LG 3D 4K TV repair - UB980 range 65UB980 79UB980 84UB980

[3DMovieList]

Recently I've been working on the repair of (IMHO) three of the best 3D TVs ever manufactured - the "UB980" range of LG passive 3D 4K TVs

Fortunately, there is a good amount of service information available online for these TVs.

The LG Service manual for the 79UB980T is available here:
https://www.manualslib.com/manual/1293719/Lg-79ub980t-Ta.html

The T-CON board is missing from the above service manual, but fortunately it is summarised on page 123 of this document:
https://pdfcoffee.com/collection-of-lg-t-con-board-testing-points-amp-voltages-pdf-free.html

However, there is one more board (or actually 8 small long boards) still not described and it's the boards connected directly to the LCD panel.
I've included a photo of two of those boards here:


Unfortunately, one of those boards experienced some water damage, as pictured here:

I guess some water or other liquid was spilt on the TV.
Fortunately, these boards are fairly simple, mainly consisting of some capacitors and resistors.
After cleaning off the corrosion, I've found three capacitors had gone short circuit and I've been able to replace those.
However, there is also a 5-legged SMD IC labelled "J516T" which is performing unusually.

Unfortunately, I have not been able to find a datasheet or other information about this IC.
Does anyone have any ideas?

Another item I have to contribute and summarise here is that many of the LG TVs have a problem that the main chip gets very hot and causes dry joints in the solder joints.
Reheating the chip using a hot air solder station can help bring TVs back to life (I have repaired three LG 3D TVs (LW6500 models) this way).
Here is a video which illustrates the technique of reflowing the solder on the main chip and overcoming the dry joints:


So... who else is repairing these LG passive 3D 4K TVs?

I have looked at the service manuals for other TVs, including other brands, to see if there is a hint as to what this chip might be, but no luck.
Anyone have any ideas?
Thanks!

[3DMovieList]

I found a datasheet (42 page PDF) for the LCD & backlight module used in the LG 84UB980T (LC840EQD-SEM1):
https://datasheetspdf.com/pdf-file/907195/LG/LC840EQD-SEM1/1
But unfortunately, there's still no circuit diagram for the circuit boards in this module, or identity of the mystery "J516T" component.

[EEVblog]

I did a video on almost exactly this:

[sinewave]

If you could remove the chip that would help and also post photos of the underside to better understand where signals/voltages are going. Doing a quick search I see that part number: R1154N001C matches quite well in terms of pinout if the middle pin on the side with three pins connects to the copper fill, looking at the application in the datasheet, I see that the VOUT pin is connected to a resistor, with an accompanying footprint for the resistor between the VOUT/ADJ and GND.

[3DMovieList]

G'day Dave and Sinewave,
Thanks for the pointers.
Attached are photos of the board (front and back) with the chip removed.

Unfortunately, the board looks like it is a multi-layer so not quite so easy to trace out visually.
U1 is the chip in question.
Pin 2 is definitely ground.
Pins 3,4,5 immediately disappear off to vias.  I would need to do some poking around with the multimeter to see where they go.
Pin 1 connects to a via _and_ a resistor.  I could measure the value of this resistor but not sure it would reveal anything definitive.
I was wondering whether the chip was a linear regulator, BUT the lack of any caps immediately nearby is a counter-indicator.
Looking up "_J51_6T" on the SMD marking code database suggests it might be a linear regulator, or an "ElectroStatic Discharge Protection diode".
https://smd.yooneed.one/code4a35.html
The first ESD SOT-23-5 that I looked up could be a winner.
https://au.mouser.com/datasheet/2/240/Littelfuse_TVS_Diode_Array_SP05_Datasheet_pdf-777043.pdf
The SP0504BAHTG has pin 2 GND, with a single diode going to each of the other four pins.
This looks to me to be the most likely bet - I'll check the chip with a semiconductor tester to see if this confirms anything.
Thanks again for the hints!

[sinewave]

I doubt it is the ESD diode as in the datasheet it says the package will be marked 504B which it is not and wouldn't seem needed, as there is a VCC silkscreen and a bank of 4 capacitors nearby makes me believe it is a regulator, I'd suggest you check if R_93 is continuous to pin 4 if it is I'd say with a bit more certainty that it is a regulator, most of the time with these components it's pretty hard to find a match anyway.

[projectgus]

(Moving here from our twitter discussion.)

I had a similar guess to you, sinewave, of R1154N051B. J516T marking matches (J51 prefix and then a two digit date code, which for that family seems to be number+letter). But pin 4 would be NC for that fixed voltage variant so I'm not very confident about it, unless the part is dual-sourced with an adjustable voltage version and pin 4 runs off to some unpopulated pads.

