Best strategy hard drive swapping?

[HaHaHoHoHeeHee]

There is no vibration problem with any HD mfg-ed since 1999.

I don't believe that for a moment. It is contrary to common sense with a
disk that spins at high speeds and your having the hard drive sitting
out in the open where it is subject to the slightest jar. That is one
reason why the caddys you recommend often feature extra shock absorbers.
Hard drives need to be securely mounted, imo, except for short testing
periods.

Depends on the mobo BIOS and mobo ATA controller. There is no ATA
80GB limit per se.


IDE interfaces are NOT "so touchy".


Good 24 inch cables, while not in spec, generally work.

It is amazing to me that the recommended limit is 18 inches and even
given that 24 inch cables might work, that shows how touch the IDE
interface is. Or maybe it's the MBs themselves or general design. Point
is that the whole standard is rather flaky if at electron speeds they
cannot design cables/etc that can be extendable for reasonable lengths.

 

[Ron Reaugh]

I don't believe that for a moment. It is contrary to common sense with a
disk that spins at high speeds and your having the hard drive sitting
out in the open where it is subject to the slightest jar.

Right but that is NOT a 'vibration problem'.

That is one
reason why the caddys you recommend often feature extra shock absorbers.
Hard drives need to be securely mounted, imo, except for short testing
periods.


It is amazing to me that the recommended limit is 18 inches and even
given that 24 inch cables might work, that shows how touch the IDE
interface is.

Nope, the cable length limit with ultra SCSI and 4 or more devices is 5
feet....that's tounchy.

Or maybe it's the MBs themselves or general design. Point
is that the whole standard is rather flaky if at electron speeds they
cannot design cables/etc that can be extendable for reasonable lengths.

HUH?

 

[Rod Speed]

Nope.

Yep, have a look at the specs of the cable in the standard.

Those bays use nothing like that between the
drive connector and the controller connector.

Nope.

Yep, have a look at the specs of the cable in the standard.

Those bays use nothing like that between the
drive connector and the controller connector.

 

[Rod Speed]

I keep the cases off most my machines and have HDs hanging or sitting on the
desktop for this purpose

Some drives get stinking hot surprisingly quickly when used like that.

but trays work.

Sometimes they do, sometimes they dont

Yes, using an SATA addon card.

The mobo BIOS may not support large HDs when using the mobo's IDE controller.
Use an addon PCI card that does support large HDs.

Makes more sense to flash the bios if they dont.

Ignore such nay sayers.

Ignore fools so stupid that they didnt even
notice the problem with the IBN 75GXPs.

Ignore all such recommendation from wacko speedo.

Ignore fools so stupid that they didnt even
notice the problem with the IBN 75GXPs.

Assume no problems as the SATA will be via an addon card.

I use the KF-83.

 

[overload]

It is amazing to me that the recommended limit is 18 inches and even
given that 24 inch cables might work, that shows how touch the IDE
interface is. Or maybe it's the MBs themselves or general design. Point
is that the whole standard is rather flaky if at electron speeds they
cannot design cables/etc that can be extendable for reasonable lengths.

The problem is electron speed. Electrons flowing through wires
generate an induced magnetic field that effects other wires. Better
(not IDE) cables use twisted pairs and differential signaling so that
any induced signals affect both sides of the signal oppositely and
cancel out. Flat cables don't work well at high speeds.

 

[Rod Speed]

(e-mail address removed) wrote

True.

Nope.

Yep. IDE went with unterminated drives for the simplicity
with those and a simple flat ribbon cable for the same reason.

That was later improved with 80 wire cables with
a ground wire between each signal wire, but that
doesnt help with the length, just the speed.

They can but chose not to. It was only intended for
internal use and the simplicity of unterminated drives.

The problem is electron speed.
Nope.

Electrons flowing through wires generate an
induced magnetic field that effects other wires.

Nothing to do with electron speed,
thats the rise time of the signal edges.

