Question Please do educate me with some of your cable stories

[gn842a]

Is DVI-D on its way out? Is it for losers? Was it always?

So five years ago I took it into my head to have a work station in one room and a home theater in the other, both running off the same computer. I am pleased to be at a point in life where, when I ask the landlord for permission to do something, he always says yes, because he is me.

So I drilled through the floor where the PC/TV screen is and ran a humongo 25 foot DVI-D across the basement ceiling all the way into the other room where I drilled another hole in the floor and brought it up to the monitor. Mouse and keyboard were to be hooked up through the monitor. Now this was no ordinary DVI-D cable, I special ordered it, it had some kind of booster.

Well this rig has always worked in terms of showing the display in both rooms, but there were from the beginning issues with both mouse and even keyboard lag. One of my earliest posts here was on this topic. Eventually I got a separate 25 foot USB line for the mouse and connected it directly to the PC. The keyboard lag I just lived with.

So the years go by and I have a number of issues with my recent rebuild. Along the way I have had many discussions here and with the people I know, who have some knowledge of what I'm trying to do. The local non-internet talent suggested that my problems with the RX 590 Nitro might be cable related (repeated crashing during benchmarks). Well, the RX 590 had already left the house on an RMA but I said OK, I'll try a new cable, and when the RMA comes back it will have a new cable to talk through.

The cable is here first. This time I got a 25 foot displayport 1.2 (looked for 1.4 in this length, couldn't find it). Anyhow I hooked it up and now no more mouse lag and no more keyboard lag. Who knows, perhaps also the graphics card issues are/were related to the same issue, though the old graphics card never faltered with the old set up (R9-380 Strix). That experiment remains to be run in the next week or so.

But never mind that, I very much wish that I had moved to displayport a long time ago. I'm kind of banging my head on the desk about it. If I read the specs right, the difference between gb/s between DVI-D and displayport is only about 10%, I never would have expected it to make a difference in my (unorthodox) build. Regardless of what happens with the eventual new graphics card, this displayport is noticeably better for the build.

I'd kind of like to know if anyone else has any "the problem was in the cables" stories so I can know what to be on the alert for. Should we as users be on the alert for the development/release of new cables? What makes cables go bad? When do we turn to cables as the issue in diagnosing a persistent problem? In my case cables were pretty much the last thing on the list. I didn't even think of it till I was reminded by others. When you look at the cables inside the case and out, that's a pretty significant diagnostic hurdle, in terms of sheer numbers.....

thanks,
Greg N

 

[Joseph_Tech]

Just like how VGA was fased out these can to with the development of newer faster cables so yes you should probably be on alert for new cables. I don't really know what makes cables go bad. But I've made it a habit of checking the cables first, before hardware only because I'm hopeful it's just that I forgot to plug it in or plugged it into the wrong place. You should too it will save you time.

 

[Eximo]

I would have to do a deep dive on the specs of DVI and the specific cable you had to place a guess on the problem. Dual Link DVI was the go to standard for high refresh displays up until the release of G-Sync and FreeSync effectively. But in reality that was two video signals.

I'm not certain DVI has error correction built into it. Pretty sure HDMI and DP do. And from brief reading HDMI and DVI were designed to be inter operable by using the same signals for driving screens.

HDMI and Display Port are basically fancy serial bus transmission cables. Max bandwidth is limited by length. You'll have more voltage drop across the wire's resistance the longer it is. A quality long cable should be using thicker wires.

Temperature cycles can certainly effect a cable, it will stretch and shrink. If it is tightly restricted, this will make it worse. That applies to any cable.

Crosstalk and shielding. It could have been simple interference from electrical sources that ran over or parallel to the cable. Could also have been grounding loop related, where you actually were pumping voltage down the wire simply due to a difference in the ground states between the two locations. (I have seen that before)

My understanding of HDMI and DP is limited when it comes to solving ground loop problems, but I understand it is the preferred audio solution these days. So they might have circuitry at either end to counteract such issues. Not sure if that applies to DP though.

 

[Eximo]

As for stories:

This was in a welding shop in the early 2000s. A new plasma cutting table was being installed. Essentially a light duty CNC that held the plasma cutter so you could program in flat shapes to be cut out.

A 100ft plus parallel cable was run from a PC in the shop's office to the table. The cable was run along the ceiling. Now this being a welding shop there was a lot of 3-phase power distribution rails up there. So it was run along one of them.

We had issues getting the PC to talk to the table. One of the techs let out a yelp while we were working on it and said "It shocked me!". Now we were all thinking static shock, but he said it felt like it is live. And we start poking around at it and eventually discover something like 50 volts floating between the shield on the parallel cable and the chassis of the table. We got a short Parallel cable, dragged the PC down to the table and tried it, worked fine. Tried another long cable (smartly a spare was purchased, welding shop...) As long as they were on the same ground, no problems. But the high voltage rails and the normal outlets in the office were definitely on different circuits.

We didn't stick around for the final solution, but the manufacturer suggested to sink a new ground rod to solve the issue. Apparently the way the building was built the ground for the ship was actually gotten from effectively another shop in the same building, something like 150ft away. And the ground for the office was from some other part of the building.

And a boring one of troubleshooting for over an hour only to find a shorted out SATA power cable. Weird thing on that one was that the PC would boot, but fail to get into Windows. I'm guessing ACPI didn't take kindly to the short. Luckily it was a modular cable.