detrimental effects of "round" cables?

Has anyone had any experience with ATA/133 drives
going to "round" cables? I've got a couple new Maxtor
7200 rpm DiamondMax drives with 8MB cache that
need a home, and since my mid-tower case will get a
CD burner and a couple new PCI adapter cards, I
figured the improved airflow afforded by "round" cables
would help control the heat. But... at 133MHz, will
the bunched physical arrangement of the wires in the
"round" cable introduce signal anomalies, resulting in
errors?

Since so few people could actually measure the
electrical difference between ribbon and "round" cables,
has anyone who had an ATA/133 HD and then had gone
to "round" cables noticed any decrease in file transfer
speed?


*TimDaniels*

Previously Timothy Daniels wrote:
Has anyone had any experience with ATA/133 drives
going to "round" cables? I've got a couple new Maxtor
7200 rpm DiamondMax drives with 8MB cache that
need a home, and since my mid-tower case will get a
CD burner and a couple new PCI adapter cards, I
figured the improved airflow afforded by "round" cables
would help control the heat. But... at 133MHz, will
the bunched physical arrangement of the wires in the
"round" cable introduce signal anomalies, resulting in
errors?

Since so few people could actually measure the
electrical difference between ribbon and "round" cables,
has anyone who had an ATA/133 HD and then had gone
to "round" cables noticed any decrease in file transfer
speed?

Round cables are more likely to introduce errors. If they are
used with newer HDDs there is usually no problem, because
these do error detection and correction on the IDE bus.
However I some problems:

1) A current CD0burner (TEAC CD-W 540E) did burn disks
with bit-errors when used with a round cable (no
read error though).
2) A 90cm rounded cable produced so many errors that
my Linux kernel decided that the disk was defective
and took it down. It could also have been the disk
that decided it was not use working on this bus, I
do not quite remember.
Cutting the cable down to 60cm and using it with
a single drive solved the problem.

So no, the bus will become unusable lomg before there
is any significant speed decrease.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus

Some pathetic excuse for a bull**** artist/pathological liar claiming to be
Joshua P. Hill wrote in
message ...
Rod Speed wrote
Timothy Daniels wrote

Has anyone had any experience with ATA/133 drives
going to "round" cables? I've got a couple new Maxtor
7200 rpm DiamondMax drives with 8MB cache that
need a home, and since my mid-tower case will get a
CD burner and a couple new PCI adapter cards, I
figured the improved airflow afforded by "round"
cables would help control the heat.

Thats basically a myth.

Not if the cable blocks the airflow, dorkbrain

It doesnt, ****wit.

-- and it often does.

In your pathetic little drug crazed fantasyworld, ****wit.

"Arno Wagner" wrote:
Round cables are more likely to introduce errors. If they are
used with newer HDDs there is usually no problem, because
these do error detection and correction on the IDE bus.
However I [had] some problems:

1) A current CD0burner (TEAC CD-W 540E) did burn
disks with bit-errors when used with a round cable (no
read error though).
2) A 90cm rounded cable produced so many errors that
my Linux kernel decided that the disk was defective
and took it down. It could also have been the disk
that decided it was not use working on this bus, I
do not quite remember.
Cutting the cable down to 60cm and using it with
a single drive solved the problem.

So no, the bus will become unusable lomg before there
is any significant speed decrease.


The longest cable that I purchased was 24" (about 60cm)
in total length (8" from the slave connector to the master
connector, 16" from the master to the board connector).
What was the length of the cable that you used with the
TEAC CD burner? Did that cable have an outer metal
braid for shielding?


*TimDaniels*

Previously Timothy Daniels wrote:

"Arno Wagner" wrote:
[...]

The longest cable that I purchased was 24" (about 60cm)
in total length (8" from the slave connector to the master
connector, 16" from the master to the board connector).
What was the length of the cable that you used with the
TEAC CD burner? Did that cable have an outer metal
braid for shielding?

I had the problem with the burner with a short cable (45cm
or 48cm) as far as I remember. I think it was a CoolerMaster
quality cable with shielding. The problem is that while newer
UDMA standards seem to require CRC-checking in all modes,
this burner seems not to do it when writing. I have no idea
whether it sends out CRC codes when reading, but I did not
observe a single read error from correctly burned CDs
and the bit-errors on the incorrect CDs did not change.

The shop where I bought the burner told me that they knew
about the problem and that some writers should not be used
with rounded cables at all, while others can be without
problem. It could be that a firmware-upgrade solves
this problem, but I am too lazy to try. My posting
was just to warn people to be careful, since a few (1..10)
bit-errors in a full CD might not be immediately noticable
but still cause problems.

I have some cheap 60cm (actually cut down 90cm) cables in
use with HDDs for some time now without problems. These are
twisted pair without shielding or brand name.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus

Some pathetic excuse for a bull**** artist/pathological liar claiming to be
Joshua P. Hill wrote in message
...
Rod Speed wrote
Joshua P. Hill wrote
Rod Speed wrote
Timothy Daniels wrote

Has anyone had any experience with ATA/133 drives
going to "round" cables? I've got a couple new Maxtor
7200 rpm DiamondMax drives with 8MB cache that
need a home, and since my mid-tower case will get a
CD burner and a couple new PCI adapter cards, I
figured the improved airflow afforded by "round"
cables would help control the heat.

