Sata cabling

  • Thread starter Thread starter Gerry
  • Start date Start date
G

Gerry

I have had a disk connection problem which seems to relate to failing sata
cables. The BIOS has failed intermittently to detect one or both hard
drives. The problem was more obvious with the master drive so I replaced the
cable 14 days ago and there was no further problem until this morning. The
problem this morning was the slave drive so I have replaced the cable for
that drive. It has now been working for a bit over two hour.

The problem first became apparent a month ago when I found the system would
freeze after it had been running some time. Resetting sometimes worked and
sometimes resulted in a failed boot. Eventually the system would boot but
the problem would happen again some hours later or the next day. Sometimes
there have been Event Viewer reports -mainly ID: 11 referring to the
Controller. Often the problem is unreported. This is probably because the
Error is occurring before Event Viewer starts.

From a friend I got these comments.

"In my view, the SATA 'Connector' is an engineering blunder. A sort-of flat
sleeve slides over a notched part on the edge of the board whereupon sit
some exposed/un-insulated traces. Flat conductors encased within a plastic
bit are slid into contact with them. There is no mechanism but friction to
keep the 'connector' in place. Entirely inadequate. It is not designed for
repeated make/break insertion/removal. If subjected even to a low number of
such operations (design spec is 50), it will fail. (5 000 for an eSATA
connector). If I have to repeatedly disconnect-connect a drive during
testing, I replace the cable as a matter of routine."

I am interested in knowing whether others have encountered this problem and
how common place it is?

TIA


--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
Gerry said:
I have had a disk connection problem which seems to relate to failing sata
cables. The BIOS has failed intermittently to detect one or both hard
drives. The problem was more obvious with the master drive so I replaced
the cable 14 days ago and there was no further problem until this morning.
The problem this morning was the slave drive so I have replaced the cable
for that drive. It has now been working for a bit over two hour.

The problem first became apparent a month ago when I found the system
would freeze after it had been running some time. Resetting sometimes
worked and sometimes resulted in a failed boot. Eventually the system
would boot but the problem would happen again some hours later or the next
day. Sometimes there have been Event Viewer reports -mainly ID: 11
referring to the Controller. Often the problem is unreported. This is
probably because the Error is occurring before Event Viewer starts.

From a friend I got these comments.

"In my view, the SATA 'Connector' is an engineering blunder. A sort-of
flat
sleeve slides over a notched part on the edge of the board whereupon sit
some exposed/un-insulated traces. Flat conductors encased within a plastic
bit are slid into contact with them. There is no mechanism but friction to
keep the 'connector' in place. Entirely inadequate. It is not designed for
repeated make/break insertion/removal. If subjected even to a low number
of
such operations (design spec is 50), it will fail. (5 000 for an eSATA
connector). If I have to repeatedly disconnect-connect a drive during
testing, I replace the cable as a matter of routine."

I am interested in knowing whether others have encountered this problem
and how common place it is?

TIA
--

Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
Click to expand...


Gerry:
Understanding that the following comments reflect only my experience working
with a fairly large number & variety of SATA HDDs and associated components
over the past half-dozen years or so...

Along with many computer technicians I've been associated with, virtually
all of us have been dismayed with the design of SATA connectors in terms of
providing a secure connection between the SATA signal cable's connector and
the motherboard's SATA connectors and components with the SATA interface.

However, our problems have basically centered about installing, changing,
modifying PC components on a daily basis and the annoyance we frequently
encounter when the usual type of SATA connector is too-easily disconnected
from its component when it is even slightly jostled. It just doesn't seem
the connection is as secure as it should be.

(We also haven't been thrilled with the design orientation of the SATA
connector in terms of instantly recognizing which way it should be plugged
into a connector/port. Would have preferred a design similar to a Molex plug
or some such.)

Because of this we prefer to use SATA cables having a locking latch or clip
as an integral part of the connector. By & large, it makes for a securer
connection in our experience. Also, the use of a right-angle SATA connector
helps at times. However because of the design/placement of some components'
SATA connector(s) these types of SATA cables are not always available to
use.

