Sony XBR-65X900F black and white image on one side (fading saturation)

[NSkyLab]

I though I've seen all possible kinds of failures but this one is new...
Colors saturation is gradual.y fading out on the left side of the screen (all colors are fading out). While B&W image is looking good.
I tried replacing TCon but nothing changed. So this is looking like the panel failure. Is there a chance it can be fixed?
Any useful measured can be done on the panel's narrow pcbs? Is there may be something similar to taping trick to fix this issue?

[Davi.p]

Hi, yes it's an uncommon situation, i think the wrong side is the one at top in picture, yes? It's most possible a problem with the Vcom signal in that side, i think it's too low, check the left side distribution strip board for some testpoints named Vcom-something, it has to be around 7v (or anyway half-avdd), if it's low check the respective pin on the tcon board, if there is ok then a connection problem, or some discrete on the strip board.. bye

[Diah]

clean LVDS from both side after inspect the contact pins no one have cut at the end or not aligned, after that if every things good for you, do hard reset the TV

[Diah]

one things more need to think about it,,, you put the TV in position where the effect shown at part the TV back close to the wall, if the wall have any Magnetic caused by wiring or metal will show this on screen..

[NSkyLab]

Hi, yes it's an uncommon situation, i think the wrong side is the one at top in picture, yes? It's most possible a problem with the Vcom signal in that side, i think it's too low, check the left side distribution strip board for some testpoints named Vcom-something, it has to be around 7v (or anyway half-avdd), if it's low check the respective pin on the tcon board, if there is ok then a connection problem, or some discrete on the strip board.. bye

Yes, top is the issue, all pictures are supposed to be plain R, G and B. Vcom is 7.25V on both sides of the panel, AVDD is 16.8V.

[NSkyLab]

one things more need to think about it,,, you put the TV in position where the effect shown at part the TV back close to the wall, if the wall have any Magnetic caused by wiring or metal will show this on screen..

Thanks, but orientation does not matter here.

[NSkyLab]

clean LVDS from both side after inspect the contact pins no one have cut at the end or not aligned, after that if every things good for you, do hard reset the TV

Thanks. Already did all of these, no change...

[Davi.p]

half avdd so vcom is supposed be 8,4v if i'm right , is there any trimmer in the panel strip board? Is vcom signal correctly reaching the panel on the last cof tab?

[NSkyLab]

half avdd so vcom is supposed be 8,4v if i'm right , is there any trimmer in the panel strip board? Is vcom signal correctly reaching the panel on the last cof tab?

Yes, 7.25V goes straight to the last cof tab. I compared both sides (left and right strip boards are almost identical) and voltages are equal...

[Diah]

Thanks. Already did all of these, no change...

in this case try to set the display under the 10 bits 8bits or less if available ,, and switch off HDR 10 and HLG, let see how it will act

[Davi.p]

Sorry, yes, since vcom is the same on 2 sides it is not the wrong voltage, now i suspect another thing, i think that when it changes between full screen colors there is a lack or scarceness of the vgoff=vgl, so the pixels energized in a certain color remains charged in the next color... check vgl testpoint in both sides and especially next before the last cof tab entering..

[Diah]

OP , don't miss with Panel buffer boards by testing VDC.. who knows will great shorts. this Android SoC GPU, we do have in this forum same effect at another Android brand.

the one who are talking on discharge... white & Black perfect.. white R=255 G=255, B=255 black all RGB = 0 why no effect on discharge when value at cells go high and down !!! funny

[NSkyLab]

in this case try to set the display under the 10 bits 8bits or less if available ,, and switch off HDR 10 and HLG, let see how it will act

Main board is disconnected. This lower level issue.

[NSkyLab]

OP , don't miss with Panel buffer boards by testing VDC.. who knows will great shorts. this Android SoC GPU, we do have in this forum same effect at another Android brand.

the one who are talking on discharge... white & Black perfect.. white R=255 G=255, B=255 black all RGB = 0 why no effect on discharge when value at cells go high and down !!! funny

Android has nothing to do with that since the issue is on the physical level, not software level. The panel physical interface is not plain RGB, it does have a signal to explicitly switch pixels off. Davi.p may be on the right track...

[NSkyLab]

Sorry, yes, since vcom is the same on 2 sides it is not the wrong voltage, now i suspect another thing, i think that when it changes between full screen colors there is a lack or scarceness of the vgoff=vgl, so the pixels energized in a certain color remains charged in the next color... check vgl testpoint in both sides and especially next before the last cof tab entering..

I can't find the vgoff testpoint, but I measured every testpoint I could find and they are identical on both sides... Pretty sure it's in the COF. So it's going to the dumpster today.
Thanks for help.

[Davi.p]

oh, so a sin, i have tried a better explanation because i was indeed not convincted, but i needed a repassing of study of panels basis, unfortunately i am at work now.. the thing is surely something of this: there's a leakage from one of the source data bits to the other data bits, this in the area from output of tcon towards the faulty side, until the rest of the faulty side panel, source data bits are mixed and almost serial in that area, when them arrives to source COFS them are disassembled in the DAC+shift registers inside COFS, somehow there is a parasite capacitance that mixes up RGB data thogether, when RGB columns have to be displayed all 3 at 0 level (=vcom) then the mixing has no effect (black displayed well).. in the ending i could suspect maybe some dirt in strip board, or some bad /missing EOL resistor near the source COFS, them has to be all 100ohm.. bye..

[NSkyLab]

oh, so a sin, i have tried a better explanation because i was indeed not convincted, but i needed a repassing of study of panels basis, unfortunately i am at work now.. the thing is surely something of this: there's a leakage from one of the source data bits to the other data bits, this in the area from output of tcon towards the faulty side, until the rest of the faulty side panel, source data bits are mixed and almost serial in that area, when them arrives to source COFS them are disassembled in the DAC+shift registers inside COFS, somehow there is a parasite capacitance that mixes up RGB data thogether, when RGB columns have to be displayed all 3 at 0 level (=vcom) then the mixing has no effect (black displayed well).. in the ending i could suspect maybe some dirt in strip board, or some bad /missing EOL resistor near the source COFS, them has to be all 100ohm.. bye..

If I properly got what you've described then I'd expect to see some artifacts or crooked colors... I didn't post those photos, but the picture looked fine, there were proper colors, except the saturation on the left side. According to the owner it was gradually getting worse over the last few months...

[Davi.p]

Excuse me, what i see is not full saturation, is something like solarization, at this purpose the there can also be an issue at gamma correction, if there are gamma test points on the strip board can be useful to check them, do you know how? about the data bits problem that i said, suppose the screen is 10 bits, every sub pixel is defined by them, suppose full red screen is displayed, and most significant byte of red shorts to most sig. byte of green & blue, the result is white or washed white, what i do not know is where the short could be, if i'm right, mini lvds data bus has one line that holds data for all 3 RGB, for example it could hold bit7+8 of each byte of RGB, so the most significant byte of all 3 components RGB, each line is made of 2 conductors for differential signal, if one of the couple is opened could determine a bad digital signal...

PS: sorry, very hard to type with cellular and my big inches..

[Diah]

Android has nothing to do with that since the issue is on the physical level, not software level. The panel physical interface is not plain RGB, it does have a signal to explicitly switch pixels off. Davi.p may be on the right track...

as you like and think... but you are wrong read about GPU , GOOD LUCK WITH YOURS TRACK