[LG TV] No Backlight, EAX65423701 2.1 PSU defect? Where to look?

[roffelkartoffel]

Hello everyone,

I have an LG 42lf5809 TV with no backlight. I have connected a raspi as a mediacenter. I sometimes had problems the TV stayed "off" when disabling the PIs screensaver but a power cycle of the TV worked. Now it just lights up for a second or two (I am able to see the LG bootup logo). Picture and sound is present, I can see the image as a "reflection" when using a flashlight in front of the screen.

Ive looked at the caps (no bulging), checked them in circuit with a cheap tester only to see good caps after I desoldered them. Now I am out of ideas and thinking about replacing it. I found a (promised as working) replacement on ebay but unfortunately the buyer does not accept returns so I hesitate since I lost money this way on an other project (backlight completely off but replacement did not work as hoped).

Have you further ideas I can look into?

Cheers from Germany

[wraper]

Most likely LED backlight failure, a very common fault of LG LED TV. To fix it, LCD panel should be disassembled and all of the LED strips replaces. Partial replacement most likely will cause repeated failure after a while. Practice shows that by the time backlight fails, all of the LEDs even though still work have cracks in the compound due to overheating and thermal cycling. To avoid degradation of the newly installed backlight, it's recommended to reduce LED current by about 25% by increasing resistance of the current sense resistor. Or just not setting backlight brightness above 75%.

[roffelkartoffel]

Thanks for your suggestion.

I think the strips are the culprit on the oter TV but I'm curious if replacing the strips solves the issues. I've measured the voltage at the backlight output and it should be around 200V (stated by one of the yt sources dealing with such issues) but it's close to zero (with a bump when the leds are on, unfortunately by DMM is to slow to capture the actual voltage)

[wraper]

Thanks for your suggestion.

I think the strips are the culprit on the oter TV but I'm curious if replacing the strips solves the issues. I've measured the voltage at the backlight output and it should be around 200V (stated by one of the yt sources dealing with such issues) but it's close to zero (with a bump when the leds are on, unfortunately by DMM is to slow to capture the actual voltage)

Protection circuit detects abnormal condition because of faulty LED backlight and LED driver shuts down.

[roffelkartoffel]

That sounds reasonable. I was able to find a replacement in china for few bucks so I'll try them.

[Labrat101]

The Backlight voltage is printed on the board usually on the left side . Normal voltage with no
LED's connected  is 90v - 100v  The LG 42inch has 6 rows of 10 Leds depending on the model .
Recommend disconnecting the LED plug from the Power supply and measure again .
should be 90 volts to each pair bank 30 LEDs to each Bank   if not the driver chip may also be damage or faulty . The protection diodes also can fail they are the 2 paralleled diodes , 4 in total

The LEDs fail mostly shorted and when enough burn out the remaining one blows open circuit.
This circuit diagram is basic to most 42 inch to 50 . the only real differences is the booster coils changes from yellow or blue .

[wraper]

The Backlight voltage is printed on the board usually on the left side . Normal voltage with no
LED's connected  is 90v - 100v  The LG 42inch has 6 rows of 10 Leds depending on the model .
Recommend disconnecting the LED plug from the Power supply and measure again .
should be 90 volts to each pair bank 30 LEDs to each Bank   if not the driver chip may also be damage or faulty . The protection diodes also can fail they are the 2 paralleled diodes , 4 in total

The LEDs fail mostly shorted and when enough burn out the remaining one blows open circuit.
This circuit diagram is basic to most 42 inch to 50 . the only real differences is the booster coils changes from yellow or blue .

BD9483 LED driver datasheet https://docs.rs-online.com/7df1/0900766b815647f7.pdf says that it will shut down if OVP protection kicks in. Exactly what will happen if LED becomes open circuit.

The Backlight voltage is printed on the board usually on the left side . Normal voltage with no
LED's connected  is 90v - 100v

The voltage that is printed on the PCB is a nominal voltage with a working backlight connected. Not open circuit voltage.
Since backlight works for a moment when switching on the TV, I guess the backlight on one of the channels is open and OVP protection kicks in.

[Labrat101]

Correct .. the voltage should read about 89 to 90volts with the LEDs not connected. On each channel.
You have one open circuit..
When you open the screen make sure not to break the retaining clips . Were cotton gloves when handling the 3 Polarized sheets and don't separate them or change there orientation. 
Bending the screen or damaging the edges will result in a dead monitor.. the cables must be removed from the screen and a small peice of anti static tap placed on the panel driver plates.
Don't touch the small copper wires around the edges..
Have fun and good luck the screens are very easy to damage. 
Also don't break the 9 or so white spikes in the reflective panel.

[wraper]

Correct .. the voltage should read about 89 to 90volts with the LEDs not connected. On each channel.
You have one open circuit..

No. That circuit is not even supposed to work with open circuit, and LED driver should shut down due to OVP is you try to do so. Not to say that nominal voltage is different for each half since a different amount of LEDs are connected to them.

