Samsung streak continues. This time I am a proud new owner of a dead 23″ Samsung 2333SW LCD monitor from September of 2009. Dead is a bit of an overstatement, but that’s what the previous owner told me. Since the alternative was for it to go into a dumpster, I figured worst case I get a 23″ LVDS panel to play with.  On to the symptoms: backlight seems to operate and the “connect you video source” banner is bouncing around the screen. When I do connect it, I see a brief attempt to display an image and then it’s back to gray. Time to take it apart.  Unlike the TV, I see almost no screws- a total of two on the back are all that I could find. The rest is held by the clips- a nightmare to put it mildly to open.

An hour of swearing later, the lid is off:

And what we see is a bit disappointing. The whole central assembly is just sitting there, not attached by any means. It was simply held in place by the lid. Cost cutting/streamlining at every step I guess. And of course all the boards face inside the cover

There is not much to the image processing board- a DVI/VGA input to LVDS conversion chip with whatever other image magic it needs to do- probably at least scaling. A few LDOs, EEPROM for the VGA interface and an EEPROM and Flash for the  TSUMU58EHJ-LF-1 chip. Looking with the scope, I see VGA signals wiggle at reasonable levels, but the chip is not seeing them. Since the levels look good, it’s unlikely that input protection is pulling them down, at least not the external ESD protection. Can’t say much about what’s going on inside the chip. On the other hand, it is driving LVDS side just fine as I see the message on the screen. Continuing to poke around, we see that all power rails seems to be looking good and clock is operating normally. The chip seems to run rather hot- almost 50C, while doing nothing. So I am starting to suspect that it itself  may be the culprit. But first we got check our voltages.

Power supply board looks pretty good, no visible cap problems, A check with a DMM and a scope confirmed that all voltages were normal. Backlight drive was operating normally as well.

So back to the imaging board. I’ve cleaned the board and inspected every pin on the image processor under a microscope- nothing obvious. Next step is to hit ebay and see what we can find. There is nobody selling exactly  TSUMU58EHJ-LF-1 as on this board, but a few sellers had similar part numbers. A few emails later I have one seller with the part I need. Part ordered. The seller is in China, so I may get a new chip, or a used chip, or a totally different one labeled as what I need. But at the price asked, I might as well try. To be continued..