(Background Image Source: EEVBlog)

As the saying goes, one man's trash is another man's treasure. And Mr. Dave certainly found a treasure in his latest dumpster dive, perhaps comparable to EEVBlog #984 where he picked up a $300,000 MSO  ( https://www.youtube.com/watch?v=d_S9YsD9Y0c ). It's not everyday you see a 4K LCD TV with no crack or signs of defect near the trash can.

For those who are unaware, the dumpsters are in fact located in a commercial business park near the building housing Mr. Dave's lab. It's not available to the public, and one would require a key to gain access to the area (hence, only tenants of the business park can scavenge for disposed electronics that are still of value/require minimum repairs).   

The TV booted successfully with no visible defects on the pixels after being transferred to Mr. Dave's new office. All buttons were working properly too. This stroke of serendipity however, was cut short when the HDMI interface was tested. The merely 4-year old LCD TV displayed a riot of incomprehensible colors on its screen, protesting against the AV signals it was being fed from a working computer. 

(Image Source: EEVBlog)

To guarantee the problem wasn't from the computer, the TV was also fed AV signals from a Raspberry Pi, which yielded the same results. Apparently, this was the reason why it was discarded. Fortunately, a  problem of this sort is tractable to any tech-savvy hobbyist or engineer.

Take It Apart!

The first step in any top-down troubleshooting session is to analyze block by block the faulty system. Of course, this can only be done effectively by taking the system apart (unless you have an X-ray machine, but that's overdoing things when a handy-dandy screwdriver will do). This also happens to be the famous EEVBlog motto.

(Image Source: EEVBlog - Mr. Dave sporting his Tough Mudder shirt)

Upon exposure of the chassis, immediately observable are the 3 main parts of an LCD TV - the power supply, the t-con board and logic control board.

 (Background Image Source: EEVBlog)

LCDs are driven by both a positive and negative rail, to bias the gates properly given a TFT LCD TV. These rails are low voltage relative to the AC supply, but draw high currents that can reach as much as 80A. The T-con and logic board consumes probably a mere 10-20% of the TV's overall power consumption in spite of the processors themselves drawing substantial currents.

The power architecture of such systems require exacting features from the PMIC. In a later section, where Mr. Dave probes all regulators for the correct bias voltage, this becomes a crucial factor to consider as an entire block will not function if just one defective regulator is detected via a feedback path. More on this later...

The Capacitor Plague Revisited

In the video, Mr. Dave expressed his disapproval over LG's use of Samyoung capacitors. This may be rooted on the 1999 to 2007 capacitor plague, where faulty dielectrics resulted in failing devices. (more info here via Wikipedia)

Hence, more undeniable proof that mistakes in verification and manufacturing can't be afforded and can damage company reputation irreversibly. Any system or product changes must be pondered and reflected on extensively to ensure yield and Cpk are unaffected.

After a thorough visual inspection, no busted capacitors were found on both boards. Debugging proved to be more of a challenge as the actual problem wasn't evident to the naked eye.

What are these!?

(Image Source: EEVBlog)

The white blocks on the picture above are new to me too. Do they perhaps serve the same function as the black blobs for COBs (chip-on-board technology)? Or maybe they provide cover for devices highly sensitive to static? Need more research on this...

Checking Bias Levels of All Regulated Voltages

(Image Source: EEVBlog)

Checking node voltages is an effective troubleshooting method. As mentioned earlier, complex systems most probably have complex power management protocols that monitor supply voltages from each regulator via a feedback path. If a channel falls below a threshold, the entire system may turn off and not boot at all.

There are many implications to the argument above. A damaged regulator with adjacent fully functional channels may mean the source PMIC also has problems. Thus, it is expected that if a regulator isn't working, the entire block is turned off.  
It is worth noting that Mr. Dave was lenient on an 11.78V reading given a 12V regulated supply. Aside from load regulation effects from current-hungry processors, (in a competitive market, a regulator with a 120mV drop at the max operating point will never see the light of the manufacturing line) the DMM may have loading effects, the DMM terminals aren't on the correct "sense" terminals (hence, voltage drops from the wires are affecting the reading), or the regulator may not be trimmed

The T-con Board or Not the T-con Board, That is the Question

Unlike Mr. Dave's opinion that the defect is in between the HDMI and logic, I have a differing say on the matter because I think the real culprit is the T-con board. But why did the text show up fine then? It may have been a built-in ASIC or SoC of some sort, which did not require the T-con for display (after all, the text display is a repetitive pattern). Then again, I'm only well-versed with the power architecture of TFT LCD TVs, and may be wrong with the AV signal flow.

How to Use a Freezer Spray to Isolate a Faulty Chip

Since repair is in the component level (and not on the board level, where all one has to do is jut down the board number and search for a re-seller online), a freezer spray is needed to pinpoint the defective component on the suspected logic board.

Damaged components/shorted parts tend to dissipate more heat due to higher current. If the board is frozen over and power is applied, the first ice crystals to melt are those on or near the defective component.

Below is a helpful video on how to use a freezer spray.

More to Look Forward To!

As mentioned near the end of the video, Mr. Dave wasn't done troubleshooting the TV. Signals from the antenna for digital television may be working and it was perchance that the signal was blocked or very weak in his lab. This reminds me of the last analog television broadcast via AM in Australia, which Mr. Dave also made a video of. But that is another story. 

EEVBlog #1145