Frail Scratch Disk

  • Thread starter Thread starter jeremy
  • Start date Start date
J

jeremy

How easy would it be to fake the presence of the drive? How much
information does windows need?

If I can record the drive information (either by putting the requisite
signals into the IDE port on the drive and recording the results, or just
by getting the relevant infofrom the manufacturer), would it be that hard
to build a PIC based device that pretends it is the disk? From there, it's
just a matter of swapping the connectors and issuing a spinup command I
assume.

Does anyone know of a good place to read up on what goes on with detection
and control of IDE devices?

Ewan

Wow, this man is on a mission! While most people would just grab one of those
WD800JB 80GB 8MB cache drives for $49 after rebates, he
wants to design new hardware using a microcontroller.

Sorry, without googling I don't have any good detailed IDE links to give.

I do not believe the IDE controller, OS drivers, or drive itself was
designed to be hot swapped. Even with your proposed "emulator" in place
at boot, I have a feeling your physical swapping of devices later is not a
good idea.

There may not be any spinup commands on IDE drives. If you just attach the
power connector (no data conn) it will power up immediately if the power
supply is on.

I hate to shoot down your project, because it would definately be a
learning experience, but you should also face the reality that your drive
may stop functioning before you even finish building your device :/

PS: I'll pass on this thread to you, you may find it helpful when shopping for
a new hard drive:
http://forums.anandtech.com/messageview.cfm?catid=40&threadid=1160291

Best of luck,
~Jeremy
____________________________________________________________________
Please remove your windows partition when replying by email
 
I've got an old 30GB drive lying around that's on it's last legs and is
making funny noises.

The thing is, I'd rather like to make efficient use of its remaining
operating time. Is there any way to stop a disk spinning up on boot, and
activate it later if you feel like it?

I am assuming you would need some kind of hardware device that mediates
between the IDE bus and power cable, and the disk itself. Is there such a
thing?

Ewan
 
Ewan said:
I've got an old 30GB drive lying around that's on it's last legs and
is making funny noises.

The thing is, I'd rather like to make efficient use of its remaining
operating time. Is there any way to stop a disk spinning up on boot,
and activate it later if you feel like it?

I am assuming you would need some kind of hardware device that
mediates between the IDE bus and power cable, and the disk itself. Is
there such a thing?

Ewan

You would be trying to power up the device way after the hardware
detection phase of Windows bootup sequence and way past when any driver
for it would load and detect the device.

If you manage to find such a device which would have to include hardware
and software (to force the device load), be sure it costs under $50
since you can get a new 40 GB for that.
 
You would be trying to power up the device way after the hardware
detection phase of Windows bootup sequence and way past when any driver
for it would load and detect the device.

How easy would it be to fake the presence of the drive? How much information
does windows need?

If I can record the drive information (either by putting the requisite
signals into the IDE port on the drive and recording the results, or just by
getting the relevant infofrom the manufacturer), would it be that hard to
build a PIC based device that pretends it is the disk? From there, it's just
a matter of swapping the connectors and issuing a spinup command I assume.

Does anyone know of a good place to read up on what goes on with detection
and control of IDE devices?

Ewan
 
Ewan Sinclair said:
I've got an old 30GB drive lying around that's on it's last legs and is
making funny noises.

The thing is, I'd rather like to make efficient use of its remaining
operating time. Is there any way to stop a disk spinning up on boot, and
activate it later if you feel like it?

I am assuming you would need some kind of hardware device that mediates
between the IDE bus and power cable, and the disk itself. Is there such a
thing?

Ewan

once the drive gets noisy (clunking sounds) there will be little time left
....it would really be a waste of time to attempt to use it

otoh: if it was simply a matter of a lot of bad sectors...
it could be lo level formatted and used as a *spare*
 
Wow, this man is on a mission! While most people would just grab one of
those
WD800JB 80GB 8MB cache drives for $49 after rebates, he
wants to design new hardware using a microcontroller.

