21 inch or 18 inch or what?

G

George Hester

I know that the specs for an IDE cable calls for an IDE cable to be no
longer than 18 inches with I think it is 4 inches seperating the two
connectors that go to devices. I am finding that these specs are making it
very diffiicult for me to put IDE drives in the box hooked to a PCI I/O
card.

I have a cable that is 21 inches and have seen other IDE cables that do not
abide by the specs. Is there any reason other than "what's within specs"
why I would not want to use out of spec IDE cables if they work? Thanks.
 
P

paulmd

George said:
I know that the specs for an IDE cable calls for an IDE cable to be no
longer than 18 inches with I think it is 4 inches seperating the two
connectors that go to devices. I am finding that these specs are making it
very diffiicult for me to put IDE drives in the box hooked to a PCI I/O
card.

I have a cable that is 21 inches and have seen other IDE cables that do not
abide by the specs. Is there any reason other than "what's within specs"
why I would not want to use out of spec IDE cables if they work? Thanks.

Eventually, you get timing difficulties and signal loss, which could
result in performance degradation or data corruption. 3 inches
shouldn't be a big problem, though.

If they work fine, I wouldn't worry.
 
R

Rod Speed

George Hester said:
I know that the specs for an IDE cable calls for
an IDE cable to be no longer than 18 inches
Correct.

with I think it is 4 inches seperating the
two connectors that go to devices.

Nope, its a max of 6" with the 80 wire cable
and a min of 5" just with the 40 wire cable.
I am finding that these specs are making it very diffiicult for
me to put IDE drives in the box hooked to a PCI I/O card.

Which one is the problem, the total length ?
I have a cable that is 21 inches and have seen other
IDE cables that do not abide by the specs. Is there
any reason other than "what's within specs" why I would
not want to use out of spec IDE cables if they work?

They dont necessarily work as well if they appear to work.

Corse 21" isnt obscenely over length.
 
G

George Hester

Rod Speed said:
Nope, its a max of 6" with the 80 wire cable
and a min of 5" just with the 40 wire cable.


Which one is the problem, the total length ?


They dont necessarily work as well if they appear to work.

Corse 21" isnt obscenely over length.

Yes Rod the total length. The PCI I/O card has to be just below the AGP
cionnector on the board. That is OK but 18 inches just does not reach the
device trays everything in this box is tighter than Candy Barr's brasierre.
I think I can build a 2-device container closer to the PCI card but if I
don't need to I'd rather not. Will yes total length.

I have an IDE cable that is 7 inches in length 3 inches seperating the
device connectors. Which seems to obviate the reasons why a long cable
would not be good. What are the reasons why a short length canble not be
good?


George Hester
_________________________________
 
S

Starz_Kid

George Hester said:
I know that the specs for an IDE cable calls for an IDE cable to be no
longer than 18 inches with I think it is 4 inches seperating the two
connectors that go to devices. I am finding that these specs are making it
very diffiicult for me to put IDE drives in the box hooked to a PCI I/O
card.

I have a cable that is 21 inches and have seen other IDE cables that do not
abide by the specs. Is there any reason other than "what's within specs"
why I would not want to use out of spec IDE cables if they work? Thanks.

Hello, I've used 21" ide cables and longer ones in my main computer (very
tall case) for many years with NO problems at all....................!

Starz_Kid...
 
R

Rod Speed

George Hester said:
Yes Rod the total length. The PCI I/O card has to be just below the
AGP cionnector on the board. That is OK but 18 inches just does not
reach the device trays everything in this box is tighter than Candy
Barr's brasierre. I think I can build a 2-device container closer to
the PCI card but if I don't need to I'd rather not. Will yes total
length.

I have an IDE cable that is 7 inches in length 3 inches seperating the
device connectors. Which seems to obviate the reasons why a long
cable would not be good. What are the reasons why a short length
canble not be good?

It was always pretty theoretical and there is no minimum now with 80 wire cables.

You see much shorter ones in external housings particularly.
 
R

Rod Speed

Hello, I've used 21" ide cables and longer ones in my main computer
(very tall case) for many years with NO problems at all....................!

The many years is completely irrelevant. The problem is with ATA133 speeds.
 
G

George Hester

Rod Speed said:
It was always pretty theoretical and there is no minimum now with 80 wire cables.

You see much shorter ones in external housings particularly.

Well if it is just theoretical I guess my dummy machine won't know the
difference. In any case I built a cage close enough so that everything is
in specs. I'll sleep better I am sure. Thanks for everyone's help it is
good to know for my next machine.
 
T

Todd

I would run the drive test utility from the drive manufacturer with a cable
shorter than the spec, and then again with the 21" cable. That will let you
know if it is making a difference.

If you can't detect a difference, and everything seems to be working, keep
an eye on it, but go with it.

Todd
 
P

Poly-poly man

George said:
I know that the specs for an IDE cable calls for an IDE cable to be no
longer than 18 inches with I think it is 4 inches seperating the two
connectors that go to devices. I am finding that these specs are making it
very diffiicult for me to put IDE drives in the box hooked to a PCI I/O
card.

I have a cable that is 21 inches and have seen other IDE cables that do not
abide by the specs. Is there any reason other than "what's within specs"
why I would not want to use out of spec IDE cables if they work? Thanks.

The "Official" IDE cable specs limit total length to 18". However, tests
have shown that 80-connector cables will work reliably up to about 27".

http://www.t13.org/technical/e00151r0.pdf
has a proposal to change the specification showing 'negligible
differences in Ultra DMA Mode 5 signal integrity between a 27",
80-conductor cable and an 18", 80-conductor cable.'

Just to note: the extra 40 cables in the 80-ridge design are grounds so
that stray bad signals are "eaten up".

hih,
poly-p man

P.S.

Pages 562 & 563 in your Upgrading and Repairing Pc's 17th edition. :)
 
K

kony

The many years is completely irrelevant. The problem is with ATA133 speeds.


Agreed, but some what to argue about it without ever trying
it. I recall a past thread where a certain troll (not you)
continually argued against dropping ATA133 speeds down to
ATA100 or 66 when using over length cables... though IIRC
the length was more than 21".
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top