Which came first - hole or the melt - ACER S220HQL LCD [FIXED]

[Inflex]

A tale of frustration and expense...

A while back a client requested a field service, complained that their display was no longer working.  After checking the system with another screen there was no output, trying my field test screen on DVI also showed no output so, the diagnosis goes "Dead GFX card".  Replace the card, comes up just dandy fine with the test screen.   Plug the client monitor back in and it's dead again *scratches head*.  Okay, maybe the replacement GFX card doesn't agree with the LCD for what ever reason, plug the test screen back in, it's definitely dead.   "Maybe a bad GFX card out of the box perhaps?" (seriously, thinking back, that was just a terrible mistake).  I request to take the whole system back to the workshop.

Back at the workshop, like a FOOL I plug the screen in to one of my workstations - nothing comes up "Huh, must be a dead screen too"... plug my original screen back in to the workstation, nothing comes up :-O    It now makes sense to me, the screen is killing the GFX cards when on the DVI port.  I say my curse words, get a new screen, new GFX card for the client machine, they're all good again, and I put the LCD in to the "naughty corner" for me to one day take a look at for curiosity sake.

....  today ....

Finally I open it up.  On first glances there's two things that strike me; 

• there's a faint splattering of coke-coloured drops all over both top and  bottom of the black portion of the DVI socket ( you can see the effect on the underside of the DVI connector too in the picture )
  
• there's a small blob of plastic at  the base with very faint charring around it

Firing up the hot-air I manage to extract the DVI socket off to reveal a pitted PCB at where I saw the blob,  but more interestingly, all the keying/guides have been melted down on the DVI connector, and the PCB key recesses don't line up with those of the DVI socket.

So, now I'm wondering;  was it just bad luck / coincidence that this PCB failed right under where one of the molten keys were, or did the process of driving the DVI socket down start the process of ultimately creating this GFX killer fault?

The VGA side still works a treat, so I'm almost considering using it in the workshop as a status display.

Only sorry I didn't video the whole process, would have made another interesting YouTube video on my channel

[wraper]

Firing up the hot-air I manage to extract the DVI socket off to reveal a pitted PCB at where I saw the blob,  but more interestingly, all the keying/guides have been melted down on the DVI connector, and the PCB key recesses don't line up with those of the DVI socket.

So, now I'm wondering;  was it just bad luck / coincidence that this PCB failed right under where one of the molten keys were, or did the process of driving the DVI socket down start the process of ultimately creating this GFX killer fault?

For me it seems that guides perfectly line up with a PCB and they melted because of your hot air as you were blowing on them directly through the holes. IMO this connector has nothing to do with the fault and you should check what is connected to that burned via/trace.
Also sometimes guides may a little bit melt during wave soldering process.

[Inflex]

Sorry, I should have circled the one I were referring to; picture attached.

Yes, the rear ones indeed were blown out by the hotair as you suggested, likewise as this was a wave-soldered board (interestingly they do have the "Wave flow ->" arrow on it too).

When initially inspected, there  was a ball-shaped black crusting extended beyond where the DVI socket foot was resting on the PCB, and you can see the cavity taken out from the foot, different to the one on the right in the original picture).

Anyhow, not really expecting to be able to safely repair this,  so  either I'll find a replacement controller  board, or just be content with the VGA (at least now with the DVI socket removed there's no chance of a false connect.

[wraper]

That is not a guide, just standoff. It was melted by burning via. Obviously there was high current flowing through that trace, and via burned because had the highest resistance in the current path. It is possible that there was a short between those two vias either.

[Inflex]

That is not a guide, just standoff. It was melted by burning via. Obviously there was high current flowing through that trace, and via burned because had the highest resistance in the tureen path. It is possible that there was a short between those two vias either.

Seems a bit obvious now in hindsight.  Nothing like posting everything up to come back an hour or two later and realise, tx.

[Fraser]

A bit of research on the burnt  via's purpose and effect on the DVI link may also enable a repair of the video cards if you are so inclined?

Maybe not commercially viable to do such but I am the curious sort who likes to know the cause of a failure in case I see similar again in the future. Some faults like capacitor failures turn out to be very common on some models of equipment.

Fraser

[amyk]

Check primary-secondary isolation of PSU, then check voltages at DVI connector. It seems something may be driving current back through the connector into whatever is connected.

[Inflex]

Here's as close as I can get without having to start using a microscope, which I do not have (yet).

From what I'm tracing, the burned out via goes to left pad of R77 (Between D7 & Q5) - comes from Pin 14 (+5V) of the DVI [ Power for monitor when in standby ]
The other via (intact) goes to left pad of R78 - comes from Pin 15 (GND) of the DVI  [ Return for pin 14 and analog sync ]

Check primary-secondary isolation of PSU, then check voltages at DVI connector. It seems something may be driving current back through the connector into whatever is connected.

This particular display uses an external SMPS pack.   I'll probe the voltages once I've rigged it back together.

[Monkeh]

+5V for the EEPROM and its ground pin - check the cards, money's on a small fuse protecting the 5V output. No +5V, no EEPROM, 'dead card'.

Seems unlikely the card is sourcing anything (diode ORed), possibly a fault in the monitor drawing excess current, more likely just somehow shorted under the connector, considering the melting.

[Inflex]

+5V for the EEPROM and its ground pin - check the cards, money's on a small fuse protecting the 5V output. No +5V, no EEPROM, 'dead card'.

Seems unlikely the card is sourcing anything (diode ORed), possibly a fault in the monitor drawing excess current, more likely just somehow shorted under the connector, considering the melting.

I'll have to now go back through the old parts and find that card, thanks.

The workstation unfortunately was a Gigabyte J1900N-DV3 which is a mini-ITX board with integrated CPU/GPU.  The 5V line is intact so far as I can see (put a probe in Pin14, pick up continuity at several points across the board, likewise Pin15 is also intact to ground.  Might see if I can source a schematic of something similar.    Would appear it's not sensing anything plugged in to DVI because as soon as I plug HD15/VGA in it immediately lights up the display.

[Monkeh]

Check it's not shorted on the monitor end, and that the diode is functioning properly.

[Inflex]

I powered the board up this morning on the LCD and much to my amusement I'm seeing 0.9~1.8V on pin 14 of the DVI, when you'd think that there should be 0V as it's a blown via and the trace is visibly disconnected.   Resistance though measures 6MR, so it would seem that the original destruction has made the area slightly conductive. 

I'll clean up the area, probably run wires direct from the resistors to the pins on the other side and cut the traces.

[Inflex]

With nothing else to go on, I cleaned it all up, routed a new trace, reinstalled the DVI socket, found a sacrifice mainboard... and crossed my fingers.




Now the hard thing is getting the damned push-button array at the bottom left to sit in the right place when I clip down the front bezel, seems to be more difficult to get it right than fixing the screen!