Thats basically a myth.

Not if the cable blocks the airflow, dorkbrain

It doesnt, ****wit.

-- and it often does.

In your pathetic little drug crazed fantasyworld, ****wit.

Uh, right.

Even you should be able to do better than that pathetic effort, child.

Try another lie and see if anyone is silly enough to believe it.

On Mon, 4 Aug 2003 09:52:07 +1000, "Rod Speed"
wrote:

Some pathetic excuse for a bull**** artist/pathological liar claiming to be
Joshua P. Hill wrote in message
.. .
Rod Speed wrote
Joshua P. Hill wrote
Rod Speed wrote
Timothy Daniels wrote

Has anyone had any experience with ATA/133 drives
going to "round" cables? I've got a couple new Maxtor
7200 rpm DiamondMax drives with 8MB cache that
need a home, and since my mid-tower case will get a
CD burner and a couple new PCI adapter cards, I
figured the improved airflow afforded by "round"
cables would help control the heat.

Thats basically a myth.

Not if the cable blocks the airflow, dorkbrain

It doesnt, ****wit.

-- and it often does.

In your pathetic little drug crazed fantasyworld, ****wit.

Uh, right.

Even you should be able to do better than that pathetic effort, child.

Try another lie and see if anyone is silly enough to believe it.

There is no sense in debating you: when, as not infrequently happens,
someone demonstrates that you are wrong, you begin to say the same
thing over and over again, like a petulant child.

Josh

Some pathetic excuse for a bull**** artist/pathological liar claiming to be
Joshua P. Hill wrote in message
...
just the pathetic excuse for bull****/pathological lies that it always
ends up having to resort to when its got done like a dinner, yet again.

"Arno Wagner" wrote:
I had the problem with the burner with a short cable (45cm
or 48cm) as far as I remember. I think it was a CoolerMaster
quality cable with shielding. The problem is that while newer
UDMA standards seem to require CRC-checking in all modes,
this burner seems not to do it when writing. I have no idea
whether it sends out CRC codes when reading, but I did not
observe a single read error from correctly burned CDs
and the bit-errors on the incorrect CDs did not change.

The shop where I bought the burner told me that they knew
about the problem and that some writers should not be used
with rounded cables at all, while others can be without
problem. It could be that a firmware-upgrade solves
this problem, but I am too lazy to try. My posting
was just to warn people to be careful, since a few (1..10)
bit-errors in a full CD might not be immediately noticable
but still cause problems.

I have some cheap 60cm (actually cut down 90cm) cables in
use with HDDs for some time now without problems. These are
twisted pair without shielding or brand name.


I just checked the cables I bought from Fry's (which are
the Vantec brand). Each of the 40 colored wires is twisted
together with a white wire (making 80 conductors in all).

A call to Vantec USA in Fremont, Calif., (webpage at
http://vantecusa.com/ide.html ) confirmed that their round
cables have 80 conductors, 40 of them ground wires -
twisted with one of the colored wires.

I also see that the website for SVC says that their round
cables have twisted pairs. (See their webpage at:

http://www.svc.com/cables-ata-100-13...evice-ata-100-
133-round-cables.html

I have no idea whether other round cables have only 40
conductors, but I'd bet that the ones described as "ATA133"
have the 80 conductors with each colored conductor
twisted with a white ground conductor. Could it be that
the round cables that you had problems with only had
40 conductors?


*TimDaniels*

Previously Timothy Daniels wrote:

"Arno Wagner" wrote:
[...]
I have no idea whether other round cables have only 40
conductors, but I'd bet that the ones described as "ATA133"
have the 80 conductors with each colored conductor
twisted with a white ground conductor. Could it be that
the round cables that you had problems with only had
40 conductors?

No, definitely not. Rounded-non-twisted pair cables would have
atrociously bad signal characteristics. I never had/saw such a
cable. What I had where 1-10 single bit errors in a 500MB CD. This is
a relatively low error rate and I at first suspected the burner.
Note that this are not errors on the CD, but just wrong data.

The problem is just that even high-quality rounded cables
have a significantly higher error rate. From UDMA66
on there never was a specification that allowed data transfer
without error checking. For UDMA33 error checking was only
added later. Since CD-R(W) drives usually use UDMA33 or
below, they just might not do the error checking, and I
actually have one that indeed does not check, at least when
writing data. It is a little funny that I never had any
read errors with the drive, so maybe it adds checksums
in the data it sends, but does not verify checksums in the
data it gets. Or maybe it just sends data slower than it
can receive data, which also makes the signal more immune
to disturbances.

I am not saying that this is correct behaviour, but merely
that this is observed behaviour. Burning several CDs
and then verifying their contents against the images
they where burned from can be used to verify whether this
problem exists with a burner. I now just use an old
40-conductor flat cable and have no problems at all with
the burner.

It is possible that new burners do not have this problem
at all.

Regards,
Arno


--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus