LG 27GL63T-B completely dead

[Pandor]

its funny how we now have 3 'completely' different dumps of the firmware. The EEPROM content is also completely different.
But upon some investigation, and trying to understand how the CPU/Scalar works, I believe the EEPROM only contains the user settings for the screen (source: HP Compaq LA2205wg 22-inch Widescreen LCD Monitor)
I tried searching for the latest firmware, using these techniques: RFC 3514: How Smart Is My Monitor? (Part 1) (davidmanouchehri.com), Discovering a Revoked LG Monitor Firmware Update | by David Zech | Medium, [Monitor Firmware] LG DualUp/MST9U Firmware Hacking - Blur Busters Forums This do no longer work with the latest Win_OSC_8.26. though.. no visible strings in the exe. I'm not sure if the 27GL63T-B ever got any updates (the firmware in the GD25Q80C does mention the gl650f product name, so same firmware, just the secondary HDMI port and components not soldered?)

I can't find any version number related string in my or the posted firmware files in this thread, so there's also that... MStar seems to be verry secretive about their scalar chips, so that also doesn't help...

We're really going down a rabbit hole here.. Why, if all power stages are fine, would it not enter its standby state...

One thing is certain, i'll never buy any LG products ever... my bedroom LG 4K tv's also acting up... LG's now just creating disposable products it seems.

[Pandor]

markismus here are my dumps, from the desoldered chips.

[Pandor]

Did you try to erase, program and verify it with one of the uploaded bin-files? You can do that without desoldering,right?

I think there are to much devices on the 3.3 line, and my programmer can't handle that. I have a hot air station, so desoldering is not that big of a deal. Judging from other LG monitor model malfunctions and some research, it could be possible that the eeprom actually contains a bootloader/management for the cpu, so that could possibly be the culprit. But i'm speculating..

[markismus]

, but for the IC 202 flash, i get FF. I guess i'll have to desolder.

Did you try to erase, program and verify it with one of the uploaded bin-files? You can do that without desoldering,right?

[Pandor]

....
Testing MP2315 a.k.a. IC402.
I have 18V on pin 2 (IN).
I used a USB buck converter (See image) set to 2.0V and connected the negative to GRD (The metal of one of the connectors, JK102) and the postive to pin 6 (EN/SYNC).
I get 12.0V on pin 3 (SW).

So this one works. Pandor Do you get these test results, too?

Currently all three buck converters work as expected.

Next step is apparently flash memory.....

i have not tested IC402, as that is only for the screen power, and i'm not even getting standby. But i could test it.
i suspect it'll function as it should. I was more focussing on atleast getting the standby led to turn on, and have the board in a standby state. I just tried reading both the eeprom and flash in place with a clip, and for IC201 i get something, but for the IC 202 flash, i get FF. I guess i'll have to desolder.

[markismus]

I used a CH341A Mini Pro programmer and read the EEPROM IC201 (24C32A, 4KB) and the SPI Flash IC202 (GD25Q80C, 8MB). The data didn't appear totally random. (See attached IC201READ4KB_24C32 and IC202READ1MB_GD25Q80.rar.)

Comparison of the result of GS25Q80C reading 1MB with the binary file uploaded a few posts back shows differences but also a lot of identical data. So it rather looks like different firmwares/settings than a corrupted IC202.
No idea what the data on the EEPROM is meant to be.

Could anyone post a dump of their IC201 EEPROM for comparison?

[markismus]

Testing MP2315 a.k.a. IC402.
I have 18V on pin 2 (IN).
I used a USB buck converter (See image) set to 2.0V and connected the negative to GRD (The metal of one of the connectors, JK102) and the postive to pin 6 (EN/SYNC).
I get 12.0V on pin 3 (SW).

So this one works. Pandor Do you get these test results, too?

Currently all three buck converters work as expected.

Next step is apparently flash memory.....

[markismus]

With the microscope I am also able to read the IC101:

IC101

SD642A0

TI 99A6801 G4

12 channel 1:2 or 2:1 bi- directional multiplexer/demultiplex

So if the HDMI would crap out on a later date, I'll have to get new skills to replace that IC.

