Electronics      How-To


An "LED monitor" is really just an LCD flat panel monitor that's backlit by LED's.

This one developed corner shadows, where some of the LED's are not lighting up.

Is it fixable?  Let's see what we can find out.

CAUTION:  The circuits in a monitor may have high voltage.  The information in this article is not guaranteed to be correct, and it might not even match your monitor.  Do not attempt electronics repair if you lack the skill, knowledge, and experience.  Disclaimer.

(For non-electronics people, just buying a new monitor is the cheaper option anyway.) 

Reader-Supported Site

Articles like this one are possible only with the support of readers like you, when you use the links on here to purchase your gear. 

The small commissions from sales are what allow me to keep this site going.   Thank you in advance for your help.

In This Article

The Problem


Teardown Notes

Visual Inspection


Multimeter Time!

That Ribbon Cable

The LED's


The Problem

So there's an Acer flat-panel monitor that mostly works.  But the screen has a dark corner, like there's a shadow or something. 

For a while, the shadow would briefly go away when the monitor left sleep mode.  Gradually it became permanent.


There were two screws holding the back on. 

The plastic trim around the screen also had to be pried apart. 

Underneath the plastic housing, there was a metal cover on the main circuit board. 

There were three or four screws holding the cover to the mainboard.  Those hex-sided fasteners that hold in the DVI and D-SUB jacks had be removed, as well.  An adjustable wrench worked OK for this.  The actual DVI and D-SUB jacks stay attached to the mainboard. 

Teardown Notes

The power supply is in a separate box (power cord).  I didn't look at that circuitry.

What I'm going to call the "display driver board" is along the bottom, on the back of the panel itself.  Looks to be a bunch of resistor-capacitor networks on this board.  There are flat flex cables for each display drive group:  SD1, SD2, etc.  The test points on this board probably wouldn't give much information about the backlight problem.

The LED backlight driver is on the mainboard itself.  The actual driver IC is an EUP2589 with 28 pins (near the ribbon cable header).  This is supposed to be a pretty good chip, with short-circuit detection, over-temp protection, over-voltage protection, and current limiting.  (Bad caps might change all that, though.)

I'm glad there are voltage test points here.  The SMD's on the mainboard are often so small that you couldn't test them directly.  (Fine-pitch IC's, too.)  Voltage test points are a great idea, as we saw in the WWII Radio article.

By the way, many components are simply not there.  Three places for zener diodes, and they didn't use them:

Were these for another, better model?  No idea.  From what I gather, the schematics for these monitors (when you can even get them) are confusing because they don't exactly match the one you have.

Next up, some solder pads for an IC they decided not to use.  Instead we have what looks like a #1084 low-dropout linear regulator.  There's diode D2 going toward the input pin, then looks like diode D5 at the output.  The tab is actually "V out".  The SMD capacitor C6 goes from input to GND.  The 470 ohm SMD resistor (R13) goes from the "GND" pin to GND;  the way they have this configured with R12 looks like what I found in a datasheet for this regulator.

Table of Contents

Visual Inspection

So I checked the mainboard for obviously-fried components, bad solder joints, etc.

The underside of the board shows signs of heat: 

There's a surface-mount IC on the other side of this. 

That's a QFN48-packaged silicon rectifier that can handle up to 1 kilovolt at 1 amp.  They get hot, so the mild scorching could be normal. 

Notice the vias which go through the board;  these look like they're supposed to be test points.  And then we have the two larger test pads near pins 24 and 48 of the SM4005 chip.  The chip handles quite a bit of wattage.  QFN48's can develop bad solder joints after some thermal cycling.  So this might be something they'd want the ability to test with a multimeter.  (Smart design tries to minimize unnecessary replacement of surface mount IC's, whenever possible.  This is why you include things like protection circuitry and test points.) 

Next, I checked the electrolytic capacitors on the mainboard.  Sure enough, a couple of the caps were bulged. 

These might have caused that corner shadow.

Table of Contents


The bad caps were both 220 uF, 25-volt electrolytics.  One was Teapo brand;  on the other I didn't see a brand name.

The high melting-point solder made for a difficult repair.  A solder pad lifted.  I had to build a solder bridge from a cap lead to one of the circuit board traces.  The whole thing would have been easier if I'd used the correct soldering tools;  a better iron would have really helped.


Table of Contents

Multimeter Time!

Replacing the bad caps didn't fix the shadow problem. 

Before going to the LED's, I carefully tested some voltages on the mainboard.  With these types of things, a mistake can instantly short out something on the board and ruin it.  There may also be lethal voltage hazards.  Definitely not a good place for beginners.  (Disclaimer, again.)

Starting with the SM4005 QFN chip, I checked a few test points for DC voltage.  Without a service manual we don't know if these are the factory-correct voltages, but here they are.   

Some of these voltages could vary with brightness adjustment.

