A Link To The Past
VGA input may have fallen out of favor, but it still makes for pretty PCB layouts from the neatly ordered resistor-capacitor (RC) low-pass filters on the left to the resistor-capacitor-diode (RCD) input protection to the right. DDI on the VGA port is provided by a 2Kb STMicro 24C02RP I2C EEPROM, which is the same for DVI and HDMI.
A Look At The DVI Input
The DVI input protection looks somewhat similar, despite being digital-only. At the top-mid, we see the DDI chip and its own protection circuitry. Below and starting from the right, we have some decoupling caps for the input protection, input clamping diodes and series termination resistors with more decoupling caps in between signal pairs. A TI TMDS351 triple-input DVI/HDMI switch occupies the bottom-left corner, though only two of its inputs are implemented on this PCB.
… And The HDMI Input
Things look much simpler on HDMI's side. I could not find information about that D1380747 part, but seeing how all four TMDS signal pairs go straight to it on their way to their respective resistors and the TI mux beyond, I would bet it is simply an integrated octuple diode clamp network in lieu of the discrete networks seen on the DVI and VGA inputs. Of course, we also have the third of those DDI EEPROMs near the bottom-left with its own support components scattered more loosely than the other two ports.
Let There Be Sound!
The audio coming out of the rear 3.5mm audio jack has to originate somewhere. That'd be a Cirrus Logic CS4352 24-bit 192ksps DAC with 96dB THD+N, which is on par with the theoretical maximum for CD audio. I have never used it before, though at a glance this looks like a neat and tightly-packed layout located out of the way from obvious ground current paths, so I would expect reasonably clean output.
I briefly listened with the first earphones I could find, and while I did not hear obvious noise or distortion at low volume, the output did start having obvious clipping with the volume set as low as 20% in Windows, which corresponded to loudness somewhat above normal for me.
Controller Chip And Friends
The display controller is an STMicro gm5868 with the required external DRAM chips to support frame rate conversion. While ST's specifications say only 75MHz 16x512k (yes, that's 1MiB) SDRAM is required, LG uses 200MHz ESMT M12L16161A-5T chips, likely because slower low-density SDR memory was effectively extinct when the board was manufactured in 2008.
The small chip partially cropped off at the bottom is a Macronix MX25L4005 4Mb serial NAND device to store the display manufacturer's UI bitmap resources, and the other chip at the top is a 16Kb STMicro 24C16WP I2C EEPROM to store user settings.
And Now, The Main Course
I said this was a (reverse) tear-down and repair, and I have not shown anything repair-ish yet. We're getting there.
Since the PCB is single-sided, all of the surface-mount components are located on the bottom, including two integrated circuits. The top IC is obviously the PWM regulator for the 12V and 24V outputs, while the smaller chip near the bottom-right corner does active power factor correction.
Now That's Convenient
All of the major top-side components and every jumper link are silkscreened on the bottom for easy signal tracing and troubleshooting. The silkscreen also interrupts some solder puddles by acting as a presumably unintended solder mask. Apart from that, soldering quality looks quite good overall.
Markings on the APFC chip must have seen better days: it looks like small flakes chipped off the packaging. Skinny dipping through a molten solder fountain can do that to you.
Searching for APFC controllers in SOIC-8 packaging with “280” in their model number spits out TI's UCC28051D as the most likely candidate. At a glance, the pin-out looks like a probable match.
Going On Top
Looking at either side of the board, most of the power handling path is obvious. Power arrives at the IEC plug on the left, goes up through the input filtering, passes through the bridge rectifier on the top heat sink, starts its way back down through PTCs, some more filtering with a choke and film caps, the APFC circuit in the bottom-left corner and then back up along the isolation boundary to feed the two output regulators.
Can you spot the issues from this angle?
If you ever want to reuse this display's power supply for something else, the input and output ratings are all conveniently printed on the board.
In the top-left corner, we also see that this board was manufactured during the 13th week of 2008, just like the CCFL one. The UL listing just below the manufacturing date tells us the PCB was made by Shanghai Wanzheng Circuit Board Co. Ltd., and has passed 94V-0 flammability tests.