[markismus]

So the next thing on the flowchart is the flash memory. I have the ICs:

IC201	FT24C32A	9eHBAC		Two-Wire Serial EEPROM
IC202	BE1941	25080CSIG	(Tags rechecked with microscope.)	3.3V Uniform Sector Dual and Quad Serial Flash
IC500	98TG4	CISZ	TPA6138	40-mW, stereo, analog input headphone amplifier with adjustable gain
IC600	MPS1928	MP3378R	K4871538X	24V, 4-Channel WLED Controller Plus High-Efficiency Buck Converter

The IC202 is called the flash memory, however the EEPROM seems like flash memory to me, too. If I remember correctly, I have a flash memory programmer/reader laying around somewhere. So I should be able to check whether those still work...or maybe rig up an Arduino to read them....

As the LED's are out, it could also be the IC600. I remember that it used to get really hot. Which it doesn't do anymore, so either that's been helped with the replacement of the AKX and AGC buck converters or it's died on me, too.

Pandor Did you replace the AGC buck converter(, IC402 on my board&image from a few posts back)? I did a crappy job at it without the microscope, so I am wondering whether I shouldn't replace that one again.
I could give it 2.5V on EN and see what SW does. If it changes to 5V, I should expect it to work.

How could we check IC600, though?
All three IC's are not that expensive, though. Just hard to see which parts offered on AliExpress are actually the same and which aren't.

Looking at CapLeaker's post #4, we could also try the 4-pin header (P200) and see whether we can get a TTL-connection going. Google doesn't really help for now for a working approach. Up till now I only used 3-pins for a serial connection and this page suggests there could even be 5 pins. Well, I begin with looking at the voltages, so that I at least don't damage my interface-board.

[Pandor]

Pandor Resoldered AKX with stereomicroscope and now it measures

PIN	1	2	3	4	5	6	7	8
Reference	FB2	EN2	SW2	EN1	GND	SW1	IN	FB1
Voltage	0.6	3.3	1.17	3.3	0	1.8	5	0.6

So that seems good. Also replaced the crystal oscillator with a new 12.000 Hz one.

AQT changed slightly to

PIN	1	2	3	4	5	6	7	8
Reference	AAM	IN	SW	GND	BST	EN	VCC	FB
Voltage	0.3	19	3.3	0	8.3	2.5	5.0	0.8

AGC is now

PIN	1	2	3	4	5	6	7	8
Reference	AAM	IN	SW	GND	BST	EN	VCC	FB
Voltage	0	18	0	0	~0	~0	~0	~0

I guess we're in the same boat now. I injected 1.8v and it was only pulling a few mA. In a last attempt I reflowed 401 and now i have 1.8v back. So currently all voltages seem correct, as yours. Still no life. Firmware maybe? I did not replace or check the crystal yet.

[Pandor]

Hello,

I have the same error, the monitor has just "died".

Symtomp: The monitor has stop working suddently, it does not show any light neither work any button, completely dead.

The power supply works fine.


Inspecting the PCB board with a thermal camera the IC401 shows a higher temperature than the rest of the board.

My readings for the IC401 are:

Pin 1	Pin 2	Pin 3	Pin 4	Pin 5	Pin 6	Pin 7	Pin 8
0,7V	3,3V	1,2V	3,3V	Ground 0V	5V	5V	1,6V

Does those readings OK, or should i look for another component?

Edit -> The IC400 just burned, i will replace that, does anyone know the part number?


Thank you!

Are you sure you're getting 5v on pin 6 (SW1), did you doublecheck the input at the coil L408? That is way to high! Should only be 1.8V. I your reading is correct you're sending 5v into the CPU DRAM/IO...

[faza1984]

Hi all. I want to share my success story. I got a monitor with the same symptoms - totally dead. I ordered new parts for IC401 and IC402. When I tried to replace IC401, it was stuck on the board. Even 450 temperature did not help to remove it. So I used a polisher machine to remove it (be careful, you can easily break the board). Next, I soldered new parts. After I'd done, I checked the voltage and got 1,8v but no 1,2v. I opened the datasheet for IC401, and 1,2v should be on 3rd pin. I double-checked the microscope and discovered that the leg did not connect to the board. I soldered twice leg three unde microscopes and check. My monitor turned on!!! Thank you all who shared their story, and good luck!

[Pandor]

No joy, though.

No life at all? Atleast you're getting all the correct voltages afaict...
This monitor is a tough nut to crack it seems. I was thinking about getting another cpu on alli, but they're like 30€ and no gurantee. Thats a lot to spend on this POC..