Pin 1:  0 volts
Pin 2:  0.483 volts
Pin 3:  1.222 volts
Pin 4:  no test point?
Pin 5:  no test point?
Pin 6:  3.328 volts
Pin 7:  3.328 volts
Pin 8:  3.304 volts
Pin 9:  3.303 volts
Pin 10:  1.095 volts
Pin 11:  1.096 volts
Pin 12:  0.950 volts
Pin 13:  0.950 volts
Pin 14:  0.011 volts
Pin 15:  0 volts
Pin 16:  -5 volts (didn't write down the exact decimal.)
Pin 17:  -5 volts
Pin 18:  4.33 volts
Pin 19:  4.33 volts
Pin 20:  5.82 volts
Pin 21:  5.80 volts
Pin 22:  28.7 volts, dropping to -5 volts.  The multimeter (10 megohms to ground) triggers some kind of flip-flop (high to low).
Pin 23:  28.7 volts, dropping to -5 volts.  Here again... interesting;  is this for sleep mode / screensaver?
Pin 24:  -5.21 volts steady
Pin 25:  28.85 volts steady
Pin 26:  28.85 volts steady
Pins 27-29:  unknown / no test points?
Pin 30:  1.216 volts
Pin 31:  1.214 volts
Pin 32:  0.213 volts
Pin 33:  4.88 volts
Pin 34:  4.88 volts
Pin 35:  no test point?
Pin 36:  12.46 volts
Pin 37:  no test point?
Pin 38:  1.243 volts
Pins 39-40:  no test points?
Pins 31-42:  no test points?
Pin 43:  4.93 volts
Pin 44:  4.96 volts
Pin 45:  3.293 volts
Pin 46:  1.125 volts
Pin 47:  1.523 volts
Pin 48:  no test point?

The pads near pin 24 and 48 are GND.  Note that some of the "no test points" could actually be tested if you pick the correct solder joint.  Steady hand!

By the way, I'm sure some of the pins would have AC voltage to them, but I didn't test that.

The header for the LED ribbon cable has two test points.

Pin 1:  18.19 volts
Pin 10:  26.40 volts

Capacitor CE1 (220 uF electrolytic):  19 volts

Capacitor near center of board (220 uF electro.):  12 volts (didn't write the decimal, might have been 12.4).  Looks like the positive lead has continuity with Pin 36 of the SM4005 chip, basically a 12-volt rail. 

By the way, when troubleshooting digital circuits, don't use a junk multimeter.  Get a good one.  Sometimes a very slight change in voltage or current can trigger a logic switch.  So, if you're going to do this stuff, get yourself a good electronics meter...

Table of Contents

That Ribbon Cable

It's very possible that the bad caps caused LED's to go bad, too.  Too much current can fry LED's in a matter of microseconds.  It could have been a change in signal that caused one of the chips (such as that EUP2589) to overdrive the LED's.  Reverse polarity can also damage LED's. 

By the way, remember that flat cable that goes from the board to the LED strips?  The voltages at Pin 1 and Pin 10 may have been a clue. 

The one was at 18.19 volts relative to ground. The other was at 26.40 volts.  Subtract 18.19 from 26.40 and you get 8.21.  Divide that by two and you get about 4.11, which is about right for the voltage drop across a high-brightness LED.  So there's the clue that two LED's are bad.

The higher voltage of 26.40 means less of a voltage drop.  That's because there are fewer working LED's in that strip.  If a whole strip of these LED's were bad, we might be seeing a voltage of perhaps 62 volts.  The EUP2589 can output as much as 65 volts, so that's very possible. 

Just remember, when current flows, there is a voltage drop.  You won't read the full voltage unless there's an open circuit.  But if there were an open circuit, we wouldn't be able to see there's something amiss with the voltage drop.

It took a while before I thought of what those voltages meant.  So the next step, where I checked the LED's, was actually first.  And the thing is, even if you know there are two bad LED's, you want to know which ones so you can replace them.

Table of Contents

The LED's

Replacing the bad caps didn't fix the shadow problem.  (They had to be replaced anyway, though).

So I set the multimeter to "diode test" to look for shorted or open LED's. 

The hardest part was actually getting to the LED's.  They're soldered to a strip along the left-hand edge of the monitor (as you're looking at the screen).  It's difficult to get at without lifting some tape and prying apart a bunch of snap-together plastic and thin, stamped metal. 

So, the LED tests.  Apparently the diode tester applies enough voltage to light them up!

Mostly good, except for two of them which didn't light up.  And sure enough, they're at the lower left-hand corner of the screen. 

I didn't measure the LED dimensions while I had it apart.  Just a guess here, but they look like 7020 LED's.

Table of Contents



This was a repair attempt on an Acer flat-panel monitor.  The LED backlights are not fully working, which means there is a corner shadow.

I found what was wrong.  There were a couple of bad capacitors, but two LED's were also burned out.  If I can get the right LED's, maybe I'll try replacing them.

This was a fun project, and I learned a couple things about surface-mount electronics.  Hopefully you did, too.

If you found this page useful, informative, or entertaining, please help me out by purchasing your stuff through these links.   Your help is greatly appreciated;  it allows me to keep this website on-line and adding more articles to it.

Thanks for visiting this page!

Contact me:

3 p o.t o . 1 2 0 s t u d i o.. c o m

This won't directly copy and paste.  Please manually type it into your mail program.
No spaces between letters.

Home Page

Audio Guide


Copyright 2015-2018. All rights reserved.


Back to Top of Page