LG W2234S Repair Help...

[docksteaderluke]

I have an LG W2234S that isn't working. The screen turns on and shows the LG logo and then about 2 seconds later turns off (black screen, no power LED). Then it turns on again and shows the logo (and power LED is on)... Repeat.

I've pulled it apart and none of the caps are bulging or leaking. The big filter cap (150uf 450V) only charges to about 160V though... I feel like this could be a problem? I've heard also that the u108 chip (25L1005) on the logic card can get corrupted. I'm not sure this is the issue since it does actually display the LG logo for a few seconds but I could be wrong.

Any help would be greatly appreciated!

[PlainBill]

The symptoms could indicate bad caps, or possibly another problem.

1. Please provide sharp, well lighted pictures of top and bottom views of the power supply, and a top view of the signal card. Take the pictures from as close to directly over the cards as possible. A resolution as close to 2000 x 2000 is also helpful.

2. For best results, the test you described above should be done with the monitor hooked to a computer displaying a stable image - a the desktop, for example.

3. Please confirm that the backlights and power light go off at the same time, and come back on without any action on your part.

PlainBill

[docksteaderluke]

Thanks for your help! In response to your questions:

1. See the attached photos.

2. Unfortunately I can never get to the point where I have a stable image since the monitor is turning on and off every 2 seconds (literally). I did perform the same measurement anyways. When the screen is powered on it measure 148Vdc across the big cap and when it's off it's 157Vdc across that same cap.

3. The backlight and power LED do infact turn both on and off at the exact same time. The controls are unresponsive during the time that the monitor is on.

[seanc]

If the power light is staying on, but the *backlights* are turning on and off every 2 seconds then I would go after the CCFL cables - one of them is likely loose/disconnected from the bulb. Had the same issue with a W2043S

You can confirm that by hooking up known good bulbs to the inverter.

[docksteaderluke]

Nope... the power LED is turning off too... Both the power LED AND the backlight are turning on and off at the same time...

[PlainBill]

Originally Posted by docksteaderluke

Nope... the power LED is turning off too... Both the power LED AND the backlight are turning on and off at the same time...

Thanks for confirming that. The 'smoking gun' in this case is the voltage across the large cap. If Charlottetown, PE is tied into the Canadian power grid, you have given a very solid clue as to the problem. The voltage across the large cap should be 1.4 x your AC line voltage, and quite stable whether the monitor is on or off. With the variation you describe, either you have SOMETHING drawing way too much current, a thermistor that is not working, or the main filter cap is bad. Su'scon does not have a good reputation.

There are a number of ways to test this theory. Choose one or more of the following methods.

1. Disconnect the signal card from the power supply. Measure the voltage across the large cap. It should be VERY close to the 1.4 x line voltage figure (for 120 VAC the figure is 167 volts), and stable. If it is much lower, CAREFULLY run your hands near the board, looking for a source of heat. Do NOT touch any heat sinks!!!

2. With everything hooked together and the monitor plugged in and flashing on and off, short across the thermistor. That is the green device just below the fuse. Since I'm asking you to do this while power is on, I'm also asking you to be very careful. If you elect not to try this, I understand.

3. If you have access to a capacitor rated at at least 200 volts, and 100µF, wire it in parallel with the existing cap. If the flashing stops, that virtually proves the theory.

4. Buy or borrow an automotive tail light. My favorite is the 1157 dual filament bulb. With the power supply / inverter disconnected from the signal board, temporarily hook it across C819. Check the voltage across the main filter cap.

PlainBill

[docksteaderluke]

Hi PlainBill,

Here's what I've got:

1. With the signal card disconnected from the power supply I get 160.4-160.7V. Seems low.

2. Shorting the thermistor didn't do anything from what I could tell. Not sure if this was the desired effect or not...

3. 100uF, 63V is the biggest cap I have in my collection at the moment... so I didn't try this step.

4. I don't have an automotive tail light handy so I didn't try this step either.

[alexanna]

Wile you are wating for some more sugestions.Try reflowing solder in the red and blue circled areas.

[docksteaderluke]

Forgot to mention. In #1 I couldn't detect any noticeable heat sources...

[PlainBill]

Originally Posted by docksteaderluke

Hi PlainBill,

Here's what I've got:

1. With the signal card disconnected from the power supply I get 160.4-160.7V. Seems low.

2. Shorting the thermistor didn't do anything from what I could tell. Not sure if this was the desired effect or not...

3. 100uF, 63V is the biggest cap I have in my collection at the moment... so I didn't try this step.

4. I don't have an automotive tail light handy so I didn't try this step either.

Measure the voltage at the outlet (or the end of the power cord). Multiply it by 1.4. Subtract 1 from that number. 160 VDC corresponds almost exactly to a line voltage of 115 volts, well within the normal range.

One other test: Remove L802. This will disable the inverter. Test the monitor. Obviously you won't have backlights, but if the power LED behaves normally that will provide a small bit of information.

