Disk Boot Failure - Suggestions?

> Candidates:
> - The PSU. Maybe it needs to long to assert its voltages.
> Since PSUs loose power over time (their capacitors loose
> capacity, making them overall less efficient at peak load),
> that could be it. It is also possible that your PSU is too
> weak. What PSU do you have and what graphics card(s)?

I am starting to suspect you are right here, based on comments I heard
elsewhere as well. I never thought about the fact that I keep tossing
more and more hardware in the system.

The power supply I am using is COOLMAX CX-400B ATX v2.01 400W. And on
this, I am powering:

1. The SOLTEK MB
2. The Athlon 2700XP
3. CD/RW drive
4. DVD drive
5. SATA HD
6. IDE HD
7. IDE HD
8. MSI 6600GT video (!)
9. M-Audio DELTA 1010LT (!)
 >> Stay informed about: Disk Boot Failure - Suggestions? 

Last night I did some trouble shooting. I discovered that on power up,
the machine *never* boots. And furthermore, whereas previously, the
reset button seemed to always work, that "fix" decreased its success
rate to less than 50%. I tried disconnecting EVERYTHING from the power
supply except the SATA (i.e. disconnect both IDE drives AND both CD
drives) and the problem remained unchanged - this makes me less
suspicious of the power supply. Agree?

I then tried booting from my IDE drive (which is also bootable), and
when I do that, it boots up just fine - BUT the SATA drive is not
present in Windows.

Further, I should note that during the boot (either on an SATA boot
failure, or when I boot off IDE and the SATA isn't visible), I do NOT
see a message on the boot screen showing the SATA controller coming up.
On the occasions where the SATA drive DID boot, I would see text on
the screen showing the RAID configuration during bootup. This makes me
suspect the SATA controller on the motherboard. Anyone?

Finally, last night, I tried messing around with the SATA cable between
the hard drive and motherboard. I removed it. Dusted it. Reseated it
and did a little wiggling. When I did this, all of a sudden the SATA
drive booted 100% even from complete power down. I tried that 4 or 5
times and had 100% success. So I thought, problem solved.

Sadly, this morning, problem back. Now the SATA won't boot at all,
under any circumstances.

Are you guys thinking controller, drive, or still possibly power supply?

One plan I have is to go to a friend's system, and try my drive on his
MB, and then his drive on my MB and try to answer the question that way.
He also has a spare power supply with separate 12V rails, so we may
try that too.

Any other thoughts? (other than that I am going to have an annoying
time dealing with this!)

Thanks.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

RSS <rsilverst RemoveThis @yahoo.com> wrote

> Last night I did some trouble shooting. I discovered that on power up,
> the machine *never* boots. And furthermore, whereas previously, the
> reset button seemed to always work, that "fix" decreased its success rate
> to less than 50%.

That's significant and indicates that the motherboard is likely
going bad or the hard drive subsystem you are booting from is.

> I tried disconnecting EVERYTHING from the power supply except the SATA
> (i.e. disconnect both IDE drives AND both CD drives) and the problem
> remained unchanged - this makes me less suspicious of the power supply.
> Agree?

Yes.

> I then tried booting from my IDE drive (which is also bootable), and when
> I do that, it boots up just fine

Always, completely reliably, from power up ?

If that is so, its likely the drive you normally boot off is
going bad or its controller is going bad or the connection.

> - BUT the SATA drive is not present in Windows.

Likely just because that boot doesnt load the driver for the sata
controller.

> Further, I should note that during the boot (either on an SATA boot
> failure, or when I boot off IDE and the SATA isn't visible), I do NOT
> see a message on the boot screen showing the SATA controller coming
> up. On the occasions where the SATA drive DID boot, I would see text
> on the screen showing the RAID configuration during bootup. This makes
> me suspect the SATA controller on the motherboard. Anyone?

Yes, that looks likely.

> Finally, last night, I tried messing around with the SATA cable
> between the hard drive and motherboard. I removed it. Dusted it.
> Reseated it and did a little wiggling. When I did this, all of a
> sudden the SATA drive booted 100% even from complete power down. I
> tried that 4 or 5 times and had 100% success. So I thought, problem
> solved.

> Sadly, this morning, problem back. Now the SATA won't boot at all,
> under any circumstances.

> Are you guys thinking controller, drive,

More likely a physical problem with the sata cable
or one of the sata connectors used for the boot drive.

Try another sata cable in case that one has a
cracked wire or poor connection to one of the pins.

If a new sata cable doesnt fix the problem you most likely
have a dry joint on one of the sata connectors where its
soldered to the pcb at either end, motherboard or drive.

The obvious test there is to try a different sata drive.

> or still possibly power supply?

Very unlikely now.

> One plan I have is to go to a friend's system, and try my drive on his
> MB,

Yes, tho just moving the drive may be enough to disturb the
bad connection enough to see it working again for a while.

> and then his drive on my MB

Thats not a great approach, because the drivers wont
necessarily be appropriate for your MB unless its identical.