Needless to say those of us who work on a daily basis with building,
maintaining, repairing PCs were delighted to see the demise of IDE data
(ribbon) cables and their ilk. Very few, if any of us would want to go back
to those types of cables.

But having said all this there really shouldn't be a problem with the
average PC user insofar as problem connections with SATA cables/devices. For
the overwhelming number of users once the connections are made that's
probably the last time they will be manipulated.

As far as your friend's comments go - we certainly agree with him/her re the
basic connection situation as I have indicated in my above comments. But
while we've connected/disconnected a SATA signal cable multiple times in
this or that system I really can't say I've experienced any problem in that
the cable/connector became defective as a result of these multiple
connects/disconnects. Of course these "multiple" times (again, in my
experience) would be not much more than in the order of a dozen or so
instances in most cases. Again, for the average PC user I don't think this
latter situation would be a problem of any consequence.

One would assume that the average PC user would be making many more
connects/disconnects when using an eSATA cable with some external device.
And as your friend insinuates, the eSATA connector seems to result in a more
secure connection than the SATA connector.

BTW, as you and others may know, the SATA-III data interface is under
development and there has been some talk about redesigning the SATA
connector to achieve a more secure connection. But since the SATA-III data
interface will most likely have to be backwards compatible with present SATA
connectors it's hard to see how they will manage that.

In terms of defective SATA signal cables we have experienced very few
problems in that area other than when the cable's connector (or cable
itself) has obviously been physically damaged or the cable is DOA. Certainly
the amount of defective SATA cables I've come across is far, far less than
the IDE ribbon cable types.

Obviously I don't know (or can even guess) whether a defective SATA cable or
SATA connector caused the problem(s) you've experienced. In my own
experience when I encountered a problem with a defective SATA
cable/connector (rare as it might be), the problem immediately surfaced - it
did not result in an intermittent problem. But I guess it's entirely
possible.
Anna
 
Gerry said:
I have had a disk connection problem which seems to relate to failing sata
cables. The BIOS has failed intermittently to detect one or both hard
drives. The problem was more obvious with the master drive so I replaced the
cable 14 days ago and there was no further problem until this morning. The
problem this morning was the slave drive so I have replaced the cable for
that drive. It has now been working for a bit over two hour.

The problem first became apparent a month ago when I found the system would
freeze after it had been running some time. Resetting sometimes worked and
sometimes resulted in a failed boot. Eventually the system would boot but
the problem would happen again some hours later or the next day. Sometimes
there have been Event Viewer reports -mainly ID: 11 referring to the
Controller. Often the problem is unreported. This is probably because the
Error is occurring before Event Viewer starts.

From a friend I got these comments.

"In my view, the SATA 'Connector' is an engineering blunder. A sort-of flat
sleeve slides over a notched part on the edge of the board whereupon sit
some exposed/un-insulated traces. Flat conductors encased within a plastic
bit are slid into contact with them. There is no mechanism but friction to
keep the 'connector' in place. Entirely inadequate. It is not designed for
repeated make/break insertion/removal. If subjected even to a low number of
such operations (design spec is 50), it will fail. (5 000 for an eSATA
connector). If I have to repeatedly disconnect-connect a drive during
testing, I replace the cable as a matter of routine."

I am interested in knowing whether others have encountered this problem and
how common place it is?

TIA
Click to expand...

The SATA connector design was centered around "server backplane" applications,
making it easy to "plug" a drive into a backplane, for a cable free installation.
The usage of the connectors for desktops, was an afterthought. (The SATA
committee has done a few things, showing a lack of judgment, like their
naming conventions.)

In a backplane application, the drive goes "straight down" onto the connector,
avoiding wiggling and breaking of the wafer. Mechanical guidance of the drive
insertion, helps prevent problems. Motherboard/desktop applications, on the other
hand, have less protection from that (depending on the connector brand). I've
even heard of some users, managing to pull the connector right off
the motherboard (Asrock).