Unfortunately, if alexanna's suggestions don't help, I don't see many alternatives to either providing an artificial load, measuring the capacitance of the large cap, or replacing the large capacitor.

PlainBill

[docksteaderluke]

Thanks for the help guys... here's what I've got:

165.6V is the calculation I get.

When I resoldered those points the voltage went up to 161.5V(max).

Removing L802 didn't change anything. Still blinks.

I found a cap on another monitor that I have (whose backlight or inverter is gone) that is 100uF 450V. I'm not 100% sure if the cap is in perfect working order but I'm gonna try wiring it in parallel with the big cap after I re-solder L802 back on.

[docksteaderluke]

Decided to wait on soldering L802 back on. Soldered the second cap in parallel with the first and no difference (on and off in 2-3 second intervals). Still measuring about 160.5V across them.

With the second cap wired in I could hear a slight charging then tick (discharge) of the caps.

I'm trying to think of something else I can use as a load across c819. What kind of resistance are we talking about here (for the load)? I've got tons of resistors lying around that might do the trick.

[PlainBill]

Originally Posted by docksteaderluke

Decided to wait on soldering L802 back on. Soldered the second cap in parallel with the first and no difference (on and off in 2-3 second intervals). Still measuring about 160.5V across them.

With the second cap wired in I could hear a slight charging then tick (discharge) of the caps.

I'm trying to think of something else I can use as a load across c819. What kind of resistance are we talking about here (for the load)? I've got tons of resistors lying around that might do the trick.

One idea. Obviously you have the signal card connected. Do you also have the cable from the signal card to the LCD panel connected? SOMETHING is overloading the power supply or a power rail. I've been assuming (I know, I KNOW!!) it's the inverter, or a bad filter cap. If you haven't already done so, try this with the LCD panel disconnected.

Do you have a pair of resistors in roughly a 3:2 ratio? For example, a 3.3K and a 2.2K would be ideal.

PlainBill

[docksteaderluke]

When you say to try it with the LCD panel disconnected do you mean the VGA cable or the cables going from the inverter side of the power supply to the lcd?

Yes I have a 3.3k and a 2.2k right here.

[alexanna]

Disconnect the LDVS cable from the logic board to the LCD panel

[PlainBill]

Yeah, what Alexanna said.

PlainBill

[docksteaderluke]

Ok... gotcha. Did it and still no change.

[PlainBill]

Originally Posted by docksteaderluke

Ok... gotcha. Did it and still no change.

I'm getting very close to recommending a total recap, but let's try this first.

Work from the back of the power supply / inverter.

1. Take the two resistors and twist one lead of each together, but staggered so one lead extends beyond the other.

2. On the power supply / inverter lay a strip of tape - electrical, duct, or surgical between the pins of CN801 - the output connector - and the edge of the board.. You want to avoid accidental shorts between the leads of the resistors and the ground bus.

3. Solder the two twisted leads of the resistor together.

4. Solder the longer lead of the twist to pin 5 of CN801.

5. Solder the free end of the 3.3K resistor to pin 3 or 4.

6. Solder the free end of the 2.2 K resistor to pin 1 or 2.

7. Check your work; make sure you don't have any shorts.

8. Lay the power supply / inverter on an insulating surface - a piece of dry cardboard is fine.

9. Plug the 4 CCFL leads into their sockets.

10. Take a deep breath, cross your fingers, and plug the power cord into the power supply.

At this point the CCFLs should come on. If they are flashing at about the same rate as when the signal card was plugged in, the problem is the power supply / inverter card. If they are on steadily, the problem is the signal card.

Disconnect the power cord, unsolder the resistor network, and report the results.

PlainBill

[docksteaderluke]

Should the signal card be plugged back in for this?

I plugged it back in. Soldered up the resistor network and plugged both a VGA cable in and the power cable in.

What I got was a white screen for about a second, then the LG logo for a second or so and then off for a second or two. Repeat.

If I didn't have the signal card connected I couldn't see anything on the screen... it was just black and with no backlight (or power LED) either.

[PlainBill]

Originally Posted by docksteaderluke

Should the signal card be plugged back in for this?

I plugged it back in. Soldered up the resistor network and plugged both a VGA cable in and the power cable in.

What I got was a white screen for about a second, then the LG logo for a second or so and then off for a second or two. Repeat.

If I didn't have the signal card connected I couldn't see anything on the screen... it was just black and with no backlight (or power LED) either.

No, I didn't want the signal board hooked up. The resistor network was to turn the inverter on. Perhaps it would be a good idea if I also had you feed a voltage into the brightness input - pin 6 of CN801. The idea was to see if the inverter would work properly if it were fed appropriate voltages without the influence of the signal card.

Make up a voltage divider similar to the one driving pin 5. But this one should have a 3.9K resistor going to 5V and a 1K going to GND. The common point would go to pin 6 of CN801. And while power is applied, measure the voltages at pins 5 and 6, if you would.

PlainBill