Is it safe to remove power from a SATA hard drive leaving data connectedto disable it?

The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

mike wrote:

The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

https://electronics.stackexchange.co...ial-interfaces

Mobo Disk - loss of power

RX+ --+--------------------------||--- TX+
| 10nF GND \
100ohm TX+ \
| 10nF TX- \
RX- --+--------------------------||--- TX- GND \__ SATA
RX+ / 7pin
TX+ ---||--------------------------+-- RX+ RX- /
10nF | GND /
100ohm
10nF |
TX- ---||--------------------------+-- RX-

On a loss of power on the Disk end, the AC coupling
provides DC isolation, removing the failsafe requirement
for the pad design.

The termination at the receiver, prevents reflections back
to the transmitter, and should match the characteristic impedance
of the red cable.

https://datasheets.maximintegrated.c...s/MAX4951C.pdf

"Input/Output Terminations

Inputs and outputs are internally 50ohm terminated and
must be AC-coupled to the SATA controller IC and
SATA device for proper operation."

"Ccl=10nF"

*******

The power connector is the one that has "optional" functions.

For example, you could tie a LED for activity blinks, to pin 11.
But, nobody did.

Pin 3 is "PWDIS", a function perfect for your application,
but could only be relied upon, if the drive was compliant
with SATA version 3.3 spec. And how would you figure that out ?
Spec sheets with that level of detail, are hard to come by.
Manufacturers won't give us a platter count, or admit
publicly when a drive is shingled. What hope would we have
of finding "PWDIS: Yes".

https://en.wikipedia.org/wiki/Serial_ATA

HTH,
Paul

On Sun, 06 May 2018 04:59:34 -0700, mike wrote:

The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

You may try a multi-OS boot arbitrator...
http://btmgr.sourceforge.net/about.html
You can boot, enter the arbitrator configuration, to hide drives;-
FWIW, if judging from MS/W10 and a free handjob, perhaps not much.

Easy .and. Good. There's probably a saying about that. The reason
I'd rather not, rather never, ever have to open a case ever again, is
simply because I don't and won't: I quit putting on the sides of
cases ages ago;- sometimes I also take metal shears to drive bay case
cages: works wonders for getting out drives, as I'm not especially
prone to shredding my hands from resulting jagged edges.

Hm, can't recall a conditional array for "grayed-out" drives, Windows
displays, either for a BIOS or SmartBoot Manager "hidden-drive"
configuration. I personally wouldn't count on a disabled safeguard
condition, not to a hardened hacker, at least until I know, if ever,
better.

I would expect, some likelihood at least, an E-SATA dock is going to
provide more protection: a failsafe condition would be nice, yes
indeed, were plugs haphazardly yanked. It's not, though, that you're
then replicating exactly a brown/black-out condition, which drives of
course, with more or less risk of data corruption, otherwise have
reasonably well to convey and withstand.

Grab and Yank . . . me, no way. Not to underestimate my impatience
factor, nor an aspect I haven't as well considered. Then again I'm
not explicitly on regular Windows 10 update schedules, however that
works;- it's also part of the Windows install regime since Windows 7:
that MSFT will actively seek out to annihilate a boot methodology
other than their own.

I think you'll find whatever you want to find here...
https://superuser.com/questions/9258...g-sata-hotplug

On 5/6/2018 9:00 AM, Paul wrote:
mike wrote:

The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

https://electronics.stackexchange.co...ial-interfaces


Mobo Disk - loss of power

RX+ --+--------------------------||--- TX+
| 10nF GND \
100ohm TX+ \
| 10nF TX- \
RX- --+--------------------------||--- TX- GND \__ SATA
RX+ / 7pin
TX+ ---||--------------------------+-- RX+ RX- /
10nF | GND /
100ohm
10nF |
TX- ---||--------------------------+-- RX-

On a loss of power on the Disk end, the AC coupling
provides DC isolation, removing the failsafe requirement
for the pad design.

The termination at the receiver, prevents reflections back
to the transmitter, and should match the characteristic impedance
of the red cable.

https://datasheets.maximintegrated.c...s/MAX4951C.pdf

"Input/Output Terminations

Inputs and outputs are internally 50ohm terminated and
must be AC-coupled to the SATA controller IC and
SATA device for proper operation."