The initial connector design had no positive retention features. Later
connectors fixed that. (But for the locking latch type, both the motherboard
connector and the cable must be compatible. A locking latch cable with a
non locking motherboard, won't help.) My current computer uses no locking
latch, but they did manage to incorporate retention into the design. The
cable won't fall off if I wiggle it. It has a moderate insertion force
to install it (spring loaded dimple ?).

SATA connectors incorporate keying, in the form of the L shaped plastic.

http://upload.wikimedia.org/wikiped...SATA_Data_Cable.jpg/150px-SATA_Data_Cable.jpg

Western Digital SecureConnect, was something that shipped before cables
had lock latches. They used mechanical features present on their brand of hard
drive, to guide the connector into the drive. So that solves the problem
at the hard drive end. Notice that at this point in time, the drive
was still using "Molex" power. AFAIK, other drive brands would not
have the square holes, to fit this cable. The square holes help guide
the connector during insertion.

http://www.wdc.com/en/library/sata/2579-001075.pdf

All of this means, people will have seen a variety of user
experiences. All the way from "no problems here", to "my cable
keeps falling off, so I glued it on" :-)

Paul
 
Not crazy over the SATA connectors either.
SATA or Serial ATA has another design flaw
in that by nature serial data transfers can never
be as fast as parallel.
 
Thanks Anna for your detailed response.

The sata cables could be 3 / 4 years old. I can't say how many times I have
removed and reconnected but it has not been a frequent occurrence. To me
there is no obvious sign of damage.

--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
JS said:
Not crazy over the SATA connectors either.
SATA or Serial ATA has another design flaw
in that by nature serial data transfers can never
be as fast as parallel.
Click to expand...

So that is why SATA2 burst, at over 200MB/sec,
isn't as fast as a burst across Ultra133 at
133MB/sec :-) ?

The article here, shows some *sustained* read
rates above 200MB/sec via SATA2, using recent
SSD products. Hard drives can burst like that,
but they can't sustain a transfer rate like that.
At least not currently. Velociraptor can only sustain
120MB/sec on its SATA2 interface (from early benchmarks).
And that is why, for hard drives as the target of these
technologies, SATA2 is a bit silly. But for other devices
like SSDs, SATA2 or faster might be a welcome addition.

http://www.techreport.com/articles.x/16255/10

The next version of SATA is coming out soon,
at 6Gbit/sec serial rate. So the rate could
potentially double again (236MB/sec from that
chart in Techreport, could become 472MB/sec
with a little work.) It'll be interesting
to see how smooth this introduction of
"SATA3" goes...

http://en.wikipedia.org/wiki/Sata

One purpose for a faster SATA link, is for port multipliers.
A box like the following one, connects five SATA drives
to one SATA cable. If this box was redesigned for
SATA3 cable rates, then reading from all hard disks
simultaneously becomes more practical. But for an
individual hard drive, the SATA3 cable won't be of
much additional benefit. We'll have to wait a
while, to see if Intel makes a SATA3 capable SSD.

http://www.sataport.com/

Paul
 
Paul

The computer is a desktop with a Gigabyte 915/910 Series motherboard. The
two drives are a Seagate and a Maxtor. The cable connectors do not match any
illustrated in your links. Both ends are identical and push vertically down
onto the motherboard. On one side of each end of the cable is a spring
loaded metal plate but I would not say it locks. The cable stays in place.


--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
JS

Interesting. Statistically how do data transfer rates compare?

--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
Well, in a crude sense think of SATA
as a two lane highway compared to
PATA as an 8 lane highway. If the PATA
cable design was updated to handle higher
transfer rates I would think that PATA could
be at least 4x faster than SATA.

And an updated PATA cable need not be a
giant size ribbon cable either. Just imagine if
your ram memory was serial access instead
of DDR2 or DDR3.
 
On the surface, that makes sense - but only on the surface. Let me explain:
In practice, it's a false assumption due to the inherent crosstalk problems
between adjacent signal carrying conductors in the Parallel ATA cable. And
THAT limits the max transfer rate. OTOH, Serial cables do NOT have that
problem, since only a single line is carrying the data. Hence, serial
cables (like in SATA) can be, and are, much faster.
 