It may be a tad OTT, but it would be rather interesting. Besides, this is a
unique opportunity to stop myself from stuffing my disk with files. If I
know it has a limited lifetime, it stays fully reserved as scratch space
only :-)

Besides which, I'm sure I could think of all sorts of fun things to do with
such a device later.
I do not believe the IDE controller, OS drivers, or drive itself was
designed to be hot swapped. Even with your proposed "emulator" in place
at boot, I have a feeling your physical swapping of devices later is not a
good idea.

You know, all of this just reminded me of those IDE hot swap bays. Why
didn't I think of that before? Does anyone know if they are any good for
this task, or are they meant to have a drive in at boot that can be replaced
with another of the same model during operation?
There may not be any spinup commands on IDE drives. If you just attach the
power connector (no data conn) it will power up immediately if the power
supply is on.

Well, putting power into it seems to work as a spinup command then :-)
There must be some sort of spinup command, because you can set timeouts in
the BIOS. Maybe there's just a spindown or set timeout command. One of the
three, anyway. If there's a spindown command, maybe it can be used to
suppress initial spinup? It's probably part of an unavoidable self-check
sequence though, as you point out.
I hate to shoot down your project, because it would definately be a
learning experience, but you should also face the reality that your drive
may stop functioning before you even finish building your device :/

Nope, it's made those noises for a while, and they aren't very loud, just a
bit dodgy. The drive is currently in my drawer anyway, so it won't be dying
anytime soon.
PS: I'll pass on this thread to you, you may find it helpful when shopping for
a new hard drive:
http://forums.anandtech.com/messageview.cfm?catid=40&threadid=1160291

Thanks for the link anyway, I'll probably end up doing that, but the problem
will still remain that the available reliable storage size (X), my disk
usage (Y), and any size between 0 and 500MB have relationship X=Y-A

I'm far more willing to learn advanced microelectronics than I am self
control.

Ewan
 
Your trying to save a penny by spending a pound...

If a hard drive is reliable, just use it. If it's not, tag it defective and
toss it out or pile it in the corner where it might be a spare part or
interesting paperweight.

If you can't afford to replace the drive, you definately can't afford to do
what you're proposing.
 
I've got an old 30GB drive lying around that's on it's last legs and is
making funny noises.

The thing is, I'd rather like to make efficient use of its remaining
operating time. Is there any way to stop a disk spinning up on boot, and
activate it later if you feel like it?

I am assuming you would need some kind of hardware device that mediates
between the IDE bus and power cable, and the disk itself. Is there such a
thing?

Ewan

The most feasible alternative towards your intentions would be an
external USB or firewire drive enclosure, which you only turn on when
needed.

On the other hand I agree with the other posters who feel that this
drive isn't worth the time or data risk at this point. A used 30GB
drive that still works perfectly is only worth $20 at best, maybe only
$10.


Dave
 
I've got an old 30GB drive lying around that's on it's last legs and is
making funny noises.

The thing is, I'd rather like to make efficient use of its remaining
operating time. Is there any way to stop a disk spinning up on boot, and
activate it later if you feel like it?

I am assuming you would need some kind of hardware device that mediates
between the IDE bus and power cable, and the disk itself. Is there such a
thing?

Ewan

Put it in an external USB (2.0) enclosure, and only power it up when you
need it. It is hot swappable that way. Of course you will have nothing on
it you don't mind losing.

JT
 
Put it in an external USB (2.0) enclosure, and only power it up when you
need it. It is hot swappable that way. Of course you will have nothing on
it you don't mind losing.

If I was to use an external enclosure, how would USB 2.0 or firewire
performance compare to a drive attached via the IDE bus? I have an IEEE 1394
header on my mobo, is it worth buying the ports to plug into it, rather than
use the USB 2.0 ports I already have?

Does anyone know if there are any issues with firewire on Asus A7V333
motherboards?