3DMovieList, you said on twitter that there are a few J516T chips on the PCB and the others all probe differently with your multimeter. FWIW, this is what I'd be focusing on next - can you post any photos of how the other J516Ts are connected on the board, and also can you give any details about what you've measured between the various pins? Bear in mind that probing in-circuit you may be measuring a lot of other things as well as that chip, but it still could be a good clue.

If you haven't already, another thing you could try doing is test continuity between the pins and the marked test points or bulk capacitors nearby on the board.

[sinewave]

There is indeed an adjustable version according to the datasheet, https://www.n-redc.co.jp/en/pdf/datasheet/r1154-ea.pdf R1154N001C is the adjustable version whereas R1154NxxxB is the fixed version, I do agree on you with it being dual-sourced as it seems to have the correct circuitry for an adjustable LDO.

[EEVblog]

On twitter I speculated a dual mosfet.
A regulator without any caps doesn't make any sense.

[EEVblog]

Maybe a load switch, but again, lack of caps.
https://www.farnell.com/datasheets/2304069.pdf

[sinewave]

I do agree with it more than likely being not an LDO, due to the lack of caps around, figuring out the actual package would help as there are 4 or so footprints with a similar layout (2 pins on one side and 3 on the other)

[3DMovieList]

A few folks have mentioned the Twitter thread.  For those following this discussion, the Twitter thread is here: https://twitter.com/3DMovieList/status/1583982843344326656?s=20&t=qZjoL39L_oq3PdYxA8_C-w
BUT let's keep the detailed discussion here.  ;-)

[sinewave]

3DMovieList, can you measure the height/width/length of the package?

[3DMovieList]

I don't have my notes with me at the moment, but when I originally tested the "good" J516T, it looked more like a diode package with several 0.6V diode drops identified using the multimeter.  But I didn't test thoroughly, and it was in-circuit, so at the time, I was just noticing a difference between the "good" and the "bad".
Hopefully I'll have time today to desolder a good J516T chip to allow some more detailed testing to check what the chip might be.
Thanks for all the good suggestions everyone.

[EEVblog]

3DMovieList, can you measure the height/width/length of the package?

I overlaid a SOT23-5 package on it visually and it matched.

[3DMovieList]

So I was able to desolder a good version of the "J516T" chip and do some tests.
Buzzing out the chip with a PEAK diode/transistor/semiconductor tester indicated that the chip contains 4 diodes, as indicated in the diagram below.

The configuration is (to me) unusual and the forward voltage of the diodes falls into two groups (two are 0.65V and two are 2.8V) as indicated in the diagram.
For the bad "J516T" chip, the diode on pin 1 was open circuit, but otherwise the same as the good chip.
I also tried testing various combinations of the SMD legs to see if there was a transistor or FET present, but nothing came up - just independent diodes.
So... the plot thickens.
I didn't spot any SOT-23-5 devices with two sets of different diodes in the package when I was looking around.
Any ideas?

[3DMovieList]

I just had a look through a range of 4 diode SOT-23-5 packages available on Mouser and couldn't find anything remotely similar to the diode configuration measured on the good J516T.
The two different forward voltages, and the alternating directions of the diodes I found seems very unusual.
I did test for transistor and FET modes too using the semiconductor tester, but the device always identified just four separately diodes.
I did some quick tracing of connections to the chip today and found just bypass capacitors so nothing special to suggest this chip is anything other than ESD protection.

[projectgus]

Is it possible you're measuring pin protection circuitry here?

Bit of a wild guess, but could it make sense for a differential line driver or receiver? Something like DS90LV011A (but not that exact one). Many SOT-23-5 all with the same basic pinout for driver or receiver.

Pins 3 & 4 are the differential pins so you may be measuring overvoltage for an LVDS pin relative to GND (pin 2)? (The LVDS maximum voltage is not referenced to VCC in these chips, although I don't know how they are actually implemented.)

However if this was the case then you'd also expect to be able to measure some protection circuitry directly between pin 1 (VCC) and pin 5 (the LVTTL single-ended pin), and probably also undervoltage protection between 3 & 4 and GND. Also, I have no idea why these boards might need this.

The only similar (equally wild guess) option there would be an op amp, which are available in similar SOT-23-5 standard pinouts and pins 3 & 4 are typically the op amp inputs.

