and Curt. Very pleased that the desired result could be achieved without any board replacement!