Ewan
 
If I was to use an external enclosure, how would USB 2.0 or firewire
performance compare to a drive attached via the IDE bus? I have an IEEE 1394
header on my mobo, is it worth buying the ports to plug into it, rather than
use the USB 2.0 ports I already have?

Does anyone know if there are any issues with firewire on Asus A7V333
motherboards?

Ewan

USB2.0 is similar in speed to Firewire. Both are probably faster than your
failing hard drive. It would probably be better as errors on the USB drive
would not hang your internal hard drives waiting for retries or error
recovery. How much money do you want to sink into this project? If you have
working USB2.0, use it as it will be plenty fast enough. If you want to add
firewire connectors for other uses, or to "play" with it. Its your system
and money. Just remember, hard drives are getting cheap. An external
USB/Firewire enclosure is handy, even after your drive fails, so can be a
good investment.

JT

JT
 
USB2.0 is similar in speed to Firewire. Both are probably faster than your
failing hard drive.

Are you sure they are quite equivalent? I had heard that Firewire can be up
to 77% faster than USB 2.0 in real world tests. I know they are
theoretically about the same, but that doesn't really count for much :-)
How much money do you want to sink into this project? If you have
working USB2.0, use it as it will be plenty fast enough. If you want to add
firewire connectors for other uses, or to "play" with it. Its your system
and money. Just remember, hard drives are getting cheap. An external
USB/Firewire enclosure is handy, even after your drive fails, so can be a
good investment.

That's what I figured. I just want to make sure I get an enclosure that will
allow any drive I put in to perform at it's best. If I put an ATA 100 drive
in a firewire enclosure, would it perform worse than if on the IDE bus?

Also, does anyone know of any enclosures that can handle USB *and* firewire?
Alternatively, anything that adapts firewire to USB?

Ewan
 
I wonder why you care about saving the ~$54 cost of a new drive by
retaining the old failing hard drive rather than just replace it (after
copying the old drive). Unless you have OTHER requirements for an
externally attached hard drive, you're wasting money.

USB/Firewire enclosure: $41
(http://www.mwave.com/mwave/doc2/660628.html)
(http://www.addonics.com/products/external_hdd/combo_external.asp)
Don't know if the prices includes the cables (Addonics price is $75 with
the USB adapter cable and $89 with the Firewire adapter cable).

Maxtor 40 GB 7200 RPM hard drive: $54
(http://www.mwave.com/mwave/SpecA/AA15192.html)
WD 40 GB 7200 RPM hard drive: $53
(http://www.mwave.com/mwave/speca/AA11495.html)
Seagate Barracuda 40 GB 7200 RPM hard drive: $56
(http://www.mwave.com/mwave/SpecA/AA21270.html)

So you could spend $41 to get the extenal enclosure and stick in your
flaky drive (also presuming you consider your own time for hardware fab,
setup, and configuration of no value). But eventually you'll be
replacing the drive anyway, so the real cost will be for both the
enclosure and the eventual replacement drive for a total of $94, or
more. Seems like just getting the ~$54 replacement drive makes a lot
more sense - unless you REALLY have a need for an external hard drive,
like portability between hosts which are NOT networked or for offline
storage of backups (but 40 GB won't going to give you much backup
space).

Your old 30 GB hard drive probably only supports ATA-33, spins at only
5400 RPM, and sounds like it will need to get replaced soon, anyway.
 
If I was to use an external enclosure, how would USB 2.0 or firewire
performance compare to a drive attached via the IDE bus? I have an IEEE 1394
header on my mobo, is it worth buying the ports to plug into it, rather than
use the USB 2.0 ports I already have?

Does anyone know if there are any issues with firewire on Asus A7V333
motherboards?

Ewan

I take it this is the last hard drive on earth?

If you already had the enclosure, the cables, ports, and IF USB2 or
firewire were as fast as pure IDE (which they aren't) the advice would
still be the same, the drive is now unreliable and should not be used.

If you want an external enclosure, buy a new drive for it.


Dave
 
Back
Top