Installing UDMA IDE cable backwards with HDD in middle!

[GTS]

Apologies, I know this has come up before but can't find the posts.

Can a UDMA Ampthenol IDE cable be mounted backwards? I.E. the end marked
'Master' used in the motherboard. Then to add to the question, if that
works, can the Master drive be connected to the middle connector instead of
the end (marked system). All cables these days are marked and colour coded,
which makes you think they are wired to only work one way...
Reason? The usual thing - even with a 24" cable there is no way I can
connect my master hard drive to the slave optical - they're simply too far
apart for the distance between the master and slave plugs (it's actually the
same on an 18" or 24") - if I keep the hard drive in the 3.5" caddy. Yes, I
could move the HDD into a 5 1/4 bay and use a caddy or 3.5-51/4 adaptors,
but then I have a cooling issue I'll need to worry about, and I already have
two intake fans blowing over the hard drive holder..
TIA!

 

[Arno Wagner]

Apologies, I know this has come up before but can't find the posts.

Can a UDMA Ampthenol IDE cable be mounted backwards? I.E. the end marked
'Master' used in the motherboard. Then to add to the question, if that
works, can the Master drive be connected to the middle connector instead of
the end (marked system). All cables these days are marked and colour coded,
which makes you think they are wired to only work one way...
Reason? The usual thing - even with a 24" cable there is no way I can
connect my master hard drive to the slave optical - they're simply too far
apart for the distance between the master and slave plugs (it's actually the
same on an 18" or 24") - if I keep the hard drive in the 3.5" caddy. Yes, I
could move the HDD into a 5 1/4 bay and use a caddy or 3.5-51/4 adaptors,
but then I have a cooling issue I'll need to worry about, and I already have
two intake fans blowing over the hard drive holder..
TIA!

I don't know about backwards. But if you do jumper selection on
the drives for matser and slave, their position can be exchanged
without effect on their (logical) behaviour.

Arno

 

[Mike Redrobe]

Apologies, I know this has come up before but can't find the posts.

Can a UDMA Ampthenol IDE cable be mounted backwards? I.E. the end
marked 'Master' used in the motherboard.

Yes, no problem.

Then to add to the question, if that works, can the Master drive be connected
to the middle connector instead of the end (marked system).

If you only have one device/drive, make sure its on the end of the cable,
to prevent signal ringing. I.e. leave the middle connector empy

Even this wouldn't stop a system running though.

All cables these days are marked and colour coded, which makes you
think they are wired to only work one way...

All 80 conductor cables support cable select, in which the drives
decide which is master/slave depending on the position on the cable.

If you've jumpered the drives as master/slave as normal, then
it doesn't matter which order you connect them up in.

 

[Folkert Rienstra]

Apologies, I know this has come up before but can't find the posts.

Can a UDMA Ampthenol IDE cable be mounted backwards? I.E. the end marked
'Master' used in the motherboard.

That will limit UDMA mode to UDMA33 although the Pdiag line on the
drive in the middle may cause a higher mode to be possible if it is not
ATA compliant.
It will short the Pdiag line on the drive at the end.
It may cause that drive to not work or the slave drive not detected.

Then to add to the question, if that works, can the Master drive be
connected to the middle connector instead of the end (marked system).

That is always possible.

All cables these days are marked and colour coded,
which makes you think they are wired to only work one way...

True for UDMA cables, they are keyed and need to be connected in a
very specific way.

Reason? The usual thing - even with a 24" cable there is no way I can
connect my master hard drive to the slave optical - they're simply too far
apart for the distance between the master and slave plugs (it's actually the
same on an 18" or 24") - if I keep the hard drive in the 3.5" caddy. Yes, I
could move the HDD into a 5 1/4 bay and use a caddy or 3.5-51/4 adaptors,
but then I have a cooling issue I'll need to worry about, and I already have
two intake fans blowing over the hard drive holder..

Use a 40-wire cable instead of 80-wire.
Still 30MB/s but none of the other problems

 

[Lil' Dave]

Physically? Yes. Will it work properly 100% of the time? No. The distance
presented on the ide cable between slave and master, and the fact that
master should be at the end of the cable are functions of ATA.

 

[CJT]

Yes, no problem.

I would have said, "No! Problem!"

If you only have one device/drive, make sure its on the end of the cable,
to prevent signal ringing. I.e. leave the middle connector empy

Even this wouldn't stop a system running though.




All 80 conductor cables support cable select, in which the drives
decide which is master/slave depending on the position on the cable.

If you've jumpered the drives as master/slave as normal, then

That was "normal" years ago, but I doubt it is now. Every retail
drive I've bought lately has come jumpered out of the box as CS.

 

[GTS]

Yes, no problem.


If you only have one device/drive, make sure its on the end of the cable,
to prevent signal ringing. I.e. leave the middle connector empy

Even this wouldn't stop a system running though.


All 80 conductor cables support cable select, in which the drives
decide which is master/slave depending on the position on the cable.

If you've jumpered the drives as master/slave as normal, then
it doesn't matter which order you connect them up in.