Better (not IDE) cables use twisted pairs and
differential signaling so that any induced signals affect
both sides of the signal oppositely and cancel out.

Correct. But that has nothing to do with electron
speed which is the same in both types of cable.

Flat cables don't work well at high speeds.

Well enough if they arent too long.

 

[Timothy Daniels]

Thanks for the reply. Hate to sound ungrateful, but an earlier thread in
this group (see google) argues that these Kingwin drives are not ATA
standard compatible and have caused problems especially when using
multiple drives. This is not my experience but comments by others here.

I have 3 hard drives in my PC, and one of them is in a removable
tray. I have them all connected to a PCI ATA/133 controller card via
shielded round cables. I've had no discernable problems with them
in the 2 years I've had the setup. And yes, such hardware does not
conform to ATA specs. So? Does that mean they don't work?
Some of the student labs at UCLA are also set up with such removable
trays for students' hard drives, and they've had no problems with them.
Engineering specs are just guidelines - like clock rates on a CPU.
That doesn't stop gamers from overclocking. Sometimes you just
gotta try something for yourself. In my case, the convenience is an
overwhelming argument for using removable trays. They make
archiving multiple bootable copies of the entire system practical,
and they allow fast switchover to a working system possible in the
event of a primary HD failure. Given the cost of a rack/tray setup
(as low as $14 plus shipping), why not give it a try?

*TimDaniels*

 

[Timothy Daniels]

Thanks for the reply. Hate to sound ungrateful, but an earlier thread in
this group (see google) argues that these Kingwin drives are not ATA
standard compatible and have caused problems especially when using
multiple drives. This is not my experience but comments by others here.

"Hard drive problems" are frequently *heat* problems. That's why I
chose the Kingwin tray with the flat fan in the bottom of the tray. Kingwin
also makes the kind with 0 thru 2 fans in the front with room for a 3rd,
but a hardware review of the flat fan gave it high marks for cooling
effectiveness. My experience bears that out. At no time does the HD
mounted inside the removable tray feel any warmer than one mounted
normally inside the case. In fact, it feels cooler. Maybe that's why I've
had no problems with the removable drive.

*TimDaniels*

 

[Rod Speed]

"Hard drive problems" are frequently *heat* problems. That's why I
chose the Kingwin tray with the flat fan in the bottom of the tray. Kingwin
also makes the kind with 0 thru 2 fans in the front with room
for a 3rd, but a hardware review of the flat fan gave it high marks
for cooling effectiveness. My experience bears that out. At no time
does the HD mounted inside the removable tray feel any warmer than
one mounted normally inside the case. In fact, it feels cooler. Maybe that's
why I've had no problems with the removable drive.

Nope, plenty have had other problems like the one whoever it was
rubbed your nose in, the drive not always being visible at boot time.

 

[J. Clarke]

Thanks for the reply, but I think the above is a very bad idea. Even if
you are careful you still have problems with vibrations, etc and the
airflow to the box is not correct with the case removed (i usually use a
outside fan when I remove the case).

I've noticed no difference in reliability or stability or operating
temperatures with the case open or closed on most machines. If there's a
problem it's generally on a tightly cased machine with marginal cooling and
often leaving the case open improves the situation. PCs just aren't that
fragile.

I want to limit the amount of new hardware at this time. Will stick to
ATA, though I think this limits me to 80GB per drive?

What ever gave you that idea?
<http://www.newegg.com/Product/Product.asp?Item=N82E16822148068>

Your particular motherboard may have an 8, 32, or 128 gig limit, depending
on its age, and that may or may not be addressable by a BIOS upgrade.

Well it makes some sense to me since the IDE interfaces are so touchy
such that adapters and longer cables are not recommended. I would think
that their could be problems with the additional connections with a
caddy.

They're not _that_ touchy. The caddies work fine as long as you don't try
to hot-swap them. Cold swap is no different than pulling a drive out and
replacing it.

looks good, but as I said others have had problems with them at least
reported in this group. maybe not related to the hardware, maybe their
configurations or user error.

