Weird computer monitor problem

[HPsauce]

One for the IT/electronics experts this...…

The monitor on SWMBO's computer started behaving oddly at start-up, sort of "interference" as on an old-school analogue TV screen.
BUT:
1. It fades over time, gone completely in a couple of minutes apart from a very light "speckle" on some plain coloured backgrounds (need to peer closely) that takes a bit longer to disappear.
2. It's exactly the same on analogue (VGA) and digital (DVI) inputs. Normally runs at 1920*1080 on DVI.
3. It's biased to the left side of the screen and is mostly noise with a tiny bit of in initial horizontal jitter.
4. I eventually realised it's nothing to do with the input signal or external interference. Just power on the monitor (see attached picture) standalone and you see it.
6. There is an external mains "brick" with a 12v DC supply to the monitor. I've tried multiple different power supplies and can rule that out.

If you power it off briefly (once "warmed up") the interference doesn't reappear. It needs to be off for a while. This leads me to suspect either a capacitor issue or something heat-related.

Looking at the back I can see no obvious way in unfortunately, though if it gets worse I'm happy to get quite brutal!



Any suggestions from people more knowledgeable than me about LCD screens what it might be and how likely fixable?

[ainarssems]

Pretty sure it's the screen itself and there is no fix, at least not an economical one. You could replace screen if you can find one for it but it's likely to be more expensive than another monitor.

[moltuae]

I used to repair TVs and monitors .... a loooong time ago. But they're pretty disposable these days. Even IF you can get hold of the service manuals, schematics and components, it'll almost always cost more in time than it's worth just to strip the thing down and rebuild it.

My guess would be some sort of video controller SMD IC has a 'dry' joint or two. You might be able to fix it if you can identify the offending IC and resolder it. If it's of the 'ball grid array' variety ('legless'), your best chance would be to 'reflow' the solder paste by carefully heating the IC, which requires a hot air gun (or sometimes putting the entire board in an oven *might* work. It might also destroy it of course if you get the timing/temperatures wrong).

Given than the fault is thermally related, probably the best way to identify the culprit would be with a can of 'freeze spray'.

My second guess would be a video controller ribbon cable connection. You could try waggling some internal cables or applying a little pressure to see if you can make the fault come and go.

If you haven't done so already, you could try Googling the model number to see if it's a common fault and whether anyone has posted an easy fix method.

[ainarssems]

In my expierience if it's random spots it's the fault in screen it self. If it's something external you get whole rows or columns messed up. Sometimes spots go away if you apply a pressure on the screen near it or if you apply bending pressure to it.

[moltuae]

Originally Posted by ainarssems

In my expierience if it's random spots it's the fault in screen it self. If it's something external you get whole rows or columns messed up. Sometimes spots go away if you apply a pressure on the screen near it or if you apply bending pressure to it.

For a failure of an odd pixel or three, I'd agree, but if it's dozens of pixels that have suddenly failed and it's a fault that comes and goes, I'd say it's unlikely to be the screen itself. It looks from the photograph to be more than just individual pixels too, more like short lines of 3 or 4 pixels. That's almost certainly a video controller fault. It's possible of course that the controller circuit responsible is built into the back of the screen, and therefore replacing the entire screen 'module' would fix it, but either way I reckon that's a fault at the video 'processing' level.

[HPsauce]

The pattern moves VERY dynamically, it's not tied to specific pixels.

I tend to favour the controller/cable as the problem area rather than the LCD panel itself.

I might try cooling it (freezer might be terminal!) to see what that does to the symptoms.

[vagdream]

Have you tried connecting a different screen to the port to confirm the problem is actually in the screen hardware?

[HPsauce]

It's not relevant, though I did that in the early stages of testing (and tried the monitor on other PC's), see first post:

4. I eventually realised it's nothing to do with the input signal or external interference. Just power on the monitor (see attached picture) standalone and you see it.

That picture was taken with the monitor disconnected and nothing nearby powered on.

[HPsauce]

Originally Posted by moltuae

Given than the fault is thermally related, probably the best way to identify the culprit would be with a can of 'freeze spray'.

Well, I actually did the opposite, with interesting results!
There's a bulge on the back which includes all the external connections and has a cooling vent on its top edge, so I guessed that encloses most of the electronics.

I pointed SWMBOs hair dryer (I have no need of such frippery!) at it for a couple of minutes to warm it. Turned it on and it's working fine.

What does that imply now?

[moltuae]

Originally Posted by HPsauce

What does that imply now?

Probably a slight crack in a solder joint somewhere. Expansion will be re-establishing a connection as it warms. To pinpoint the exact component you'll need to apply heat/cold more precisely, one component at a time, or you could try applying a little pressure. Use something like a rubber-ended pencil or something long and plastic -- fingers and metallic objects can affect such sensitive signals, giving inconsistent results.