Thanks for the info. I have installed the master drive in the middle, and
the optical at the end, but managed to do this without having to have the
cable reversed in the end. All is working as it should, and the HDD has not
lost any speed
Ta

 

[Mike Redrobe]

That was "normal" years ago, but I doubt it is now. Every retail
drive I've bought lately has come jumpered out of the box as CS.

You think more installations are running cable select than
jumpered master/slave ?

Whatever state the drive arrives in, most people set it to master or
slave when installing simply because CS is less widely known about,

CS needs the cables to support it - it won't work with a "normal"
40 conductor cable ( all 80conductor cables support it though).

 

[CJT]

You think more installations are running cable select than
jumpered master/slave ?

Certainly I think more NEW installations are.

Whatever state the drive arrives in, most people set it to master or
slave when installing simply because CS is less widely known about,

I think the manufacturers set the default to what they expect most
people to want.

CS needs the cables to support it - it won't work with a "normal"
40 conductor cable ( all 80conductor cables support it though).

How many new installations use 40 conductor cables (thereby losing
much of the speed the drives are capable of)?

 

[Folkert Rienstra]

I would have said, "No! Problem!"

And you would have been correct.

 

[J. Clarke]

You think more installations are running cable select than
jumpered master/slave ?

Whatever state the drive arrives in, most people set it to master or
slave when installing simply because CS is less widely known about,

CS needs the cables to support it - it won't work with a "normal"
40 conductor cable ( all 80conductor cables support it though).

Well, sort of. It actually works fine with a 40 conductor cable on which
the correct wire is cut in the right place, and I've encountered 80
conductor cables that did _not_ have that cut (verified at the connector).

If you can't see the hole then don't assume that it's a cable-select cable
until you have checked at the connectors.

 

[chrisv]

You think more installations are running cable select than
jumpered master/slave ?

Whatever state the drive arrives in, most people set it to master or
slave when installing simply because CS is less widely known about,

I know that I just jumped on the CS bandwagon about a year ago, after
many years of "if it ain't broke don't fix it" fooling around with
master and slave settings. CS is great.

 

[Mike Redrobe]

Well, sort of. It actually works fine with a 40 conductor cable on
which the correct wire is cut in the right place,

Its no longer a "normal" 40 conductor cable after its been cut, is it?

and I've encountered 80 conductor cables that did _not_ have that cut
(verified at the connector).
If you can't see the hole then don't assume that it's a cable-select
cable until you have checked at the connectors.

So much for simplicity...

 

[Lil' Dave]

Well, sort of. It actually works fine with a 40 conductor cable on which
the correct wire is cut in the right place, and I've encountered 80
conductor cables that did _not_ have that cut (verified at the connector).

If you can't see the hole then don't assume that it's a cable-select cable
until you have checked at the connectors.

Some 80 wire ide ribbon cables use an internal terminator, not visible to
the naked eye. Don't rely on seeing a "cut".

 

[Folkert Rienstra]

Well, sort of. It actually works fine with a 40 conductor cable on which
the correct wire is cut in the right place, and I've encountered 80
conductor cables that did _not_ have that cut (verified at the connector).

If you can't see the hole then don't assume that it's a cable-select cable
until you have checked at the connectors.

80-conductor cables never have holes for CS, they have a missing or an
unconnected pin in the middle connector. Some 80-conductor cables do
have a hole (wire cut) for seperating CBLID and PDIAG though. Others
don't and the hostside connector has the seperation inside the connector.

 

[J. Clarke]

Some 80 wire ide ribbon cables use an internal terminator, not visible to
the naked eye. Don't rely on seeing a "cut".

Actually, the presence of the cut is quite reliable as an indicator--if the
cable has it then it's a cable select cable. The _absence_ of a cut is
less so. If it doesn't have it then it may still be one but you can't
determine that by looking at it, you have to test it with a continuity
checker or meter.

 

[Mike Tomlinson]

Yes, no problem.

I'm not so sure. The blue (motherboard) connector grounds one of the
pins to indicate to the controller that an 80-wire cable is fitted. If
the cable is reversed, quite apart from the cable select issues already
mentioned, the controller may think a 40-wire cable is fitted and reduce
data transfers to ATA33.

 

[Lil' Dave]

Actually, the presence of the cut is quite reliable as an indicator--if the
cable has it then it's a cable select cable. The _absence_ of a cut is
less so. If it doesn't have it then it may still be one but you can't
determine that by looking at it, you have to test it with a continuity
checker or meter.

Don't remember saying not seeing a cut is not reliable. Read again.

 

[CJT]

Well as I posted I managed to do it without reversing the cable. However, I
do have the master drive in the middle and the slave at the end. All works
as before, Sandra benchmarks the master drive well above similar ATA-100 8mb
buffer models in the comparison list - so I was curious about your statement
that 'master should be at the end of the cable are functions of ATA.
When I first posted the original question, I did have an inkling that the
master / slave markings on the cable were only for cable select purposes?

I think that's true, but why not just use CS and not worry about it?

 

[CJT]

Just an afterthought - it seems much more gracefull in a standard case to
have the IDE cable from mobo-to HDD-to Optical, as the alternative is to
have the cable doubling back from the optical to the HDD, which is just
messy!

So let the optical drive be the master and be happy.