In January, I did a partial recap of my LNT3242HX/XAA because it would not power on. Several caps were swollen, and out of an abundance of caution I replaced every cap on the board that was made by that manufacturer (except possibly some tiny ones that are pretty unlikely to fail). A couple of weeks ago, the set started misbehaving, with the blackest parts of the picture bizarrely showing up as bright cyan. The picture is otherwise normal except for those cyan sparkles in black areas (or cyan over the entire screen if you unplug the HDMI cable).
Suspecting a bad T-Con board, I ordered a replacement. However, the replacement is much worse than the original; everything is bright pink. I'm assuming that the replacement board was simply bad, though the seller claims that all their parts are "tested". I'm not sure what I expected from someone who ships circuit boards wrapped in bubble wrap without an antistatic bag. (What the ****? I'm amazed it passed a signal at all!)
Just to be certain I wasn't missing something else, I checked the power supply voltages with my oscilloscope and noticed what strikes me as an unusual amount of ripple at 120 Hz, with some rails that seem out of tolerance:
Note that these voltages are measured in-circuit, at the input to the main board, so it is possible that the 13V rail is low because of loading. (What does this rail do?) But the 5.3V rail seems too high to me — not out of spec enough to fry anything, but enough to maybe cause subtle misbehavior like the blue sparkles, or cause the T-Con board to fail prematurely.
Is it typical for power supplies in these sets to be so badly out of spec? Am I right in suspecting that the giant film capacitor right after the full-wave bridge is failing to adequately filter the AC here? And is that 13V rail really under so much load that it sags by almost half a volt? Could any of these issues be causing the black noise problem, rather than the T-Con board? Or maybe a defective FWB?
Any other ideas?
Suspecting a bad T-Con board, I ordered a replacement. However, the replacement is much worse than the original; everything is bright pink. I'm assuming that the replacement board was simply bad, though the seller claims that all their parts are "tested". I'm not sure what I expected from someone who ships circuit boards wrapped in bubble wrap without an antistatic bag. (What the ****? I'm amazed it passed a signal at all!)
Just to be certain I wasn't missing something else, I checked the power supply voltages with my oscilloscope and noticed what strikes me as an unusual amount of ripple at 120 Hz, with some rails that seem out of tolerance:
- The 5.3V rail is actually 5.4V ± .2V (+5.6% / -1.9%)
- The 12V rail is actually 12.2V ± .2V (± 1.16%)
- The 13V rail is 12.6V ± .4V. (-6.2% / +0%)
Note that these voltages are measured in-circuit, at the input to the main board, so it is possible that the 13V rail is low because of loading. (What does this rail do?) But the 5.3V rail seems too high to me — not out of spec enough to fry anything, but enough to maybe cause subtle misbehavior like the blue sparkles, or cause the T-Con board to fail prematurely.
Is it typical for power supplies in these sets to be so badly out of spec? Am I right in suspecting that the giant film capacitor right after the full-wave bridge is failing to adequately filter the AC here? And is that 13V rail really under so much load that it sags by almost half a volt? Could any of these issues be causing the black noise problem, rather than the T-Con board? Or maybe a defective FWB?
Any other ideas?
Comment