Windows keeps "losing" secondary HDDs

[ummduh]

Having a bit of an issue.

First off. System is stock. I haven't OC'd anything yet. AHCI mode. SSD boot drive in SATA0, data drive(s)s in SATA1, DVD drive SATA2.

Gigabyte z68ma-D2H-B3
i5 2500k (stock until this is figured out)
2x2Gb G.Skill 1600
Antec Basiq 430W
No extra GPU, stock Intel is good enough for right now.
Some random USB2 HDD enclosure
Seagate Go Flex "desk" 2TB USB2 (put the destroyed WD20EARS in the USB enclosure and recovered the files to the Seagate .. both worked flawlessly in 6 days' time )
No other peripherals at this time have been installed.

My problem is that I keep losing my "data" drive. No matter what drive I happen to put in the system.

Initially I thought my WD20EARS had just died (again) but it has been flawless in a USB enclosure for the last 140-something hours recovering the 1.4TB of data this caused me to lose. In that time another drive I installed in its place actually has come and gone several times without reboot. It has shown up without any input from me and left without my input. It isn't showing up anymore as the last time it did Windows said it needed to be formatted.

I have run 2 other drives in that time and they ALL do it. They're fine in USB enclosures though.

They always reappear at reboot. They drop out after no certain amount of time. no rhyme or reason. After a few times of it dropping out, it is no longer recognized and Windows says it needs to be formatted. (this is why I have spent so much time recovering the files from the first drive... )

I have tried with and without Intel RST drivers. I updated the BIOS on my board. I've tried all the other SATA ports.

I have not tried not running AHCI mode cuz that'd be stupid.. I NEVER lose my boot drive. I have not had any BSODs.

System is set to "Always On" power mode, hard drive turn off to "never", screen always on, etc.. S5 is enabled in the BIOS but the computer hasn't slept in the last 6 days due to recovering the first drive since then.

There seem to be a lot of Event logs that seem to point to the drives going bad, but I know the drives are good. They all work flawlessly in USB enclosures FWIW.

any help? The mobo/cpu are new, swapped from a h67 board and i3-2100 to a z68 board and i5 2500k.
I attempted to open a ticket with Gigabyte, but it's been 3 days and I haven't heard back from them yet. Kinda p/o'd about that.

 

[rraod]

As you have tried all the other SATA ports, I am suspecting the SATA cable connecting the Data drive to the motherboard. I had similar problems with IDE ribbon cables, and I had to always open the case, unplug and replug the cables to get back the drives.

There could be a loose connection to the HDD either in data cable or power cable.

 

[ummduh]

I will check that tomorrow.

BTW, it appears possible that one of my drives actually did die. I finally got to reboot and it didn't show up in BIOS this time. I will to swap to a third I had been using and go from there.

Started to get sick of HD issues. 2 SSDs failed in the last 3 months, my WD20EARS failed twice in the last year, now this.. arrrgh.

 

[TVeblen]

I'm thinking power supply.
Do you have a multimeter you can test with? You could also take a look at your voltages in the BIOS settings, or a program like Speedfan.

That rate of hard drive failure is excessive, and knowing that 'dirty power' can kill drives I'm wondering if that could be the case here. Dirty power is far worse than a failed power supply. Because the power level fluctuates, it can reap havoc with the drive's controller. It would also explain why the drives cut out.

Check the data cable as rraod suggests, that is a common issue, but then consider the issue with the power supply. That 430 watt is relatively small for a system like yours. It will work brand new and at peak performance, but there is very little "overhead" to work with (excess capacity). Any minor defect, including normal capacitor aging, could push it over the edge.

You could also try a different (spare) power connector. While quite rare, sometime the connector just does not make good contact (loose wire, etc).

 

[ummduh]

Sorry haven't updated anything. Haven't messed with this for a couple days. I did take the HDD out that I thought was dead (wasn't even showing up in BIOS anymore) and put it in a USB enclosure and it works fine. I ran WD Data Lifeguard short test on it without issue. I have since reinstalled it and am running the extended test to see what will happen.

After this if it passes I will put it back to SATA1 (i put it to SATA5 right now) and restest. I also have it on a new SATA cable.

BTW.. 430W not enough for a basic system? Really? What would you suggest? 1KW for a system that won't even peak at 150W? 430W is easily 70% more than needed.

The HDD failures up to this point have been in 2 different systems. But all at my 1 home. My first WD20EARS died within a couple months. WD sent a refurb under warranty. That one died in a week. I called and whined, tried to get a different line. They could only offer 1TB drives in the black line.. I declined. They sent a new WD20EARS instead of another refurb. Thats what I am using now, and what I lost all the data from.
I also have a second system (kids comp) running an OCZ Agility 60GB SSD. Their monitor (probably 10 years old.. ) died and I haven't replaced it yet so their comp sat around for a month or so. I fired it back up for something or other and the drive was missing for no apparent reason.
Then on my main comp, I have an Agility 2 ssd. It would only work for about 2 minutes after power on and then it would die. OCZ RMA'd that one already. I am currently using that replacement as the boot on my main comp right now. Still have to ship the Agility 1 back to them.

