IDE rounding question

C

clone#13

Limp Gawd
Joined
Aug 7, 2004
Messages
238
Hi, I am in the process of rounding my IDE cables for my HD and optical devices, and I was separating the idvidual wires with a razor, everything is going good except for one thing. I cut too close to one of the wires and I exposed the actual wire :eek: but I didn't cut throught it, but rather just shaved off a bit too much of the wire covering. I used electrical tape to seal the exposed wire......any one knows if that would or would not affect the functionality of the cable. Has it happend to any one? thanks for any help.
 
If you did not cut through the actual wire then it will be fine. As long as the wire does not touch any other metal then nothing will happen. You should be fine by just covering it up with electrical tape.

JEEVES
 
First off, rounded cables, homemade or not, do not conform to the ATA specification in any way, shape, or form. This is for a very good reason. By separating the individual wires from the ribbon, you are separating active data lines from the grounds that are supposed to accompany them and protect them from crosstalk. Crosstalk can wreak havoc on performance, or worse, data integrity. I certainly wouldn't trust by data with split wires from a ribbon cable, let alone bare wires that have been patched up.
 
lol, as much as that makes since it not 100% acurate. I am not seeing where these grounding lines are. I have split and taken apart many of ide cables and there is nothing that runs between the data lines. It is just rubber. If what you are saying is true then my lcd cable would not work, my VFD cable would not work. And my crappy IDE cables would not work. Like I said before, this will not effect the performance as much as he is making it out to be.

JEEVES
 
Ever wonder why those ATA-66/-100/-133 cables have 80 wires but only 40 pins on each end? That's because those extra 40 wires and then some are placed adjacent to the data lines to prevent crosstalk. These wires look identical to every other wire in the ribbon, but their purpose is for the sole reason of preventing crosstalk.

And why does this matter? Crosstalk corrupts the input signal, causing an unexpected output signal. Luckily, this is addressed by error correction signals that accompany the data flow, but when error correction needs to be used, it slows down the data transfer, reducing performance. Severe corruption, which is possible when wires are split apart from their grounding wires, can overcome the abilities of the error correction circuitry and allow corrupt signals to pass, causing errors and data corruption.

Will all this doom and gloom definitely happen to anyone who rounds their cables? No, but do you want to worry about it in the first place?
 
Well thanks for all the input guys, but I have to agree with Jeeves, each wire is isolated from the other in its own sheath.....as long as the wire is not exposed and touching another exposed wire. All the wires are in their own respective cover. I mean Asus and all other companies are selling premade rounded IDE cables, I am sure they are aware of crosstalk.

Clone#13
 
Still, Asus will never claim that your data is safe with a rounded cable. Besides, who knows of the internal decisions between marketing and engineering that led to the final product from Asus or any company? It doesn't mean the engineers approved of it--and they shouldn't.

Lost Circuits has a well written article explaining why the ATA cable is layed out as it is, and explains why it needs to stay that way. Here you go:

http://www.lostcircuits.com/advice/sata150/4.shtml

It may seem like a foreign concept to you since you're probably not an electrical engineer, but two completely insulated wires can affect each other while running parallel to one another. That's why ground wires (that are also insulated) must be placed between data wires to avoid interference with one another. This is a big problem when dealing with compatively high speed data buses like the IDE bus. With a parallel bus it's especially challenging.

Well, whatever. It's your data, not mine.
 
xonik is right on about this although some don't get crosstalk while others do. Round cables can lead to crosstalk (I got data curroption like that) but not all the time (it happened only once with a cable that I made), your chances decrease with the quality of the cable.

I've rounded my own cables before with only one out eleven going bad from it but now I just buy the round cable (Akasa brand haven't done me wrong yet) if I need them (I've actually gone back to some regular IDE cables since they are so flexable).
 
I am glad to get all these feedbacks...thanks..I actually had and extra IDE cable that I was not using and decided to give it a try and see how it will look. I guess installing them would tell me if it works, but data corruption might no occur right away, am I right?. I guess the ultimate solution to those wide cables would be to get SATA drives since the cables are thinner and can be hidden around the case.
 
It took something like a month before I saw data corruption (or noticed it), the "ulitmate solution" would be to just get some high quality ones I guess (with some SATA drives :D ). I think I saw some round Akasa cables around 18"-24" for something like $12.99 at CompUSA, not bad IMO. Antec also has some good ones (I believe they are shielded too, I think I've seen them there too).
http://www.performance-pcs.com/cata...d=346&osCsid=a674812aaaf50c813090118dbea1ed93
 
I think they probly use twisted pairs (data/ground or w.e)

what do you guys think about cables longer than the 18" I think it is specification for ata133 etc

I think 24" should be ok, as I might get a couple vantec 2 device black rounded ones.
 
clone#13 said:
Hi, I am in the process of rounding my IDE cables for my HD and optical devices,
Click to expand...

I know it's an old thread. But this still shows up in google searches. And they are still selling these 'rounded' cables at various commercial sites.

