Is vertical mounting a bad idea?

[J.Clarke]

That's sounds like a reasonable explanation. A round-wire finger-guard
like what a power supply uses wouldn't work well enough for that?

If it's making metal-to-metal contact with the case and the wire spacing
is close enough. But that's an additional assembly step that most PC
manufacturers don't want to spend the labor to perform--on one machine
it's peanuts, on a few thousand it adds up.

I wonder how many of us have cases that meet those regulations, with
missing slot covers and the like. I had to 'snap out' a slot cover to
install my HD in a 5.25" slot and there's no way to cover it up again.
I also have the side covers off my case for the time being.

I suspect that most hobbyists machine leak like sieves, but unless our
machine is generating enough noise to result in a complaint being filed,
we don't get investigated. OTOH, a PC manufacturer, even the strip-mall
computer shop variety, can get in fairly serious trouble if their
machines, as-shipped, don't comply.

 

[Rod Speed]

Necessary for ME, to do what I wanted.

Wrong. Again.

And you previously claimed that that config was NECESSARY
for on the fly CD copying. You are just plain wrong.

Judging from this group, both configurations are common.

Separate issue entirely to what is
NECESSARY for on the fly CD copying.

Your "massive overkill" is probably no more than an extra $20 if you
compare cases with 4 or 8 3.5" bays from the same manufacturer.

'same manufacturer' is completely irrelevant andyou are
just plain wrong on that $20 claim anyway with some of them.

What's your problem with that?

Thats its stupid to be claiming that that is anything like
the 'best mid-tower case for a home system' when
almost no one demands that stupid requirement.

So a $1500 system goes up to $1520. Big deal.

You've just plucked those utterly bogus numbers out of your arse.

Most all cases intended for home use are cheap
as dirt now compared to what they used to cost.

Completely irrelevant to whether 'best mid-tower case for
a home system' needs to have that many 3.5" drive bays.

I usually spend more for one night out.

Your problem.

The Lian-Li PC-60 is no taller, about an inch wider and
5 inches deeper (~23") than my present case (~18").

Completely irrelevant to whether its anything like
the 'best mid-tower case for a home system'

Pity its MUCH more expensive than
perfectly adequate mid tower cases.

That's not a problem for me so I think it's a good
trade-off for the extra good cooling and add-on room.

More fool you.

Other cases with 6-8 3.5" bays

Only a fool demands that for a home system.

may be smaller - I haven't looked around much yet.
I've had much taller cases (full tower) but not one
this deep. Maybe the case is too big for you,

Wrong. Again. Just ridiculously over priced.

and that's fine, and why there's hundreds of
cases to choose from - one size does not fit all.

Gone blind yet ?

I know, the Lian-Li has 8.

Pathetic, really.

I could have bought an add-on cage to hold two more
3.5" drives for the case I have now but I didn't do it because
I started with one drive and didn't plan ahead very well.
Too late now. The cooling would have sucked anyway.

You dont need a case with 8 3.5" drive bays.

Very helpful.

Pathetic, really.

 

[Rod Speed]

Some pathetic excuse for a bullshit artist desperately cowering behind
just the puerile attempts at insults any 3 year old could leave for dead.

 

[Timothy Daniels]

That cant be done with a PATA drive because the differential
signals that are used in that config arent even present.

What can be done is to have a twisted
pair with one of the wires a ground wire.

I just checked my round cables (not yet installed), which
were made by Vantec, and there are 80 conductors where
each of the 40 colored conductors is twisted with a white
conductor. A call to Vantec confirmed that the 40 white
conductors are grounded. I also looked at the SVC website,
and it says that its round IDE cables consist of 80 conductors
with 40 twisted pairs. I have no idea if other brands have
only 40 conductors, but it would appear that the twisting of
a signal conductor with a ground conductor would go far to
cut down crosstalk - as would opting for shorter cables
when one has a choice.


*TimDaniels*

 

[Rod Speed]

I just checked my round cables (not yet installed), which
were made by Vantec, and there are 80 conductors where
each of the 40 colored conductors is twisted with a white
conductor. A call to Vantec confirmed that the 40 white
conductors are grounded. I also looked at the SVC website,
and it says that its round IDE cables consist of 80 conductors
with 40 twisted pairs.