[Labrat101]

No . Your not correct. In open circuit the protection should not operate.  Will give minimum set  voltage. Just under 89v per channel.. I have repair 1000s of these monitors . Over the years .
Your meter or test gear may not be suitable .
The voltages per channel may well differ per model . Just check the colour of the 2 booster
coils if both are the same colour then the outputs are the same .. If 1 is blue and the other
is yellow .. Then one channel is 100 - 120 volt  or as specified on the panel version .
These panels should be checked with a bench meter that can read at least 300hz on DC
other wise you will get false reading .and also have a good led panel tester .
 

[wraper]

No . Your not correct. In open circuit the protection should not operate.  Will give minimum set  voltage. Just under 89v per channel.. I have repair 1000s of these monitors . Over the years .
Your meter or test gear may not be suitable .

Maybe read the datasheet? There is no voltage regulation/sense in that driver chip. Only OVP which shuts it down. Not to say boost converter is not supposed to operate without any load, nor it can output regulated voltage without voltage sense.

[Labrat101]

No . Your not correct. In open circuit the protection should not operate.  Will give minimum set  voltage. Just under 89v per channel.. I have repair 1000s of these monitors . Over the years .
Your meter or test gear may not be suitable .

Maybe read the datasheet? There is no voltage regulation/sense in that driver chip. Only OVP which shuts it down. Not to say boost converter is not supposed to operate without any load, nor it can output regulated voltage without voltage sense.

I guess you don't repair these monitors you can say what you like . but I would not let you within 1000yrd of our lab

[wraper]

No . Your not correct. In open circuit the protection should not operate.  Will give minimum set  voltage. Just under 89v per channel.. I have repair 1000s of these monitors . Over the years .
Your meter or test gear may not be suitable .

Maybe read the datasheet? There is no voltage regulation/sense in that driver chip. Only OVP which shuts it down. Not to say boost converter is not supposed to operate without any load, nor it can output regulated voltage without voltage sense.

I guess you don't repair these monitors you can say what you like . but I would not let you within 1000yrd of our lab

If you are so smart, please explain us a mechanism how this can output the voltage you claim without any load present?

[Chalcogenide]

From the datasheet OVP appears to have an upper and lower threshold, so OVP will kick in in case of open circuit, when the output voltage rises above the higher OVP threshold, but then the resistor network of the OVP itself will slowly cause the output voltage to drop below the OVP lower threshold, and the chip will turn back on, and the cycle repeats again and again. This effectively causes a voltage to be present at the output, albeit with high ripple and not well regulated.

[wraper]

From the datasheet OVP appears to have an upper and lower threshold, so OVP will kick in in case of open circuit, when the output voltage rises above the higher OVP threshold, but then the resistor network of the OVP itself will slowly cause the output voltage to drop below the OVP lower threshold, and the chip will turn back on, and the cycle repeats again and again. This effectively causes a voltage to be present at the output, albeit with high ripple and not well regulated.

From the datasheet

(*1)…As OVP is detected, the output GATE=L, DIMOUT=L, and the abnormal counter starts
(*2)…If OVP is released within 4 clock of abnormal counter of the GATE pin frequency, the boost operation restarts.
(*3)…As the OVP is detected again, the boost operation is stopped.
(*4)…As the OVP detection continues up to 4 count by the abnormal counter, IC will be latched off. Both channels will be
stopped. (GATE1=GATE2=L, DIMOUT1=DIMOUT2=L)
(*5)…As the latched off, the boost operation doesn't restart even if OVP is released.
(*6)…The STB=L input can make IC reset.
(*7)…It normally starts as STB turns L to H.
(*…The operation of the OVP detection is not related to the logic of PWM.

Also as OVP is about 3v typ., considering resistive dividers it should activate at about 176V for one channel and 118V for another. As OVP input has 100mV typ. hysteresis, short shut down without any load other than OVP resistive divider should not be enough to return voltage to acceptable level, therefore it should shut down completely.

[Labrat101]

No . Your not correct. In open circuit the protection should not operate.  Will give minimum set  voltage. Just under 89v per channel.. I have repair 1000s of these monitors . Over the years .
Your meter or test gear may not be suitable .

Maybe read the datasheet? There is no voltage regulation/sense in that driver chip. Only OVP which shuts it down. Not to say boost converter is not supposed to operate without any load, nor it can output regulated voltage without voltage sense.

I guess you don't repair these monitors you can say what you like . but I would not let you within 1000yrd of our lab

If you are so smart, please explain us a mechanism how this can output the voltage you claim without any load present?

If you are so smart, please explain us a mechanism how this can output the voltage you claim without any load present? 
Better still You explain why as you have  Posts: 13427 . Have never worked on these systems and seem
to know nothing more than comment telling every one else how much BS you know .
How do you think LG checks these Panels on their test lines .
 O maybe some just sticks there tongue on it to see if it Hot      
With the Right Test equipment there is No Problem checking these outputs .
 If you have a problem with this I suggest you take it up with LG tech dept .
 Subject Closed . 

[wraper]

Better still You explain why as you have  Posts: 13427 . Have never worked on these systems and seem
to know nothing more than comment telling every one else how much BS you know .

Or maybe because you have no clue what to say, you instead resort to ad hominem fallacy.

How do you think LG checks these Panels on their test lines

Maybe you mean a LED driver? With dummy load attached. The same as PSU. Do you think you can switch on something like this with no load whatsoever and call the test passed?