Probably user error. I've had no trouble with Kingwin. But then I notice
that I seem to have far less trouble with my systems than do many posters.

originally I just thought since electrons travel at the speed of light
there should be no problem with a longer cable,but others have warned me
against this.

(a) Electrons do not travel at the speed of light. If all the energy in the
universe were collected and all the matter converted to energy and that
energy was applied to a single electron it would not reach the velocity of
light, although it would be pretty close. The propagation velocity in
wires is typically 60-80% of the velocity of light.

(b) The velocity makes no difference--interference and cross talk are the
issues. Television and radio broadcasts _do_ travel at the velocity of
light, but they are not free from interference.

 

[Lil' Dave]

Can't speak for others, including you. For some oldish PCs I've had, I used
caddies. I NEVER used the onboard ide connectors. Rather, I used an
identical Promise ide controller in each. No bios recognition differences
as a consequence. Used Kingston caddies, forgot the model number, cold swap
only. Standard 18" 80 wire ribbon cable was barely long enough in tower
case. Just be careful if you're using the WD made drives. You have to
consider master alone or master w/slave jumper pin settings, if you choose
instead of cable select.

 

[HaHaHoHoHeeHee]

(e-mail address removed) wrote

Yep. IDE went with unterminated drives for the simplicity
with those and a simple flat ribbon cable for the same reason.

That was later improved with 80 wire cables with
a ground wire between each signal wire, but that
doesnt help with the length, just the speed.


They can but chose not to. It was only intended for
internal use and the simplicity of unterminated drives.


Nothing to do with electron speed,
thats the rise time of the signal edges.


Correct. But that has nothing to do with electron
speed which is the same in both types of cable.


Well enough if they arent too long.

So, the suggestion of the fellow who recommended round cables, is
that worth pursuing. IOW is there any way I can run any type of
cable that would be long enough, say around 3-4 feet so that I
could hook up drives on that ide cable outside the box? Sorry if
that's a dumb question,but my background in digital electronics is
limited.

 

[HaHaHoHoHeeHee]

I've noticed no difference in reliability or stability or
operating temperatures with the case open or closed on most
machines. If there's a problem it's generally on a tightly
cased machine with marginal cooling and often leaving the case
open improves the situation. PCs just aren't that fragile.

Did you actually test this? Seems to me that the design and
location of the fans should be strategically placed such that with
the case closed the air flow is directed where it needs to be. I'm
inclined to thing you are right about this and the important thing
is ambient temperature and operator use (heavy or light). However
i do not like the idea of using one of these caddys that has no
fan.

What ever gave you that idea?
<http://www.newegg.com/Product/Product.asp?Item=N82E16822148068>

another post from someone that sounded like he knew what he was
talking about,haha. Fool me. If I took time to check everything i
read on the net, I would waste more time than I usually do, so I
just ask and hope someone will correct me ;-).

Your particular motherboard may have an 8, 32, or 128 gig limit,
depending on its age, and that may or may not be addressable by
a BIOS upgrade.

Ok, should go to dell support forums to answer this. And I gues
the usenet for the AMD 850MHZ Duron clone, which I bought second
hand, so no docs.

They're not _that_ touchy. The caddies work fine as long as you
don't try to hot-swap them. Cold swap is no different than
pulling a drive out and replacing it.

Ok, I'll try them, but they are expensive, when you compare their
pricing to the costs of new systems and hard drives. The good ones
are anyways. Not gonna spend more than $50 tops for one of these
suckers.

Probably user error. I've had no trouble with Kingwin. But
then I notice that I seem to have far less trouble with my
systems than do many posters.


(a) Electrons do not travel at the speed of light. If all the
energy in the universe were collected and all the matter
converted to energy and that energy was applied to a single
electron it would not reach the velocity of light, although it
would be pretty close. The propagation velocity in wires is
typically 60-80% of the velocity of light.

(b) The velocity makes no difference--interference and cross
talk are the issues. Television and radio broadcasts _do_
travel at the velocity of light, but they are not free from
interference.

