Samsung 2493HM No Backlight

[Cateyerr]

Hi all,

I sat down at my computer tonight and moved the mouse to wake my monitor (Samsung 2493HM) and I heard a very audible "pop" sound and the monitor stayed black - but the power LED was solid. I tried cycling the power and reconnecting the hdmi cord several times before starting the google process which led me here.

I have read these threads already:
https://www.badcaps.net/forum/showthread.php?t=14483
https://www.badcaps.net/forum/showthread.php?t=15843
https://www.badcaps.net/forum/showthread.php?t=20520
https://www.badcaps.net/forum/showthread.php?t=13366
But I have not been able to diagnose it yet.

Things I have tried/noticed:

• The monitor is still displaying a picture, I can see it when a bright light is shone on the display. So I am assuming the problem is in the inverter or power board.
  
  The speakers still work.
  
  There is a rattling noise (sounds exactly like a baby rattle, except softer) coming from logic board when it is tapped or shaken.
  
  I didn't see any obviously blown caps/resistors/fuses.
  
  There is some glue on the board but it doesn't look bad/brown like in other threads.
  
  When I took voltage readings the 6 CFLs were not plugged into the inverter (because I could not get to the boards without unplugging them.)

Pin readouts: (Pin Pwr on/Pwr off)

CN853:

ST/P_DIM +2.0/0V
A_DIM +2.61/0V
BL +4.38/0V
SOS +5.23/0V
GND +0.00/0V
GND +0.00/0V
GND +0.00/0V
GND +0.00/0V
+24V +8.37/0V
+24V +8.37/0V
+24V +8.37/0V
+24V +8.37/0V

CN852:

DET +1.83/0V
ST/P_DIM +2.21/0V
A_DIM +2.61/0V
SOS +5.24/0V
BL +4.40/0V

CN851:

NC 0/0
NC 0/0
+5.3V +5.32/+0.45V
+5.3V +5.32/+0.45V
+5.3V +5.32/+0.45V
GND 0/0
GND 0/0
GND 0/0
NC 0/0
NC 0/0
NC 0/0
GND 0/0
GND 0/0
GND 0/0
S_B +5.22/+5.23V
PSON +3.18/0V

Big Cap CP801: +>1000(outside my multimeter range)/0V (I read this is supposed to be ~390V/165V?) Also this gave me a reading of ~170/170V earlier, but now is reading differently. Not sure why. This Cap is domed on top but it has a plastic top that looks like it was made that way.

So the CP801 is my first suspect but is there anything else I should be checking?

[euromatlox]

Re: Samsung 2493HM No Backlight

Those 24V voltage line values (8.37v) are pretty low. I don't know what they should be when inverter board is connected (=with load), but if you disconnect the inverter board (you can do that..not just CFLs) then it should read about 24v. If it still is much lower than it should be, check Sanken 8008hfe regulator. In any case it's good practice to replace all the electrolyte capacitors near it, and the other side of the transformer too, to exclude them from the 'failed parts list'.

You can also check burn mark on video board 240CT01C2LV0.2 (under metal plate), more details here:

https://www.badcaps.net/forum/showthread.php?t=26106

Are you really sure about the big capacitor voltage, should that kind of voltage destroy the cap totally ?

[Cateyerr]

Re: Samsung 2493HM No Backlight

I measured the 24V lines on the inverter side with everything plugged in and i'm still getting around 8.3V. You were right on the big cap, i think i measured the wrong solder points. I am getting 161V over it. It's not jumping up to ~390 like it should with the monitor on.

On the Sanken 8008hfe i'm getting

Pin 1 8.53V
Pin 2 5.34V
Pin 3 0V
Pin 4 0.78V
Pin 5 3.59V

[budm]

Re: Samsung 2493HM No Backlight

The 24V will not be working until the PFC Voltage booster circuits work first.
When work in the Primary side (hot side) you will use the negative leg of the large cap as the primary circuits Ground ref. for all the voltage measurement in the Primary side.
The chassis ground is for the secondary (cold side) circuits ground ref point.
I would remove all those yellow glues first (the glue become conductive and turn brown in color, it is also corrosive), you can soften them up with alcohol first before removing them.
Can you please take the straight shot pictures of the bottom side of the power supply board, also closed up in 1/4 section so we can easily see the SMD components and traces?
Check the DC voltage between jumper J808 (next the 8-PIN IC: ICB801S?) and the Negative leg of the large cap, it should have about 15VDC.
You should also check your meter's battery to make sure it is good.

