I've done a lot of web research over the last few days, but haven't found a failure mode quite like this. My Samsung LN52A650 (fluorescent LCD from 2008) has intermittent random horizontal lines across the entire screen. They occur regardless of the input source and on top of any OSD menu overlays. They seem sharply defined, as if they're digital in nature, not some analog/power issue in driving the panel. They vary in intensity from a single raster every few seconds to covering almost the whole screen continuously. When they get really intense, the TV will restart and the cycle will repeat. Attached photos show a scene with and without the lines (and also the T-con board). I also recorded a video: https://www.youtube.com/watch?v=_0RrRdSpS_U
Since this model was affected by the Samsung bad cap class-action, and bad power can cause all sorts of issues, I looked at the PSU caps first. None were bulging or leaking, but just in case, I replaced the 4 820/1000 uF caps at the low voltage output. No change in behavior and the old caps tested okay.
My best guess so far is that the issue is somewhere between the output side of the main board processor and the input side of the T-con board processor. Spraying freezing spray on the T-con processor seemed to calm the issue, though the issue can be intermittent enough to make it difficult to be sure. I also tried freezing the main board processor, but there seemed to be even less of a correlation. I've reseated both ends of the LVDS cable and the T-con end of the ribbon cables, but it had no effect. Disconnecting the LVDS cable while the noise and restarting were continuous stopped the restarting. Reconnecting the T-con and restarting, such that the main board processor was still hot and the T-con processor had cooled, seemed to calm the issue. So while I've heard claims that T-con board failure doesn't cause horizontal lines, that's my best guess so far. Assuming the data over the LVDS is horizontal raster-oriented, it seems to be getting corrupted in bursts along the way. Since the issue seems to vary with heat, and wiggling the LVDS cable does nothing, I'm assuming it's not the cable.
I'd appreciate any thoughts on the issue. Should I try replacing the T-con board next?
Since this model was affected by the Samsung bad cap class-action, and bad power can cause all sorts of issues, I looked at the PSU caps first. None were bulging or leaking, but just in case, I replaced the 4 820/1000 uF caps at the low voltage output. No change in behavior and the old caps tested okay.
My best guess so far is that the issue is somewhere between the output side of the main board processor and the input side of the T-con board processor. Spraying freezing spray on the T-con processor seemed to calm the issue, though the issue can be intermittent enough to make it difficult to be sure. I also tried freezing the main board processor, but there seemed to be even less of a correlation. I've reseated both ends of the LVDS cable and the T-con end of the ribbon cables, but it had no effect. Disconnecting the LVDS cable while the noise and restarting were continuous stopped the restarting. Reconnecting the T-con and restarting, such that the main board processor was still hot and the T-con processor had cooled, seemed to calm the issue. So while I've heard claims that T-con board failure doesn't cause horizontal lines, that's my best guess so far. Assuming the data over the LVDS is horizontal raster-oriented, it seems to be getting corrupted in bursts along the way. Since the issue seems to vary with heat, and wiggling the LVDS cable does nothing, I'm assuming it's not the cable.
I'd appreciate any thoughts on the issue. Should I try replacing the T-con board next?
Attached Files
Comment