Destroyed LVDS cable for Dell 2405WFP

[ItFinallyWorks]

I have an interesting problem. I have a Dell 2405WFP monitor (acquired second hand, history unknown). It had the usual problem with bad caps, which I fixed. It also appears that it took out one of the converters on the logic board (loads the 19V supply (checked with no load) down to about 14V if I connect the logic board, fails to power up when connected, etc), but I need to troubleshoot that further. I'm not really that worried about that since it seems to be fairly common and I work on DC-DC converters for a living.

The interesting problem is like this. The LVDS cable on this monitor is part twisted pair wires and part flat flex cable. The flat flex looks fine, but the insulation of the twisted pair wires is crumbling and falling off. Literally, just unplugging it, much of the insulation crumbled off into my hand (and I was trying to be gentle). It doesn't fall off in large chunks, its tiny little pieces, like dust. Also, this harness has straight wires in it for power or something low speed. Those look perfectly fine. Its just the twisted pairs (which go through a manufactured, shielded cable) that have problems.

So, has anybody heard of this? I'm in the electronics industry and what one of my coworkers tells me is that this is typical of wires that have been overheated, but I've never seen insulation degrade like this - the overheated PVC wires I've seen have the insulation fall off in chunks, not dry dust like this. Also, other than the wires, there's no signs of overheating. So, I'm wondering if perhaps the panel itself has been fried or if this is just a manufacturing defect?

If it is just a manufacturing defect, anybody have some suggestions for repair or replacement? I can't find any on ebay and googling the part number on the label (50.L1E04.001) shows very few results. I doubt the the manufacturer would even talk to me about getting a replacement, assuming they had any. But, repair is difficult since these are tiny high speed differential pairs, so splicing is a bad idea as is covering with heat shrink since I will never be able to get the twist back right. I might be able to get the connectors and pins, but I'm sure I don't have the right crimpers for these.

[selldoor]

Re: Destroyed LVDS cable for Dell 2405WFP

This is the third case of this I have seen in the last week try search
dell lvds in the forum


EDIT - its here https://www.badcaps.net/forum/showpo...&postcount=240
and there is another case of it further down the same post.

[ItFinallyWorks]

Re: Destroyed LVDS cable for Dell 2405WFP

Thanks, looks like I'm not the only one is this boat. Not sure if I would have found that or not. Oddly a forum search for dell lvds doesn't bring up that thread - doesn't look like he used lvds anywhere. Now melt cable does, but I had to know that was the terms used in that post.

What's really got me interested is how this cable got hot since it isn't located near anything that should be particularly hot. The power supply is located well to one side of the logic board on this monitor. Unless maybe that backlight gets warmer than I think.

The other interesting thing here is that all the LVDS wires in this cable are destroyed, but wires literally right next to it (as in one slot in the connector over) are perfectly fine. Interesting. The LVDS wires are in a shielded cable where the shield is electrically and mechanically secured to the logic board. Maybe if that board gets quite hot, the connection conducts the heat to the cable? Or maybe those are just more sensitive.

Anyway, the solutions I've seen look pretty hacky. I do have some nice twisted pair ribbon cable I could hack something up with (and probably have it be vastly better than the other solutions I've seen), but I worry about signal integrity or robustness. Best would be replacing the cable or the wires, but this could be a design flaw so a replacement might not be any better.

[selldoor]

Re: Destroyed LVDS cable for Dell 2405WFP

Well I thought kazkasneaiskaus repair attempt was much better than hacky. In any case it was only a cheap repair to see if that was the only problem with his set.

[ItFinallyWorks]

Re: Destroyed LVDS cable for Dell 2405WFP

When I said hacky, I wasn't just referring to his fix. There was one other I saw that truly was. I will admit, the repair kazkasneaiskaus did does look nice and neat, but from an electrical engineering point of view it is a hack since it throws away the differential signaling and carefully designed line impedance you need to keep good signal integrity for those high speed data lines. That's why he had those issues with artifacts on the screen. But, like you said, if the point was only to demonstrate that the monitor works, it was quite successful.

My concern is, lets say I prove the monitor is still viable. Now what? A fix like that is great to demonstrate viability, but as he mentioned in his post, it has serious issues. To fix it right, differentials pairs are, at a bare minimum, an essential. But for it to be truly right and be assured that it works as designed, the wire really should be the gauge and twists per inch of the original to obtain the same line impedance as the system was designed for. Plus, you really shouldn't just solder new wires on since that does create stubs that can reflect edges and degrade signal integrity. So, to be right, the best way is to crimp pins on the new wires and replace the old. But those crimpers are specialized and expensive.

Remember, at the high speeds that cable operates at, this is RF territory. So, think transmission lines, proper termination, source impedance, and all the rest of that. It makes fixing it a whole lot more difficult since you really do need to do it right for it to work well. For alot of things, realistically, any old hack will work just fine - being neat and clean counts though. But once you start getting to these kinds of speeds, it gets more difficult. Not impossible, but you do have to pay attention to good high speed signal design practices. Neat counts here too, but there's more at play than just electrical connection.

If I had to guess, it's probably sufficient to carefully and neatly solder on something like http://www.computercableinc.com/ccin...+Cable&id=8376 (or similar - teflon insulation would be nice). But I can't guarantee that. Part of what makes me think its probably OK is this https://cdn.badcaps-static.com/pdfs/...5414ffa38e.pdf . My cable isn't just twisted pair - it has shielding too, but I'll bet it isn't essential (but nice).