Thanks for the information, interesting to know this. So
theoretically if I could find really good cables (is that the weak
link), I could run an ide cable out of the box to whatever drive I
plugged it into?

 

[HaHaHoHoHeeHee]

Can't speak for others, including you. For some oldish PCs I've
had, I used caddies. I NEVER used the onboard ide connectors.
Rather, I used an identical Promise ide controller in each. No
bios recognition differences as a consequence. Used Kingston
caddies, forgot the model number, cold swap only. Standard 18"
80 wire ribbon cable was barely long enough in tower case. Just
be careful if you're using the WD made drives. You have to
consider master alone or master w/slave jumper pin settings, if
you choose instead of cable select.

message

Thanks for the information. Yes, I will watch the WD thing.
How expensive/difficult is it to install the promise controllers?
Why exactly is this better? Thanks.

 

[J. Clarke]

So, the suggestion of the fellow who recommended round cables, is
that worth pursuing. IOW is there any way I can run any type of
cable that would be long enough, say around 3-4 feet so that I
could hook up drives on that ide cable outside the box? Sorry if
that's a dumb question,but my background in digital electronics is
limited.

Round IDE cables are usually flat cables that have been hacked up and
wrapped in a piece of tubing. You won't get any more distance out of them
than flat cables and they're more out of spec than any halfway decent caddy
would ever be. Their main utility is to make geek-chic machines with
windows in the side look k3w1 to wannabees and provide a good laugh to
ubergeeks.

If you want more span then you need to go to a different signalling
method--you can get an SATA-to-PATA bridge or a Firewire bridge or a USB2
bridge, all of which will allow you longer cables than standard PATA, but
you'll then have the bridge to contend with which can introduce problems of
its own.

 

[J. Clarke]

Did you actually test this?

Not systematically, but I have machines that have been running for years
with the case open.

Seems to me that the design and
location of the fans should be strategically placed such that with
the case closed the air flow is directed where it needs to be.

In general the CPU has its own fan, beyond that the video board if it does't
have a fan is usually in dead air with the case closed. The only thing
that may not get sufficient flow is the disks if you're relying on the
power supply fan to pull enough air into the case, but that's iffy--all my
machines have dedicated fans for the disks.

I'm
inclined to thing you are right about this and the important thing
is ambient temperature and operator use (heavy or light). However
i do not like the idea of using one of these caddys that has no
fan.

Caddys are a special case--they tend to be tight. I've never used a fanless
one--some of the Kingwins have multiple fans.

another post from someone that sounded like he knew what he was
talking about,haha. Fool me. If I took time to check everything i
read on the net, I would waste more time than I usually do, so I
just ask and hope someone will correct me ;-).


Ok, should go to dell support forums to answer this. And I gues
the usenet for the AMD 850MHZ Duron clone, which I bought second
hand, so no docs.


Ok, I'll try them, but they are expensive, when you compare their
pricing to the costs of new systems and hard drives. The good ones
are anyways. Not gonna spend more than $50 tops for one of these
suckers.

$50?!?!?!? Try $20 for an aluminum rack with fan. And about 15 for each
extra tray.

Thanks for the information, interesting to know this. So
theoretically if I could find really good cables (is that the weak
link), I could run an ide cable out of the box to whatever drive I
plugged it into?

If you could find some kind of hypothetical super cable you might be able to
extend the span a little, but there are no super cables for IDE, no matter
what some cable vendor would have you think.

 

[HaHaHoHoHeeHee]

Round IDE cables are usually flat cables that have been hacked
up and wrapped in a piece of tubing. You won't get any more
distance out of them than flat cables and they're more out of
spec than any halfway decent caddy would ever be. Their main
utility is to make geek-chic machines with windows in the side
look k3w1 to wannabees and provide a good laugh to ubergeeks.

If you want more span then you need to go to a different
signalling method--you can get an SATA-to-PATA bridge or a
Firewire bridge or a USB2 bridge, all of which will allow you
longer cables than standard PATA, but you'll then have the
bridge to contend with which can introduce problems of its own.

