Simple ATA ribbon question

[JimboSlim]

Hi Storers

On an 80 conductor ATA100 cable ribbon for connecting master ATA100 and
Slave ATA66 does it matter which connecter goes on the Master drive
The blue end is in the mother board, There is a Black end connecter and a
grey middle connecter. Should the Grey Mid one go in the Slave Disk Or the
Black end one?

Thax for any help

 

[CJT]

Hi Storers

On an 80 conductor ATA100 cable ribbon for connecting master ATA100 and
Slave ATA66 does it matter which connecter goes on the Master drive
The blue end is in the mother board, There is a Black end connecter and a
grey middle connecter. Should the Grey Mid one go in the Slave Disk Or the
Black end one?

Thax for any help

The black one on the end is for the master.

 

[JimboSlim]

Thanx Yep I sent that off a bit to quick I found the answer on the Maxtor
site

The drive placement convention used on an 80-conductor cable is also
different from the previous generation of cable select type cables. The
80-conductor Ultra ATA cables require the master drive (drive 0) to be
installed at the end of the cable and the slave drive (drive 1) to be
installed on the middle connector.
The connectors on 80-conductor cables are also color-coded to help insure
proper drive placement and attachment to the host (system). Typical
color-coding is as follows: blue for attachment to the host (system), black
is for device 0 (master) and gray is for device 1

Cheers Jimbo

 

[Timothy Daniels]

"CJT" responded:

The black one on the end is for the master.

I believe that's true only in the Cable Select mode.
If not in Cable Select, it matters not on which
connectors the Master and Slave are. Of more
concern would be mixing drive speeds on the
same IDE channel (i.e. on the same cable). I've
read that speeds ought not be mixed, but you
might get away with it.


*TimDaniels*

 

[Mark Taylor]

Hello

This only matters if using the cable select jumper on the hard drive. I
personally never use this function and simply use Master/Slave or Single
Drive jumpers accordingly and connect the cable however it best fits into
the system. In many situations I would use the supposedly motherboard
connector to fit to a drive and the drive end to fit to the mainboard. In
many of these situations I might have a CD burner and a LS120 or ZIP drive
and if I use the standard cable connections I cannot fit theminto the case
where I want them. I use many IDE devices in my systems and have never had a
problem. The only time I ahd a problem was with a Western Digital drive
which used as a single drive jumpered as master would not boot. I had to
remve all jumpers, which is the WD default single drive mode, and then it
booted fine.

Regards
Mark

 

[Rod Speed]

CJT wrote

I believe that's true only in the Cable Select mode.
Correct.

If not in Cable Select, it matters not on
which connectors the Master and Slave are.
Correct.

Of more concern would be mixing drive speeds
on the same IDE channel (i.e. on the same cable).
Nope.

I've read that speeds ought not be mixed,

Not anywhere that knows what its about you havent.

but you might get away with it.

Works fine all the time.

The only time you end up with the lowest speed used for both
drives is with some ancient chipsets no longer used much at all.

 

[CJT]

"CJT" responded:




I believe that's true only in the Cable Select mode.
If not in Cable Select, it matters not on which
connectors the Master and Slave are. Of more

It matters if you have only one drive.

 

[Timothy Daniels]

"CJT" pointed out:

It matters if you have only one drive.

If you had only one drive, you couldn't have both
a Master and a Slave as the original poster described.
But you *are* right. And it's to avoid reflections
from that unterminated section of cable that would
otherwise be hanging free.


*TimDaniels*

 

[Timothy Daniels]

"Rod Speed" corrected:

Not anywhere that knows what its about you havent.


Works fine all the time.

The only time you end up with the lowest speed used for both
drives is with some ancient chipsets no longer used much at all.

What you say may be true, AND...SIIG, Inc., for the latest
version of their UltraATA 133 PCI controler card - the one
with 256-byte FIFO buffers - states in their installation
manual:

"In order to achieve high-speed data transfer, a 40-pin/
80-wire Ultra ATA ribbon cable is recommended. Also,
do not mix Ultra ATA/133 hard disk with slower IDE or
ATAPI devices on the same channel."

I don't say that you are wrong, but there should be some
reason for their claim and a reason given by you for your
belief that they are wrong.


*TimDaniels*

 

[Rod Speed]

Rod Speed wrote

What you say may be true,

No 'may be' about it.

