IDE HD not always detected

[Natéag]

I have two SATA drives and
one (non system) IDE HD
on A7N8X-e Asus mainboard.

Quite often, the disk is no longer detected
at startup. Rebooting corrects the problem,
which I encounter mostly when the computer
comes out of WinXP hibernation. Lots of
clicking and slow XP loading.

I then have to reboot so that all my software
and stored data can be accessed.

Could it be some CPU problem. I have
a 380W Antec CPU - that I have used for months
without problems. Some voltages are a bit low,
but that has been true for as long as I can remember.

CPU 1.65 V

+3.3 V 3.25 V
+5 V 4.52 V
+12 V 11.90 V
-12 V -11.83 V
-5 V -4.48 V


I have reset the BIOS to no avail.

 

[anc]

I have two SATA drives and
one (non system) IDE HD
on A7N8X-e Asus mainboard.

Quite often, the disk is no longer detected
at startup. Rebooting corrects the problem,
which I encounter mostly when the computer
comes out of WinXP hibernation. Lots of
clicking and slow XP loading.

What happens if you shut down instead of hibernating,
does this cure the problem ?

 

[JBM]

I have two SATA drives and
one (non system) IDE HD
on A7N8X-e Asus mainboard.

Quite often, the disk is no longer detected
at startup. Rebooting corrects the problem,
which I encounter mostly when the computer
comes out of WinXP hibernation. Lots of
clicking and slow XP loading.

I then have to reboot so that all my software
and stored data can be accessed.

Could it be some CPU problem. I have
a 380W Antec CPU - that I have used for months
without problems. Some voltages are a bit low,
but that has been true for as long as I can remember.

CPU 1.65 V

+3.3 V 3.25 V
+5 V 4.52 V
+12 V 11.90 V
-12 V -11.83 V
-5 V -4.48 V


I have reset the BIOS to no avail.

The 5 volt line seems low.
Since there is a greater load on the power supply when
drives are first starting the voltages may go lower than
what you show, especially if the supply is going bad.
Have you tried reducing the load on the supply by
disconnecting you CDrom drive/drives or any PCI
cards you don't need to boot. Then do what ever you
do to recreate the problem to see if that helped.

Jim M

 

[Ben Pope]

I have two SATA drives and
one (non system) IDE HD
on A7N8X-e Asus mainboard.

Quite often, the disk is no longer detected
at startup. Rebooting corrects the problem,
which I encounter mostly when the computer
comes out of WinXP hibernation. Lots of
clicking and slow XP loading.

I then have to reboot so that all my software
and stored data can be accessed.

Could it be some CPU problem. I have
a 380W Antec CPU - that I have used for months
without problems. Some voltages are a bit low,
but that has been true for as long as I can remember.

CPU 1.65 V

+3.3 V 3.25 V
+5 V 4.52 V

Thats 10% down. ±5% is spec, so that could easily be your problem.

+12 V 11.90 V
-12 V -11.83 V
-5 V -4.48 V

Fine. -ve voltages aren't used.

I have reset the BIOS to no avail.

Given your sagging 5V line(s) and the fact that it won't boot from a cold
start, but will from a warm (ish) start, your PSU is easily the number 1
target.

Unless of course you've been playing with the cabling or jumpers on the
drive in question.

Ben

 

[Natéag]

I have tried a round cable, that is when the problem started,
but I have reconnected the flat cable and nothing changed.

I guess it is a voltage problem. Of course the HD may be
defective, in that it would require a higher voltage to start -
if that is possible, of course. Otherwise, I checkdisk
never finds any errors on that drive. I have reformatted
it, but it did not improve anything.

I am going to try a new CPU.

Thanks.

 

[Ben Pope]

I have tried a round cable, that is when the problem started,
but I have reconnected the flat cable and nothing changed.

I guess it is a voltage problem.

Yeah, most likely.

Of course the HD may be
defective, in that it would require a higher voltage to start -
if that is possible, of course.

The HD will require something like 5V ±5%, and however much current. (As
well as 12V, most likely) When it draws that current (which is much higher
at startup than idle) the voltage line is likely so sag even further.

Otherwise, I checkdisk
never finds any errors on that drive. I have reformatted
it, but it did not improve anything.

It was pretty unlikely to. It's not the hard drive thats at fault, it's the
PSU. I would say that with a likelihood of around 5% your drive has a
problem, and you've already told us that your PSU DOES have a problem, so...

I am going to try a new CPU.

Try a new CPU by all means but it won't change anything unless it
significantly reduces the load on the PSU.

Why don't you just put in a PSU that runs inside of spec and see if that
helps?

Ben

 

[Michael-NC]

I have tried a round cable, that is when the problem started,
but I have reconnected the flat cable and nothing changed.

I guess it is a voltage problem. Of course the HD may be
defective, in that it would require a higher voltage to start -
if that is possible, of course. Otherwise, I checkdisk
never finds any errors on that drive. I have reformatted
it, but it did not improve anything.

I am going to try a new CPU.

Thanks.