[3DMovieList]

Thanks for the suggestion projectgus.  At this point anything is possible.  ;-)
I did some further testing on the board yesterday to check the chips connections to other points on the board.
Pin 1 and pin 4 of the J516T chip are connected together, and together connect to a zero ohm resistor (R394) (I think effectively a fuse), but it doesn't seem to connect to anything else on the board, which is really weird.  The only thing they might connect to is a connection on the thin ribbon cables that run to the LCD panel, but I'm unable to probe those.
Pin 3 goes to one pin on the input connector CN2 (pin 3) (which I think is a power supply pin, but I'm yet to confirm) on the other end of the board and there is one bypass capacitor (C305) on the line near the input connector.
Pin 5 goes to three sequential pins on the input connector CN2 (pins 68,69,70) on the other end of the board and it also connected to 8 bypass capacitors (C304,C316,C323,C326,C333,C336,C343,C346) going to ground along the way.
Pin 2 is ground.
I've captured a high-resolution image of the board, which helps understand the configuration.  The J516T chip is removed in this image.

Unfortunately, it seems the board is multi-layer, so this top view doesn't tell the full story.
But... given it seems there are only bypass capacitors connected to the J516T chip, to me it seems likely to me that this is a Transient Voltage Suppression (TVS) chip.

Given it's only one of the diodes (pin 1) that is malfunctioning on the bad J516T chip (which should have a forward voltage of 0.7V) I might just solder the chip back with pin 1 disconnected, and see if this allows the LCD panel to behave correctly.  I could solder a small external signal diode between pins 1 and 2 to replace the missing diode in J516T, but I wonder if that is absolutely necessary.  I was thinking of drawing up a simple schematic to illustrate what the circuit looks like. 
But as you can see from the circuit board image, the J516T chip is the only chip on the board.  Everything else is just capacitors and resistors.

All comments, ideas and theories very welcome!  Thanks for reading this far!

[3DMovieList]

3DMovieList, can you measure the height/width/length of the package?

I've measured the J516T chip with a vernier and the measurements are:
3.0mm (L) x 1.6mm (W) x 0.75mm (H)

[3DMovieList]

Good news!  There is LIFE in the TV!!!
I borrowed a "J516T" chip from another 65UB980T that I am repairing and that has brought the 84UB980T TV to life!
More details to follow, but replacing three SMD caps, and the "J516T" chip has brought the TV to life.

[3DMovieList]

So the 84" TV is alive, but actually there is one remaining problem with the TV that I need to solve to get it fully operational.
The mainboard also has a problem.  The image you saw earlier is with a borrowed mainboard from another TV.
So, the TV had four problems, of which I've solved three so far: (a) blown fuse on the T-CON board, (b) broken and twisted trace on one of the T-CON board flex cables, (c) the three caps and J516T chip on the LCD driver board, and finally (d) a mainboard problem (yet to be solved).
I'll report more as things progress!   

[3DMovieList]

I've found another two useful reference documents that can hopefully help solve faults on the UB980 range of LG 4K 3D TVs:
1. Full service manual for the LG 98UB980 (98" 4K 3D TV)
    https://www.manualslib.com/download/1163299/Lg-98ub9800.html
    This document is very similar to the 79UB980T service manual mentioned earlier in this thread.
2. "Novacom Collection of OLED/LED/LCD TV Repair Tips V5.0"
    https://lcdledsolutions.com/circuitpdf/pdf/V5%20OLED.LCD.pdf
    From page 76, there are further debugging information for the LG 84UB9800 TV.
    Firstly, the interconnect diagram, and then in following pages a series of Main Board Component Voltage measurements for chips all over the main board.

Unfortunately, this information hasn't helped solve the UB980 mainboard problems as yet.

[3DMovieList]

Some more useful information for debugging/repairing LG TVs.
The following document is titled: "LG - OWNER’S MANUAL - EXTERNAL CONTROL DEVICE SETUP"
http://www.proaudioinc.com/Dealer_Area/RS232.pdf
It contains the IR remote control codes, including the elusive code for the "ADJ" (adjust) button on the LG Service Remote Control, which isn't generally available for purchase.
This page on Reddit suggests how to generate this code using ESPhome:
https://www.reddit.com/r/OLED_Gaming/comments/vuphk0/does_anyone_have_the_ir_codes_for_lg_service/

[3DMovieList]

I found another interesting repair resource for the UB980 range of LG 4K TVs.
This video shows a repairer reflowing the CPU which returned the TV to life.
  (the reflow occurs at 5:20)
Prior to this process, the repairer checks various voltages to confirm correct operation of the various power supply sections.
In his case, the voltages were all OK, which led to the CPU reflow option, which was successful.
In my case the voltages are all OK as well, however reflow has not worked.  There's one more thing I want to try...