Trouble is that the standard specifys something completely
different, a flat ribbon cable with every second wire grounded.

Electrically they aint the same at all.

I have no idea if other brands have only 40 conductors,

I doubt it with better than ATA33 cables. They are usually
either that twisted pair arrangement or a ribbon cable rolled up.

Both approaches are electrically very different to what the standard
specifys. They are often longer than the standard allows too.

In short they flout the standard.

but it would appear that the twisting of a signal conductor
with a ground conductor would go far to cut down crosstalk

Its more complicated than that. Twisted pairs should be driven
differentially, just just a ground twisted with the signal line.

- as would opting for shorter cables when one has a choice.

Makes a hell of a lot more sense to use flat 80 conductor ribbon
cables that meets the standard instead. Even if that means that
you have to choose a case that allows the use of those.

 

[Timothy Daniels]

Trouble is that the standard specifys something completely
different, a flat ribbon cable with every second wire grounded.

Electrically they aint the same at all.


I doubt it with better than ATA33 cables. They are usually
either that twisted pair arrangement or a ribbon cable rolled up.

Both approaches are electrically very different to what the standard
specifys. They are often longer than the standard allows too.

In short they flout the standard.


Its more complicated than that. Twisted pairs should be driven
differentially, just just a ground twisted with the signal line.


Makes a hell of a lot more sense to use flat 80 conductor ribbon
cables that meets the standard instead. Even if that means that
you have to choose a case that allows the use of those.

I agree with you that round cables flout the specs. I also agree
that twisting provides the best noise immunity when the pair of
wires are driven differentially. But there is still a degree of
immunity provided when one of the wires in a pair is grounded.
The signal travelling along the pair is still just the instantaneous
value of the voltage between the "signal" wire and the ground
wire instead of the voltage between the "plus" and "minus" signal
wires. It has only half the noise immunity because the signal is
half what it would be if paired with its opposite value, and so
environmental noise seems twice as large. In the case of the 80-
conductor ribbon cable, the ground wires are on two sides of
signal wire, but they are not twisted, which makes them vulnerable
to certain noise modes that twisting protects against. The round
cable also provides an outer metal braid for shielding, which the
ribbon cable does not. The bottom line, I think, is that round cable
is harder to specify and so standards for round cable were just
never defined. So round cables do flout the specs, but that doesn't
necessarily mean that their noise immunity is worse. It just means
that "your mileage may vary" somewhere between Better and
Worse.


*TimDaniels*

 

[Timothy Daniels]

....I'm considering a Lian-Li case that has vertical HD
mounting but I worry about the possibility of gyroscopic effects
causing a head crash if the computer is accidently moved while
operating. Is horizontal mounting safer? Maybe there's a reason why
practically every other case manufacturer uses horizontal mounting
even though vertical mounting is more space efficient.

Dell makes pretty good PCs and PC cases, and my Dell Dimension
mounts its hard drive(s) vertically. The location is right at the
bottom
front of the case where it gets a steady supply of cool room temp air,
and being vertical, the hard drive(s) also get the added benefit of
convection currents - really good in case one of the exhaust fans fail.
For cooling considerations, I think vertical mounting is superior to
horizontal mounting for *anything* (as long as there is ambient air
and gravity has an effect).


*TimDaniels*

 

[Rod Speed]

I agree with you that round cables flout the specs. I also agree
that twisting provides the best noise immunity when the pair of
wires are driven differentially. But there is still a degree of
immunity provided when one of the wires in a pair is grounded.

Not really, particularly when you have sets of those
twisted pairs all crammed together in an outer jacket.

The signal travelling along the pair is still just the instantaneous
value of the voltage between the "signal" wire and the ground wire

Nope, all the grounds are commoned at each connector, the
line receiver doesnt see both the ground and the signal line.

instead of the voltage between the "plus" and "minus" signal wires.

Yep, thats the only way a twisted pair can work.