[budm]

Re: Samsung 2493HM No Backlight

More info:
Pin8 of the SMD PFC IC (ICP801S) gets the power through linear regulator transistor (TO-220, QB801) sitting next to the 5.2V SMPS IC. You should see about 25VDC between the metal tab (Collector of QB801) and the Negative leg of the main large DC filter cap. The voltage between the Base pin of the transistor and the negative leg of the main filter cap should have about 16VDC.
By the way, check resistance reading between the VCC PIN 8 and the Ground pin6 of that PFC IC.
You may want to follow this thread also:
https://www.badcaps.net/forum/showth...715#post323715

[smason]

Re: Samsung 2493HM No Backlight

A side note, that brownish glue on many of the components looks suspiciously like the stuff that gets conductive as it ages. Might not be causing problems, but something to consider.

[euromatlox]

Re: Samsung 2493HM No Backlight

I measured a failed Sanken 8008hfe using diode tester function.

red cable into pin 3
black cable into pins 1,2,4,5

1-3 017
2-3 014
4-3 681
5-3 no reading

Using ohmmeter for the broken regulator:

1-3 14.4ohm
2-3 11.6ohm
4-3 no reading
5-3 8.5ohm

Diode tester for the brand new 8008hfe:

1-3 551
2-3 no reading
4-3 676
5-3 731

If measured still connected on board can be lower readings (since diodes, resistors, etc on board too).
Anyway, for broken one, diode tester beeps and/or readings are low.

[McEugene]

Re: Samsung 2493HM No Backlight

solved case that worked for me after all procedures above failed in my case:
my psu is a BN44-00195A model 245B2 (slightly different part numberingfrom schematics above)
symptom was:
a black screen with blue power light steady on.
5,2V standby power was ok, primary PFC voltage was ok (380V for my european 220V input monitor). but the backlight 24V was not there ...(5.3V was not their either since it is a byproduct of the main 24V output...).

what I did:
A) replaced most polarized cap though all my caps looked ok (no bulging top)---> still black screen
B) checked for the yellow glue becoming conductive, clean up a couple of components ---> still black screen (no 24V out)
C) checked the operation of the MC33067P ICM801
pin 14 was always high and pin 12 alway low...check the oscillator (pin 2 ) and their was no oscillations there, so I started looking at reasons for not startup of the IC in the datasheet...
pin15 12V was ok (power supply)
pin 8 non inverting output of error was 5.1V =(ref voltage of the IC) --> force max oscillator frequency in principle...

pin 10 fault input was low when I mesured it...so in principle all was ok , no fault detected ...
By reading the datasheet of the IC , I noticed that if there was a pulse greater than 1v on the fault pin the IC would lock itself to prevent serious damage and remain in that state until power is removed...
It was a long shot but I decided to connect my DSO (! use double 12V-220V AC 3A transformers back to back on the 220V ac input to allow the whole supply to float and prevent to blast my scope inputs!!!! Ov of the scope is connected to earth in principle...)
So, i measured (captured) the Pin 10 level as I flipped on the power switch of the PSU and I noticed a tiny 10ms 1.1V pulse on the fault line after 200ms then the line was perfectly below the 1V threshold of the fault input pin (pin10)...I was nearing the solution.
There are two separate circuits tha can trigger a fault condition on pin 10:
1)- the overvoltage monitoring circuit via opto PC802S on the schematics(PC804S on my PSU model)
2) the "feedback/monitoring circuit" from the primary of the TM801S transformer (DM805-CM809-RM815 on the scematics)

on the overvoltage monitoring circuit, I disconnected the 0 ohm resistor (it not on the schematics, but on the PCB is is easy to find by following the tracks)
between the PC802S opto led cathode and the LMBT2222 transistor to desactivate that circuit and measured again the power up sequence on pin 10 ---> the 10ms pulse high was still there....
Then,
I checked carefully feedback components from the the primary of TM801S transformer and noticed nothing abnormal at first, then I notice some of that yellow hardened glue on the 47pF 1KV capacitor (CM809 on the schematics) in the feedback circuit. That glue was covering both legs of the capa!!! It was a long shot but I decided to desolder the capa and clean it up as nice as possible without breaking it...soldered it back...and it WORKED! The MC3306P was oscillating nicely and the 24V was back, I reconnected the 0 ohm resistor of the other monitoring circuit and all was OK.
My 2493HM monitor is back to life and so far it as been working perfectly again...

So my advice is CHECK the Yellow glue on all critical components and remove it if you can!!!

Now I know that PSU quite well since I have been working on it for a complete day or so....Since I have two of those monitors I am ready for the second one when it fails (hope not but...)
McEugene is online now Report Post

[budm]

Re: Samsung 2493HM No Backlight

Yep, those glues must be removed as in post 4. I found this out since the 80's.