Ok, thanks, looks like caddies are the best way to go if I must
have easily removable IDE drives. I took a quick look at three
different mfg. pages, including kingwin and all the cheaper models
were plastic, no fan for like $30. Guess I did not look hard
enough. If you have a model number for the $20 alum. fan model,
can you give that to me? Thanks.

 

[Rod Speed]

So, the suggestion of the fellow who recommended
round cables, is that worth pursuing.

Nope. they flout the standard even more comprehensively.

If you need longer cables, use SATA instead.

IOW is there any way I can run any type of cable that
would be long enough, say around 3-4 feet so that I
could hook up drives on that ide cable outside the box?

Yes, SATA can do that. So can USB2 and firewire

Sorry if that's a dumb question,

Nar, its a very sensible question.

 

[Rod Speed]

Did you actually test this?

I do, and he's right.

Seems to me that the design and location of the fans
should be strategically placed such that with the case
closed the air flow is directed where it needs to be.

In practice no covers works even better because with
modern systems the stuff that gets hottest like the cpu
and the video card has its own fans and if you do want
to stack all the hard drives up with no free slot between
them, you really do need a fan blowing air over the stack.

I'm inclined to thing you are right about this and the important
thing is ambient temperature and operator use (heavy or light).

I monitor the hard drive temp using SMART and make sure it
never does get to what I consider is undesirable, and thats
well below the max that the hard drive manufacture allows.

And you can see the hard drive temp
move up with stuff like formatting the drive.

However i do not like the idea of using
one of these caddys that has no fan.

Yeah, they mostly do severely restrict the airflow over
the drives and some drives like the Barracudas get rid
of most of their heat by conduction to the metal drive
bay stack and that isnt possible when its in a caddy.

Plastic caddys are particularly bad.

another post from someone that sounded like he knew what
he was talking about,haha. Fool me. If I took time to check
everything i read on the net, I would waste more time than I
usually do, so I just ask and hope someone will correct me ;-).

Yeah, thats the best approach.

Ok, should go to dell support forums to answer this.
And I guess the usenet for the AMD 850MHZ Duron
clone, which I bought second hand, so no docs.

Everest should be able to tell you who's motherboard
is being used and you can likely get the manual off the
manufacturer's site and check for bios updates.

Ok, I'll try them, but they are expensive, when you compare
their pricing to the costs of new systems and hard drives.
The good ones are anyways. Not gonna spend more than
$50 tops for one of these suckers.

And you really need to compare the total cost with something
as basic as just a new large hard drive and a lan etc too.

It wasnt user error with the fella that had a problem with
the drive being seen reliably at boot time in the caddy
but not when the drive was installed internally instead.

Thanks for the information, interesting to know this. So
theoretically if I could find really good cables (is that the weak
link), I could run an ide cable out of the box to whatever drive I
plugged it into?

Nope, thats not a good idea with IDE. Much
better to use SATA, USB2 or firewire instead.

 

[Timothy Daniels]

Ok, thanks, looks like caddies are the best way to go if I must
have easily removable IDE drives. I took a quick look at three
different mfg. pages, including kingwin and all the cheaper models
were plastic, no fan for like $30. Guess I did not look hard
enough. If you have a model number for the $20 alum. fan model,
can you give that to me? Thanks.

Here it is again is the direct link to the Kingwin model with the
fan in the bottom of the tray:
http://www.kingwin.com/pdut_detail.asp?LineID=&CateID=25&ID=136
As you can see, the model no. is KF-101-IPF .
If you want it in black, here is the link:
http://www.kingwin.com/pdut_detail.asp?LineID=&CateID=25&ID=135
And as you can see, that model no. KF-101-IPF-B .

Now go to Nextag.com and enter the model no. and do the search,
and you'll find the tray and rack pair going for $25. Search more
diligently, and you'll find lower prices.

*TimDaniels*