[Pandor]

markismus posted some great pictures with correct IC labels and datasheets, one page back.

[Aertop]

Hello,

I have the same error, the monitor has just "died".

Symtomp: The monitor has stop working suddently, it does not show any light neither work any button, completely dead.

The power supply works fine.


Inspecting the PCB board with a thermal camera the IC401 shows a higher temperature than the rest of the board.

My readings for the IC401 are:

Pin 1	Pin 2	Pin 3	Pin 4	Pin 5	Pin 6	Pin 7	Pin 8
0,7V	3,3V	1,2V	3,3V	Ground 0V	5V	5V	1,6V

Does those readings OK, or should i look for another component?

Edit -> The IC400 just burned, i will replace that, does anyone know the part number?


Thank you!

[markismus]

No joy, though.

[markismus]

Pandor Resoldered AKX with stereomicroscope and now it measures

PIN	1	2	3	4	5	6	7	8
Reference	FB2	EN2	SW2	EN1	GND	SW1	IN	FB1
Voltage	0.6	3.3	1.17	3.3	0	1.8	5	0.6

So that seems good. Also replaced the crystal oscillator with a new 12.000 Hz one.

AQT changed slightly to

PIN	1	2	3	4	5	6	7	8
Reference	AAM	IN	SW	GND	BST	EN	VCC	FB
Voltage	0.3	19	3.3	0	8.3	2.5	5.0	0.8

AGC is now

PIN	1	2	3	4	5	6	7	8
Reference	AAM	IN	SW	GND	BST	EN	VCC	FB
Voltage	0	18	0	0	~0	~0	~0	~0

[Pandor]

...
My 401 has 0.08V and 1.2V on resp. SW1 and SW2. Since EN1 = 1.7V and EN2 = 0.3V, I would expect them to behave differently. I'll check later when I have more time. The IN = 2.15V instead of the expected 5V, so that could also explain the odd behavior.
my 402 has 0.8V on EN/SYNC, so it should not output anything on SW. However, it also doesn't have 5V on VCC, so it doesn't work anyway, right?

Pandor Looks like we at least have the same problem situation.

I've bought a 20k crystal oscillator. I'll replace that one first, before checking the buck converters again.

Those readings seem to be way off. after replacing 401, i get:
IN=5v
EN1 and EN2=3.3v
SW2=1.16v (close enough)

But SW1 goes straight down to 0.08v. so in my case, something is overloading the SW1 side, making it go into protection mode.
I have a bad fooling the CPU is taking the rail down..

[Pandor]

I believe that would be the AKX. The ic numbering might be different, but the general power rails seem to match. The AKX is the only dual output, so hence my assumption.

For me the AKX seems to start at 2v output briefly, but goes down to 0.08v on SW1. So it could be going into safe mode/protect. The weird thing is that the SW1 side does not seem to have a short on my board, but the SW2 side has a rather low resistance to ground. If anyone can confirm it should be 2.5v, i might try injecting that after the coil.

markismus the abcence of 5v input at the AKX is worrying though. In my case the 5v is present.

the powerrails actually match up in the service manual. Only IC401 seems to be IC403 in the schematics. This also explains why IC 402 would be low, as that seems to generate the 12v for the panel.

So i seem to be missing 1.8v, or rather it drops out at power up. I'll dig some more on that rail.

I'll just inject 1.8v into that rail and see what gets hot, and the power draw. Something on that rail seems to bring it down as soon as it powers up. Resistance on the unpowered rail seems fine though.

I just hope its not the CPU, as the 1.8v seems to split into VDD_DRAM and VDDIO and those just go through a capacitor cluster, and then into the cpu..

[Pandor]

As the numbering is different from ours, I used the datasheets. However, I am interested now wat the analogue to ic705 would be.

I believe that would be the AKX. The ic numbering might be different, but the general power rails seem to match. The AKX is the only dual output, so hence my assumption.

For me the AKX seems to start at 2v output briefly, but goes down to 0.08v on SW1. So it could be going into safe mode/protect. The weird thing is that the SW1 side does not seem to have a short on my board, but the SW2 side has a rather low resistance to ground. If anyone can confirm it should be 2.5v, i might try injecting that after the coil.

markismus the abcence of 5v input at the AKX is worrying though. In my case the 5v is present.