The guy above was right. The 80 wire IDE/'PATA' cables contain 40 dummy wires. The purpose of these is to be in physical proximity to the live/data wires, in order to help maintain integrity of the data. Cutting the wires, separating them from each other, utterly defeats the design of the 80 wire ribbons.

The best thing is not to round them at all, but to position the flat ribbon so that the airflow in the case is right OVER them, like an airfoil; not into them like a brick wall. The flat ribbon provides even less obstruction than a thin SATA cable.
 
sevry said:
I know it's an old thread. But this still shows up in google searches. And they are still selling these 'rounded' cables at various commercial sites.

The guy above was right. The 80 wire IDE/'PATA' cables contain 40 dummy wires. The purpose of these is to be in physical proximity to the live/data wires, in order to help maintain integrity of the data. Cutting the wires, separating them from each other, utterly defeats the design of the 80 wire ribbons.

The best thing is not to round them at all, but to position the flat ribbon so that the airflow in the case is right OVER them, like an airfoil; not into them like a brick wall. The flat ribbon provides even less obstruction than a thin SATA cable.
Click to expand...
I have yet to hear of any HD corruption or performance issues caused by rounded IDE cables. They're well-accepted in the industry by now. You're right about the 40 GND wires in the 80-pin cable, but it's no big deal to separate chunks of wire. It is helpful, however, to separate into chunks of even numbers of conductors, so for an 80-wire cable, separate into 10 chunks of 8 or so, so that each conductor has its neighboring ground conductors to keep it company.
 
Mohonri said:
I have yet to hear of any HD corruption or performance issues caused by rounded IDE cables. They're well-accepted in the industry by now. You're right about the 40 GND wires in the 80-pin cable, but it's no big deal to separate chunks of wire. It is helpful, however, to separate into chunks of even numbers of conductors, so for an 80-wire cable, separate into 10 chunks of 8 or so, so that each conductor has its neighboring ground conductors to keep it company.
Click to expand...

If you look at the photos, however, they're not keeping anything together. At the connection, it's just all separate wires. That it's 'well-accepted' could just be from ignorance and people attributing problems to some cause other than a bad cable?

But the guy, above, was absolutely right. The whole point of the 80 wire ribbon was to keep the 'dummy' wires in proximity to the working 40 wires. Separate them out, it literally defeats the entire design.

They should just see the 'rounded' cables as 40 wire cables and say that somehow the 40 wires are insulated or otherwise protected so that the data errors at 100 and 133 are just not a problem as they would be with a 40 wire ribbon. But again, that will pose more of an obstacle for airflow that would a flat ribbon.
 
Commercial rounded cables = Good
Homemade for ATA33 devices (40-wire) = Usually Good
Homemade for ATA66+ devices (80 wire) = Dodgy.
 
cpemma said:
Commercial rounded cables = Good
.
Click to expand...

Perhaps because they are insulated in some way to preserve the data. I don't know. But it would seem only that it would suffer even in comparison with the old 40 wire cable.
 
Mohonri said:
I have yet to hear of any HD corruption or performance issues caused by rounded IDE cables. They're well-accepted in the industry by now. You're right about the 40 GND wires in the 80-pin cable, but it's no big deal to separate chunks of wire. It is helpful, however, to separate into chunks of even numbers of conductors, so for an 80-wire cable, separate into 10 chunks of 8 or so, so that each conductor has its neighboring ground conductors to keep it company.
Click to expand...


OK.

Now you've heard of both corruption and performance loss :

In the past, I have used rounded ide cables on my rigs.

I built my p4 northwood with rounded cables.From day one , every drive on the rounded cables automatically defaulted to ATA 33........I could not make them work to spec no matter what I tried.Further to this , every drive on the rounded cables experienced data corruption sooner or later during the 2 year span till I went back to ribbon cable.

The SECOND I went to proper ATA133 cable , the bios reset to the correct speed on every device I had.I have also not had a single data error since (3 years).

I won't use rounded cable at all anymore unless I have space issues.

The Issue of crosstalk is not an insignificant one.IMHO rounded cables are no good.
 
Well, there ya go. Thanks for the story. I'm personally not a huge fan of rounded cables, as I find flat cables easier to route and hide.
 
[url=http://www.provantage.com/qvs-ideu-2bbg~7QVSS004.htm]ProVantage shop[/url] said:
Connects two IDE CD/DVD or hard drives and supports speed up to 133MBps. This premium improved-airflow round internal cable has 40 twisted pairs and PVC jacket.
Click to expand...
That's why the ground wires still work in a properly constructed round cable. I vaguely remember some of the early rounded cables were more gimmick than technically-sound, but I've never had data transfer problems with mine.
 
I've been using commercially made round wires for over 3 years in my box without any data loss/corruption. The box works perfectly. They are 133MBps and were made for Thermaltake. Put some tape over the exposed wire and try it in your set-up alone and using standard flat wires on the other peripherals. If you get problems, then you will know it isn't working.
 
You must log in or register to reply here.
Back
Top