> and try to answer the question that way.

> He also has a spare power supply with separate 12V rails, so we may try
> that too.

I wouldnt bother with that at this stage.

> Any other thoughts?

Try another sata cable first.

> (other than that I am going to have an annoying time dealing with this!)

Dunno, its easier to diagnose than many.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

>
> Always, completely reliably, from power up ?
>

the IDE drive has been 100% reliable (as of this point) from startup.
also, when the SATA drive *is* running, there has been no problem with
it and no funny stuff of any sort. in fact, last night while it was
running, i was able to do a sync between it and my IDE drive to back up
all my data in case i couldn't get back in again.

it just seems unlikely to me that it's the drive... i guess the most
conclusive evidence that it's not the drive would be if i put in my
friend's SATA drive in my machine, and it doesn't recognize it at all.
but other than that, all other swapping tests would probably only be
"suggestive".
 >> Stay informed about: Disk Boot Failure - Suggestions? 

RSS <rsilverst.RemoveThis@yahoo.com> wrote

>> Always, completely reliably, from power up ?

> the IDE drive has been 100% reliable (as of this point) from startup.

Fine.

> also, when the SATA drive *is* running, there has been no problem with it
> and no funny stuff of any sort. in fact, last night while it was
> running, i was able to do a sync between it and my IDE drive to back up
> all my data in case i couldn't get back in again.

Thats interesting. That might indicate a problem with the cable
or the connection to it in the sense that once the system is
booted and gets to retry on CRC errors on the cable etc,
that doing that always succeeds, but that that can get in the
way of detecting the drive at boot time when more primitive
approaches are used over the cable in the sense of retrys.

It would also be worth trying a different power connector
for the sata drive too. Its possible that if it cant start in
time, the drive goes missing forever, shuts down. But
if it does get started, it runs fine after that when used etc.

> it just seems unlikely to me that it's the drive...

Dunno, everything points to a problem with the sata drive subsystem,
particularly the fact that its seen intermittently at boot time and the
IDE drive is always seen completely reliably at boot time.

> i guess the most conclusive evidence that it's not the drive would be if
> i put in my friend's SATA drive in my machine, and it doesn't recognize
> it at all.

Yes, that would be a useful test, but you'll screw his config on that
drive if you boot from it in your system if the MBs arent identical etc.

> but other than that, all other swapping tests would probably only be
> "suggestive".

Really depends on the result of the test.

If swapping the sata cable sees it boot reliably for weeks,
and using the original again sees a return of the problem,
that would prove that its a dud cable pretty conclusively.

Its unlikey that the result you got when fiddling
with the cable is a coincidence and not relevant.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

Rod- First, thanks for the detailed responses. Much appreciated.
>
> Yes, that would be a useful test, but you'll screw his config on that
> drive if you boot from it in your system if the MBs arent identical etc.

Understood. I was thinking we won't boot off his drive - I thought I'd
put his SATA drive on my controller, but boot off my IDE drive and see
if Windows identifies the SATA drive in the system. Seems that should
be a valid test because right now when I boot off my IDE drive I do
*not* see my SATA drive. That would not rule out the possibility that
something's wrong on the boot sector of my drive though...

My only options at this point, other than buying new hardware and hoping
for the best, are to do the power cable, the SATA cable, and the drive
swap test.

Oh, there *is* one more option that a friend suggested to me. They said
I should buy an inexpensive PCI SATA controller card and see if *that*
solves the problem. Because if it does, maybe the simplest solution
would be to just disable the SATA controller on the motherboard, use the
PCI card, and live with it. If none of the aforementioned experiments
yield any results, I may try this. Thoughts on that one?

Thanks again.

Bob
 >> Stay informed about: Disk Boot Failure - Suggestions? 

RSS <rsilverst DeleteThis @yahoo.com> wrote

> Rod- First, thanks for the detailed responses. Much appreciated.

No problem, thats what these technical groups are all about.

>> Yes, that would be a useful test, but you'll screw his config on that
>> drive if you boot from it in your system if the MBs arent identical etc.

> Understood. I was thinking we won't boot off his drive - I thought I'd
> put his SATA drive on my controller, but boot off my IDE drive and see if
> Windows identifies the SATA drive in the system. Seems that should be a
> valid test because right now when I boot off my IDE drive I do *not* see
> my SATA drive.

Dunno, since you dont ever see the SATA drive in that config, it
looks more like the install on the IDE drive doesnt actually have
the SATA drivers loaded. You'd expect to see the original sata
drive sometimes when its sometimes visible when you boot off it.

> That would not rule out the possibility that something's wrong on the
> boot sector of my drive though...

You shouldnt get an intermittent result with the boot sector and a
bad boot sector should show up in the SMART data for the drive.

> My only options at this point, other than buying new hardware and hoping
> for the best, are to do the power cable, the SATA cable, and the drive
> swap test.