AND...SIIG, Inc., for the latest version of their UltraATA
133 PCI controler card - the one with 256-byte FIFO
buffers - states in their installation manual:

"In order to achieve high-speed data transfer, a 40-pin/
80-wire Ultra ATA ribbon cable is recommended. Also,
do not mix Ultra ATA/133 hard disk with slower IDE or
ATAPI devices on the same channel."

They have their head complete up their arse on that last.

AND its completely trivial to prove that using HDTach etc.

I don't say that you are wrong,

Just as well, I'd have you summarily executed if you were stupid
enough to try that. AND send the bill for the bullet to your relos.

but there should be some reason for their claim

See above on heads and arses.

and a reason given by you for your belief that they are wrong.

Even you should be able to use HDTach and prove it for yourself.

 

[Timothy Daniels]

"Rod Speed" arsified:

They have their head complete up their arse on that last.

AND its completely trivial to prove that using HDTach etc.

Please tell more about HDTach. What does it measure?
Where is it available? Does it cost money? Do ads pop
up? Can it be uninstalled easily when one is done with it?


*TimDaniels*

 

[Rod Speed]

Rod Speed wrote

Please tell more about HDTach. What does it measure?

Basically the raw transfer rate to the drive. Bypassing the OS to measure that.

Where is it available?

Even you should be able to put HDTach into google and find it.

Does it cost money?

Not if the drive is formatted FAT32, it does if its formatted NTFS.

For that particular test you can always ghost the
drives, format them FAT32, do the test, then restore
using ghost if they are currently formatted NTFS tho.

Do ads pop up?
Nope.

Can it be uninstalled easily when one is done with it?

Yep. Or you can just ghost the partition its installed on
and restore it after the test if you are completely paranoid.

 

[Timothy Daniels]

"Rod Speed" explained:

Basically the raw transfer rate to the drive. Bypassing the
OS to measure that.

Could it measure the rate of transfer between drives
that are on the same IDE channel? That could be used
to test the effect of differing drive speeds on the same
channel versus identical drive speeds on the same
channel.


*TimDaniels*

 

[Rod Speed]

Rod Speed wrote

Could it measure the rate of transfer between
drives that are on the same IDE channel?

There is no 'the rate of transfer between drives', all
transfer is between the drive and motherboard memory.

That could be used to test the effect of differing drive speeds on the
same channel versus identical drive speeds on the same channel.

Yes, HDTach could certainly be used to see what speed you get with
a particular drive, with and without a slower drive on that ribbon cable.

Thats why I suggested it for that, to prove
that there isnt a problem in that area.

What you can see is an advantage with having the same
speed drives on the same channel with RAID configs tho.

 

[Timothy Daniels]

"Rod Speed" explained:

There is no 'the rate of transfer between drives',
all transfer is between the drive and motherboard
memory.


Yes, HDTach could certainly be used to see what
speed you get with a particular drive, with and without
a slower drive on that ribbon cable.

Are you saying, then, that there is no overhead and
no wasted mode-switching of any sort as the I/O
system serves one drive speed and then the other?
IOW, there is no "adjustment" - perhaps in something
like DMA mode - and all disk I/O can be modelled
as independent disks interacting only with the mother-
board and irrespective of each other?


*TimDaniels*

 

[Rod Speed]

Rod Speed wrote

Are you saying, then, that there is no overhead and
no wasted mode-switching of any sort as the I/O
system serves one drive speed and then the other?

Correct. A slower drive thats not being used
wont have any effect on the speed of access
to the faster drive on the same ribbon cable.

IOW, there is no "adjustment" - perhaps in something
like DMA mode - and all disk I/O can be modelled
as independent disks interacting only with the mother-
board and irrespective of each other?

Correct.

 

[bill]

"Rod Speed" explained:


Are you saying, then, that there is no overhead and
no wasted mode-switching of any sort as the I/O
system serves one drive speed and then the other?
IOW, there is no "adjustment" - perhaps in something
like DMA mode - and all disk I/O can be modelled
as independent disks interacting only with the mother-
board and irrespective of each other?


*TimDaniels*

What you are describing is called "independent timing", and has been a
feature of chipsets for some time.
Go here for a description:

http://www.storagereview.com/guide2000/ref/hdd/if/ide/confTiming.html

 

[nuke]

Also,

They have their head complete up their arse on that last.

Unless there is some deficiency in their controller such that it cannot
maintain independent timing for each disk on the bus. Some controllers still
can't.

 

[Folkert Rienstra]

Unless there is some deficiency in their controller such that it cannot
maintain independent timing for each disk on the bus.

Some controllers still can't.

And those are?