You're wasting time and money if you think the CPU is at fault.

 

[Natéag]

Yes, it did.

I think I have found the (temporary: ie until I get a new, more powerful,
CPU) solution:
I removed one of my three Kingston 400 DDRs.
I will know in a few days if that was the problem.

But everything worked fine when I disconnected my DVD-Writer, plus
one of my SATA drives. With any one of these, the problem would
reappear. I have now reconnected both, which I need, an removed
one of the DDR units.

Result of any of the above solutions (?) is that my IE1394 appears
in the device manager appears without that yellow ! sign.

 

[Natéag]

I do not think so. The voltages seem very low.

You're wasting time and money if you think the CPU is at fault.

 

[JBM]

A new CPU isn't going to help with your low voltage problem.
As Ben and others said the most likely fix is a new / better power supply.
Each item you remove reduces the load on the supply and
therefore may fix a problem but the underlying cause is
the supply.

Jim M

 

[Michael-NC]

The CPU has nothing to do with the voltages. That is the power supplies job.
The CPU is probably the last thing that will cause problems in any computer,
it's the most reliable component in a computer.

 

[Natéag]

OK. I mistook CPU for Power Supply (PSU)

My error. I do not often use those abbreviations
in english - a foreign language for me.

 

[JBM]

OK. I mistook CPU for Power Supply (PSU)

My error. I do not often use those abbreviations
in english - a foreign language for me.

I understand

 

[Ben Pope]

OK. I mistook CPU for Power Supply (PSU)

My error. I do not often use those abbreviations
in english - a foreign language for me.

Ahh, THAT makes sense.



Ben

 

[Aldo Larrabiata]

You also may have a rather normal voltage (measured with a multimeter) and a
high ripple on the +5 and/or +12V line when the disk is starting (its inrush
current may be as great as 10 Amps) or running and the heads moving.
The only way to solve the issue is to connect an oscilloscope on the power
lines or to use a true RMS AC digital voltmeter with a large bandwidth
(probably more difficult to find than a scope!).

Replacing the PSU will give good clues.


As a first approach it could be interresting to know what voltages you get
with the _old_ PSU in the different configurations (one disk removed, then
the DVD drive removed, and so on).

If you don't have a multimeter let MBM evaluate the voltages. Don't consider
the absolute values but the variations instead, depending on the different
configurations.

 

[Natéag]

I replaced the PSU and now get 4.99 to 5.02 volts
with a new Antec, same type.
So everything is fine.

 

[Paul]

You also may have a rather normal voltage (measured with a multimeter) and a
high ripple on the +5 and/or +12V line when the disk is starting (its inrush
current may be as great as 10 Amps) or running and the heads moving.
The only way to solve the issue is to connect an oscilloscope on the power
lines or to use a true RMS AC digital voltmeter with a large bandwidth
(probably more difficult to find than a scope!).

Replacing the PSU will give good clues.


As a first approach it could be interresting to know what voltages you get
with the _old_ PSU in the different configurations (one disk removed, then
the DVD drive removed, and so on).

If you don't have a multimeter let MBM evaluate the voltages. Don't consider
the absolute values but the variations instead, depending on the different
configurations.

Aldo, if what you are saying is true, then disks would never work
in 30W or 40W external enclosures. Disks are designed to a "power
footprint", so that an infinitely large power supply is not needed.

The motor drive circuit is designed to current limit, and while the
motor is spinning up, somewhere around 2amps will be drawn from +12V.
This settles back to 0.5 amps when the drive is up to speed. All the
drive specs I've ever read have a limit like that. (Of course, not
the old computer room removable disk pack disks that had a motor
running from 220V

For example, I have a detailed spec for an IBM 120GXP in my document
collection, and it lists:

12V@ 1.867amps 5V@ 0.740amps Start up (max)

Startup current can still be an issue, if you have enough disks.
And, that is why the SCSI standard supports staggered spinup, so
if you had a lot of disks, the 2A or so from each one, don't add
together when power is first applied.

Paul

 

[Aldo Larrabiata]

Ooops,
the 9.7 Amps is the total amount of the inrush currents I have with three
IDE disks (Two Maxtor DiamondPlus60 with 2950 mA peak, One Diamond Max plus8
with 1380 mAmps peak and other usual stuff).

I've also 4 SCSI disks but not considered in this budget. Their spinup is
delayed through the SCSI BIOS (more or less 10 Amps as well).

When My PSU failed, I temporarily used a small 300W PSU from another
computer. I noticed what I described in the previous mail.
During intense seeking/writing, the ripple raised up to 0.7 Vpp.

All disks aren't equal, the Maxtors appear to rather greedy.

You also may have a rather normal voltage (measured with a multimeter) and a
high ripple on the +5 and/or +12V line when the disk is starting (its inrush
current may be as great as 10 Amps) or running and the heads moving.

[...]

Aldo, if what you are saying is true, then disks would never work
in 30W or 40W external enclosures. Disks are designed to a "power
footprint", so that an infinitely large power supply is not needed.