"Ccl=10nF"

*******

The power connector is the one that has "optional" functions.

For example, you could tie a LED for activity blinks, to pin 11.
But, nobody did.

Pin 3 is "PWDIS", a function perfect for your application,
but could only be relied upon, if the drive was compliant
with SATA version 3.3 spec. And how would you figure that out ?
Spec sheets with that level of detail, are hard to come by.
Manufacturers won't give us a platter count, or admit
publicly when a drive is shingled. What hope would we have
of finding "PWDIS: Yes".

The WD drives have a jumper for power on in standby,
but that doesn't keep the drive from booting.

https://en.wikipedia.org/wiki/Serial_ATA

HTH,
Paul

Thanks for the inputs. Most of the links were vague, but
I did find what appeared to be an authoritative confirmation.

Silly me, I should just have looked at the drive.
The series caps are right there at the connector.
Confirmed by a DVM, there doesn't seem to be anything
there that could hurt the motherboard.

I'm good to go.

I'm curious. You seem to be everywhere providing detailed,
trustworthy, information to many people on a wide range of
topics. How do you find the time?

mike wrote:
On 5/6/2018 9:00 AM, Paul wrote:
mike wrote:

The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

https://electronics.stackexchange.co...ial-interfaces



Mobo Disk - loss of power

RX+ --+--------------------------||--- TX+
| 10nF GND \
100ohm TX+ \
| 10nF TX- \
RX- --+--------------------------||--- TX- GND \__
SATA
RX+ /
7pin
TX+ ---||--------------------------+-- RX+ RX- /
10nF | GND /
100ohm
10nF |
TX- ---||--------------------------+-- RX-

On a loss of power on the Disk end, the AC coupling
provides DC isolation, removing the failsafe requirement
for the pad design.

The termination at the receiver, prevents reflections back
to the transmitter, and should match the characteristic impedance
of the red cable.

https://datasheets.maximintegrated.c...s/MAX4951C.pdf

"Input/Output Terminations

Inputs and outputs are internally 50ohm terminated and
must be AC-coupled to the SATA controller IC and
SATA device for proper operation."

"Ccl=10nF"

*******

The power connector is the one that has "optional" functions.

For example, you could tie a LED for activity blinks, to pin 11.
But, nobody did.

Pin 3 is "PWDIS", a function perfect for your application,
but could only be relied upon, if the drive was compliant
with SATA version 3.3 spec. And how would you figure that out ?
Spec sheets with that level of detail, are hard to come by.
Manufacturers won't give us a platter count, or admit
publicly when a drive is shingled. What hope would we have
of finding "PWDIS: Yes".

The WD drives have a jumper for power on in standby,
but that doesn't keep the drive from booting.

https://en.wikipedia.org/wiki/Serial_ATA

HTH,
Paul

Thanks for the inputs. Most of the links were vague, but
I did find what appeared to be an authoritative confirmation.

Silly me, I should just have looked at the drive.
The series caps are right there at the connector.
Confirmed by a DVM, there doesn't seem to be anything
there that could hurt the motherboard.

I'm good to go.

I'm curious. You seem to be everywhere providing detailed,
trustworthy, information to many people on a wide range of
topics. How do you find the time?

When you're retired, anything is possible.

Paul

On Sunday, May 6, 2018 at 5:00:51 AM UTC-7, mike wrote:
The problem I'm trying to solve is that windows 10 upgrade to
one hard drive messes up the boot sector of the other bootable
drive. I've been opening the case and pulling the data cable.

I need a way to EASILY switch off a hard drive so that it is
completely inaccessible.

I've been considering disconnecting +5V and +12V
with a front panel switch
leaving the data cable connected.

I don't intend switching it hot. I'd rather not open
the case and pull the cables or have to reconfigure
the BIOS every time.

Does the BIOS really disable the drive in hardware? Or just
"make a recommendation" to the driver?

Can removing power damage the SATA data interface of the motherboard
or the hard drive?

Or are both ends of the data interface protected against such?

The same question applies to an eSATA drive dock that's not
powered on.

Another problem of Windows 10. This problem arises from Windows 10 combining multi hard drive storage into one storage called "Onedrive cloud". I don't know what good is that sync file activity from hard drive and online file with one drive. Trying to setup Onedrive Cloud is what screwed up your hard drive.