 

[ignatzatsonic]

BTW.. 430W not enough for a basic system? Really? What would you suggest? 1KW for a system that won't even peak at 150W? 430W is easily 70% more than needed.

ummduh:

I agree with you. I seriously doubt you are overloading your PSU. I have a very similar system to yours (see my system specs). By actual measurement, it never uses even 200 watts. But a malfunctioning PSU is always possible.

 

[TVeblen]

I didn't mean to say that you are overloading the power supply.
I mean that if that power supply is defective, or old, it can show symptoms sooner than a larger supply might. They do not always just blink out like a light bulb. They sometimes just fade away. Or exhibit weird spikes and dips in output. And that includes new ones.

You should test for voltages while running the box. A multimeter works best for this.

 

[DeaconFrost]

Overloading the PSU isn't the only issue that can arise with power. You aren't using a dedicated GPU, so I doubt the load is the problem, but as others have said, that doesn't mean the PSU isn't failing. If the problem exists with several different drives, that tells you it is either power related, or cabling/port issues. Port issues can be determined by swapping ports. If cables are swapped as well, then it might be time to RMA the PSU.

 

[ummduh]

Well, not sure what is going on now. The drive I thought was dead ( hence forth drive B) seems to be working fine on all ports after a cable swap. The drive I thought was fine is now failing (AGAIN!!! ) (drive A) using the same cable that allowed drive B to work fine. Failing the same way on all ports (well, 1-5 anyways, have not touched the boot drive on SATA0 since it hasn't given me any problems) .

Running Western Digital's diagnostic software returns:


Test Option: EXTENDED TEST
Model Number: WDC WD20EARS-00J2GB0
Unit Serial Number: WD-xxxxx
Firmware Number: 80.00A80
Capacity: 2000.40 GB
SMART Status: PASS
Test Result: FAIL
Test Error Code: 08-Too many bad sectors detected.
Test Time: 20:21:48, November 03, 2011

So, now I'm uncertain. This particular drive, being drive A, failed the first 2 times in a different case/different PSU. The first SSD failed in my current "new" system. The second SSD failed in a third older system.

Still suggest I monitor voltages? I do have multimeters. Just need to know what I'm looking for. BTW, PSU is 4 months old, cpu/mobo is a couple weeks old, this drive (A) is about 9 months old, (B) is .. well at least 5 years old.

IMO it looks like I just have terrible choice in hard drives.

 

[TVeblen]

Yes, we do not like to just chalk things up as bad luck here, but sometimes that is all it is. But we endeavor to try and fix things anyway.

You are most concerned with the 12v line when testing with the multimeter. I just open up the case and find one of the spare device power connectors. The yellow wires are the 12v lines, the black wires are the commons. With the machine running, you touch the black lead from the multimeter to the pole in the connector that attaches to a black wire, then touch the red lead from the multimeter to the pole that attaches to a yellow wire. You are not just looking for 12v on the multimeter, you also want to observe for any fluctuations in voltage beyond 0.1 or 0.2 volts DC. I like to observe the voltage while starting up and while using the computer. It helps if you rig the contacts so they will plug into the connector poles and stay there without hands.

The red wires in the PC are 5v. It would be a good idea to check these too.

 

[ummduh]

Just thought I'd update.

I monitored voltages and found nothing out of the ordinary.

I received the RMA'd replacement. The other drive I thought died came back to life in my USB enclosure. I have both the OLD "drive B" and the new replacement running without any problems now.

I did replace all the cables prior to this configuration and that didn't fix it.

I now have the SSD on SATA0, SATA1 is empty, SATA2 has the DVD drive, SATA3 has the "drive b", SATA4 has the RMA'd 2TB drive, SATA5 empty. Seagate 2TB external on USB2.

They sent me a WD20EACS drive instead of the WD20EARS.

Has been flawless since my last post or so. I've considered putting the case back together but don't want to jinx it now lol. I also have no use for the "drive b" but too paranoid to take it out since it's been working so well now. lol.. I use it for some 3x redundancy in backups..

Have come to realize it is the MOBO issue, but in reading GigaByte's RMA procedures.. I'm a little .. OK, a LOT po'd. I'll be buying another brand MOBO in the future and probably just throw this one out. I won't be buying any more Gigabyte products in the future.

I also got my replacement for the other failed SSD.. Surprised they still have new stock for the OCZ Agility drives to send out! It's still sitting in the box untouched for now..

 

[scrooge]

by chance are you using the (white) SATA 6Gb ports ?
if so remove the drives from the 6gb and only use the (blue) 3gb ports and see what happens.

scrooge

 

[ummduh]

I only have my SSD boot drive on one of the SATA 6Gb ports. all the other drives are on 3Gb ports. I haven't messed with that drive since the beginning and it hasn't given me any problems so I am a little hesitant to mess with it now that it is all working..

 

[scrooge]

i asked because i remember a friend had some trouble with a Gigabyte board and it had something to do with pluging both 6gb ports in at the same time. not sure if it was a bios fix or how he got it working.


scrooge