Sometimes a hard drive disappears when I boot up 1

[cpjust]

I'm running Vista x64 SP2 with 8GB RAM, 4 hard drives & an SSD (non-RAID). Vista is on the SSD.

About once or twice a week, when I boot up it sits at a black screen for a few minutes and the HDD light on my case stays on instead of flashing; then it finally gets to the login screen. After I login I notice that 1 of my hard drives doesn't show up in Windows Explorer or the Drive Manager. It's not always the same drive the disappears; the cables aren't loose; and I've run the HD Tune disk diagnostics many times and never find anything wrong with any of my drives. I'm guessing it's timing out while waiting for the hard drive to respond and then gives up and drops it from the list of available drives.

This has been happening for several years, a few different motherboards, and many fresh installs of Windows.

Any ideas on why this happens and how to fix it?

 

[goombawaho]

List hardware please: The 4 hard drives - internal hard drives/external? RAID controller or just four separate drives each with a drive letter? Vista is on a single SSD. You haven't given us much to go on.

You can rule out the motherboard if you've had different ones and viruses since you've had multiple installations of windows.

BIOS update available? (but that problem shouldn't spam different motherboards)

It doesn't happen after resuming from standby or hibernation does it?

 

[DrB0b]

Does the BIOS always see the drives?

The way I would look at this is, what is the common denominator between all of your builds? Is it a RAID card(not sure if you mean no RAID on SSD or all), same OS each time installed, same brand/chipset motherboard, same case or cabling....? Are the non-SSD drives old?

Learning - A never ending quest for knowledge usually attained by being thrown in a situation and told to fix it NOW.

 

[rclarke250]

what are the specs of the power supply? And are you implementing a hard drive spin up delay if you have raid enabled?

 

[cpjust]

First motherboard was an ASUS P5K3 Deluxe. At that time I only had 3 x 500GB Seagate Barracudas (later changed to WD) in a RAID-5 array.
Next motherboard was a Gigabyte piece of crap (don't remember the model, but maybe something like GA-G41MT) no RAID.
Current motherboard is Asus (maybe P5G41T-M? I'll check when I get home) no RAID.

CPU: Intel core2 Quad 2.4GHz Q6600 (underclocked)
RAM: 2x4GB Kensington DDR3 1333 (underclocked to 1066) -- used to be: 2x2GB OCZ Gold DDR3 1333

[no RAID on any of these. All internal SATA]
SSD: OCZ Agility 90GB
2x WD Black 500GB
1 WD Black 1TB
1 Seagate Barracuda 500GB

Ultra Chilltec CPU cooler.
Antec One Hundred case

I don't remember the power supply, but I'll also check that when I get home.
The hard drives vary between 2-5 years old, but this has been happening from day one.
I don't use sleep mode anymore because it was crashing every so often (probably because of this same problem), also I don't want to do a lot of writing to my SSD whenever it goes into sleep mode.

 

[cpjust]

OK, the power supply is an Antec EA-650 (650W).
The motherboard is an Asus P5G41T-M LE.
I also have a Syba (I/O crest) SATA III PCI-e controller card for the SSD drive.

 

[rclarke250]

That unit has 3 12v rails, how are these distributed to the drives, and video card if you have one?

 

[cpjust]

I use the video, sound & NIC on the motherboard.
1 power cable goes to the SSD & the 1TB drive.
1 power cable goes to the other 3 500GB drives.
I also have an Asus DVD-RW drive.
1 power cable goes to the DVD drive, the Ultra Chilltec & 2 fans in the case.

 

[DrB0b]

Have you tried a new PSU or separating the 3 SATA drives onto two separate molex rails in case the one is having an issue powering all three devices?

Learning - A never ending quest for knowledge usually attained by being thrown in a situation and told to fix it NOW.

 

[rclarke250]

Sounds like one of the drives maybe putting noise on the bus, and it knocks a drive off the controller. Have you tried disconnecting the 3 500 GB drives, and running for a few days with just the ssd, and 1 tb drive? then add 1 drive at a time back in till you get all the drives connected? Have you looked for firmware updates on the drives? And last, do you use the windows drivers,system board manufacturer drivers, or the drivers from each of the manufacturers of the individual components (Like realtek drivers for lan,and audio, Intel drivers for chipset, and matrix drivers for the hdds)? Because you should be using the last, the manufacturers drivers, unless they cause an issue, then the system board supplied drivers from the website, last choice would be windows supplied drivers.

 

[goombawaho]

^^^ Good advice above - elimination by subtraction then addition + updates all around.

 

[cpjust]

My first PSU started smoking, so I replaced it a few years ago.
I can check which drivers I'm using, and if updating them doesn't work, I'll try disconnecting drives and adding them back one at a time.

 

[cpjust]

I should probably also check my Event Viewer more often. I'm seeing these errors quite a bit:

Source : Event ID : Error message

atapi : 11 : The driver detected a controller error on \Device\Ide\IdePort2.
disk : 15 : The device, \Device\Harddisk0\DR0, is not ready for access yet.
disk : 51 : An error was detected on device \Device\Harddisk0\DR0 during a paging operation.
volmgr : 57 : The system failed to flush data to the transaction log. Corruption may occur.
Ntfs : 137 : The default transaction resource manager on volume H: encountered a non-retryable error and could not start. The data contains the error code.
Ntfs : 141 : {Delayed Write Failed} Windows was unable to save all the data for the file H:\; the data has been lost. This error may be caused if the device has been removed or the media is write-protected.

 

[goombawaho]

Given those errors, I would disconnect all but one hard drive at a time (excluding the SSD) and run the manufacturer's LONG test on each one. Not a generic test or third party test - the one from the manufacturer.

May be outdated, but will get you pointed.

http://www.tacktech.com/display.cfm?ttid=287

Could also be the motherboard/sata port that it's plugged into. Maybe update drivers first to see if it goes away.

 

[cpjust]

rclarke250 I think you may be right about the power supply cables.
I disconnected 1 of the 500GB drives and my system ran perfectly for a week with no errors in the Event Log.
Now I reconnected it and disconnected a different 500GB drive and I'll try that for a week and see how that goes...
If it turns out that my PSU just can't handle 3 drives on a single power cable, I guess I can get a Molex to SATA power cable adapter and connect one of my 500GB drives to one of the Molex cables.

 

[rclarke250]

Thanks for the update.

 

[cpjust]

New update, that configuration eventually failed too.
I then connected everything back the way it was, but I moved the hard drive that was usually failing from the bottom slot up one slot in my case so it would be further away from the big bundle of power cables coming out of my PSU. It's been running fine for over a week now.
So my new theory is that the power cables were producing some EM interference when they were right against the drive.