Yes, and I'd certainly try the first two when booting from the sata drive.

> Oh, there *is* one more option that a friend suggested to me. They said
> I should buy an inexpensive PCI SATA controller card and see if *that*
> solves the problem. Because if it does, maybe the simplest solution
> would be to just disable the SATA controller on the motherboard, use the
> PCI card, and live with it.

Its unlikely that the motherboard sata controller is intermittent.

> If none of the aforementioned experiments yield any results, I may try
> this. Thoughts on that one?

I'd check that the sata driver is actually loaded in the
install you boot when booting the IDE drive first myself.
You should be seeing the sata drive intermittently and since
you never see it, its more likely that the driver isnt loaded.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

End of the story:

Pretty sure it was the Serial ATA controller on the motherboard. When I
brought it to a friend's place, we tried swapping in 2 other SATA cables
- made no difference. We tried plugging the drive into the other SATA
header on the motherboard (there were 2), and that made no difference.
Connected my drive to his machine, and it worked perfectly 100% of the
time. Maybe it was a bad connection or a short between one of the
headers and the board. Who knows.

The solution:

I gave my friend my SATA drive. He gave me one of his IDE drives (a
Samsung that was actually quieter and larger than what I had). And we
disabled the SATA controller on the motherboard. End of story.

Although it was on warranty from Newegg, I saw that they don't carry
that motherboard anymore. Who knows what they would have given me
instead, and what nightmares of compatibility and configuration I may
have encountered. And I'd have had to go 3 weeks with no computer. So
I think this was the best solution. And I ended up with a larger drive,
so I can't complain.

Thanks for all the assistance.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

RSS <rsilverst.DeleteThis@yahoo.com> wrote

> End of the story:

> Pretty sure it was the Serial ATA controller on the motherboard. When I
> brought it to a friend's place, we tried swapping in 2 other SATA
> cables - made no difference.

You didnt say clearly what was happening before the swap,
whether the drive was visible or not, presumably it wasnt.

> We tried plugging the drive into the other SATA header on the motherboard
> (there were 2), and that made no difference.

Thats a real worry. The intermittent symptoms indicated a
bad joint in a cable or soldering of the physical connector
to the motherboard and these two have eliminated both of
those, particularly if all those swaps didnt allow the sata
drive to be seen at all.

> Connected my drive to his machine, and it worked perfectly 100% of the
> time.

That doesnt prove a lot with an intermittent fault tho.

> Maybe it was a bad connection

Unlikely since you got the same result with
two different sata headers on the motherboard.

> or a short between one of the headers and the board.

Also not likely. You can get a bad solder joint where
the connector is soldered into the motherboard, but
its unusual to get an intermittent short and you got
the same result with two different sata headers.

> Who knows.

> The solution:

> I gave my friend my SATA drive. He gave me one of his IDE drives (a
> Samsung that was actually quieter and larger than what I had). And we
> disabled the SATA controller on the motherboard. End of story.

> Although it was on warranty from Newegg, I saw that they don't carry that
> motherboard anymore. Who knows what they would have given me instead,
> and what nightmares of compatibility and configuration I may have
> encountered.

And it might not have been easy to convince them it
was faulty if it decided to behave itself when returned.

> And I'd have had to go 3 weeks with no computer.

Yeah, significant consideration.

> So I think this was the best solution. And I ended up with a larger
> drive, so I can't complain.

Yeah, decent result.

> Thanks for all the assistance.

Thanks for the washup, too rare in my opinion.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

Rod Speed wrote:
> RSS <rsilverst DeleteThis @yahoo.com> wrote
>
>> End of the story:
>
> You didnt say clearly what was happening before the swap,
> whether the drive was visible or not, presumably it wasnt.

at that point it wasn't visible at all. the likelihood of successful
boot had decreased to about 1 in 10 or worse. for that reason, it
seemed significant that it came up successfully 100% of the time in my
friend's system (not as the boot drive, obviously), and he mirrored that
drive for me twice (onto both of my IDE drives too), so it was
performing a ton of data movement with no glitches.
 >> Stay informed about: Disk Boot Failure - Suggestions? 

RSS <rsilverst.TakeThisOut@yahoo.com> wrote
> Rod Speed wrote

>> You didnt say clearly what was happening before the swap,
>> whether the drive was visible or not, presumably it wasnt.

> at that point it wasn't visible at all. the likelihood of successful
> boot had decreased to about 1 in 10 or worse.

OK, best situation for that swap test, particularly
when the swap didnt improve that lack of visibility.

> for that reason, it seemed significant that it came up successfully 100%
> of the time in my friend's system (not as the boot drive, obviously), and
> he mirrored that drive for me twice (onto both of my IDE drives too), so
> it was performing a ton of data movement with no glitches.

Yes, tho it isnt absolute proof since moving the
drive could have seen the fault go away for a bit.

The previous tests were more significant for that reason.
 >> Stay informed about: Disk Boot Failure - Suggestions?