JS

If pata is faster than sata why has sata replaced pata?

BTW I have Sata II

--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
They already tried that with the 80 conductor fast ATA cables. While it
helped out a lot, it sure couldn't even come close to SATA.

Now, if you want a 40 or 80 conductor parallel cable that's a meter wide,
with very wide separation between all the signal carrying conductors (of
which there are a LOT, for parallel)..... perhaps that might work. :-)

Bottom line: it's not at all practical.
 
Gerry said:
Paul

The computer is a desktop with a Gigabyte 915/910 Series motherboard. The
two drives are a Seagate and a Maxtor. The cable connectors do not match any
illustrated in your links. Both ends are identical and push vertically down
onto the motherboard. On one side of each end of the cable is a spring
loaded metal plate but I would not say it locks. The cable stays in place.
Click to expand...

The connectors on this example, don't seem to have a shroud around the body.
Just the wafer and some guides. My motherboard has ones like this. My
cables are the part that are providing the retention, as the connector
on the motherboard is not helping.

http://www.gigabyte.com.tw/FileList/Image/motherboard_productimage_ga-8i915pdualgraphic_big.jpg

This is a closeup. This one doesn't support a lock latch,
or at least, doesn't look like the one I have in mind.

http://www.coolgear.com/images/CD011C00A.jpg

This connector has a shroud around the outside,
and that is where you'd put a feature to support
latching. This still isn't the one I saw before.

http://www.coolgear.com/images/CD011D03A.jpg

This one claims to support a latch. Maybe inside somewhere. Not sure.

http://www.satagear.com/CD011D05A_SATA_Connector.html

I'm convinced, if it wasn't for the poorly engineered
first attempt at a connector, there wouldn't be so many
flavors of connectors out there. There is no way to
know, when you buy a motherboard, what you'll get. I've
even seen motherboards, where connector types are mixed,
some with the shroud, and some without. They use
it to distinguish Southbridge ports, from the ports connected
to a separate RAID controller (they're not sprinkled randomly).

And this connector comes with a fastener, to avoid the
embarrassment of one of your customers, ripping the
connector right out of the motherboard, when removing
a cable :-) A good idea.

http://www.coolgear.com/images/CD011D04A.jpg

I've seen at least one connector with a shroud,
where there is a rectangular hole in the shroud,
where the lock latch grips onto. I cannot find
a picture of it right now.

There is a nice collection of pictures here as well.

http://cooldrives.stores.yahoo.net/saiandsaiiin.html

Paul
 
Yes I know about 40 vs 80 design,
what I'm talking about is a total redesign.

And they do have round PATA cables,
not rated any faster but they are not wide and flat.
 
It's was deemed a less expensive solution at the time.

I have PATA 80 wire for ATA 100 and ATA 133 drives
(connected using either round PATA cables or the traditional flat ribbon)
plus SATA I and SATA II drives.

Don't get me wrong, I'm not against SATA
but they just don't deliver on the advertized speed.
Plus the connector design of SATA does not give
me confidence in constantly connecting Hard Drives
on my "Test" computers.
 
JS

The design specification is 50 insertions and removals.

The cable dates from December 2005.

--


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~
 
The last time I experienced similar problems, which led to disks failing the
manufacturers test utility, it was as a result of a slightly iffy pwr
supply, that when put in another PC sometime later, went bang
Pwr supply & disks (2) all replaced under warranty from reputable makers
(Samsung & TruePower)
 
The point is, there is no practical design possible that will allow a
parallel cable to reach the transfer rates of a serial one, for the reasons
stated. Note the emphasis on the word practical (meaning realizable) - not
theoretical (such as with a half meter wide cable, and/or shielded cable(s,)
or whatever). (Shielded cables suffer from increased capacitance which
limits their transfer rate, so even if each data signal cable (16 for 16
bit, or 32 for a 32 bit) were shielded, it's still a no go).
 
Back
Top