It has only half the noise immunity

Thats nothing like the electrical reality.

because the signal is half what it
would be if paired with its opposite value,

It doesnt work like that.

and so environmental noise seems twice as large.

Nor that either.

In the case of the 80-conductor ribbon cable,
the ground wires are on two sides of signal wire,

Yes, the ground wires provide significant electrical isolation between
the adjacent signal lines. And a more stable impedance as well.

but they are not twisted, which makes them vulnerable
to certain noise modes that twisting protects against.

Not when it aint driven differentially it doesnt.

AND the drivers and receivers are designed for the
electrical characteristics of that particular config.

The round cable also provides an outer metal braid for shielding,

These are digital lines, not analog. They dont need a shield braid.

which the ribbon cable does not.

Doesnt need that.

The bottom line, I think, is that round cable is harder to specify
and so standards for round cable were just never defined.

Thats must plain wrong too. Round cables with twisted pairs
are in fact defined in some standards like CAT5 and SATA.

So round cables do flout the specs, but that doesn't
necessarily mean that their noise immunity is worse.

It aint about immunity to noise, its about crosstalk
and avoiding impedance discontinuitys which
produce reflections with unterminated cables.

It just means that "your mileage may vary"
somewhere between Better and Worse.

Wrong. The drivers and receivers are designed
to get adequate performance out of the correctly
specified 80 conductor flat ribbon cable.

Pointless utterly flouting those specs.

If you do want a round cable for whatever reason,
the only thing that makes any sense is to use SATA
instead which just happens to have proper formally
specified round cable in that standard.

Round PATA cables are only used by those
who havent got a clue about the electrical basics.

 

[Folkert Rienstra]

Big deal. Some wires will cross each other a few times along the
length of the cable which is nothing like being aligned all the way.

But?

They are basically the same when the electrical specs are similar.
The SE impedance of twist and flat cables is in the 80-90 ohms.
That is within the ATA impedance spec.

Nope, not electrically.

That is exactly why (and only why) it is used in SCSI, to cut crosstalk.

Nonsense (and obviously in more than one way).

I agree with you that round cables flout the specs.

Oh? Why? (except for the pictorials in the specs).

I also agree that twisting provides the best noise immunity
when the pair of wires are driven differentially.

Oh? Why?
Twisted has nothing whatsoever to do with differential signalling.
Twisting diminishes susceptability to pick up or emitting stray fields
that go in one direction. Differential signalling accomplishes the same.
Using them together is kind of a waste except that in differential
use, crosstalk is diminished by the use of twisted conductors.

But there is still a degree of immunity provi-
ded when one of the wires in a pair is grounded.

Nope, there is no difference. There will not be any common
mode rejection but that has nothing to do with the cable itself.

The signal travelling along the pair is still just the instantaneous
value of the voltage between the "signal" wire and the ground
wire instead of the voltage between the "plus" and "minus" signal
wires. It has only half the noise immunity because the signal is
half what it would be if paired with its opposite value, and so
environmental noise seems twice as large.

Nonsense. We are talking digital signals here, not analog.

In the case of the 80-
conductor ribbon cable, the ground wires are on two sides of
signal wire, but they are not twisted, which makes them vulnerable
to certain noise modes that twisting protects against.

Well, if that was a problem they would have specified 'twist and flat'
instead of flat. Note that flat is just the most common ribbon type,
the spec itself doesnt use 'flat' anywhere other than in the pictorials.

The round cable also provides an outer metal braid for shielding,

Nope, most do not.

which the ribbon cable does not.

The bottom line, I think, is that round cable is harder to specify

Not really, just more words to use. It would certainly be a lot
harder to check whether a cable would comply with them though.

and so standards for round cable were just never defined.

So round cables do flout the specs,

Not the rolled-up flat ones.

 

[Folkert Rienstra]

"Rod Speed",
you have an interesting way of throwing light on a subject.

Your comments
Commenting.

are illucidating, and educational.

About the person making them. Not the comments themselfs, they ain't.

 

[troll shafter]

Some pathetic excuse for a troll claiming to be
message just the usual pathetic excuse for a troll thats all it can ever manage.

Try harder, child. You might actually manage to fool someone, sometime.