The Official PC building thread - 4th Edition

ddp,

I'm not sure how that happened. The message was for Oman7!

Sorry,
Russ

 

For those involved with my Hard drive situation, here's a little something interesting. My primary windows machine seems to think that the 5000AAKS drive that I'm working on, is completely empty. When windows acknowledges the drive for the brief moment that it does, it shows 100% free space. THat's rather curious. There's some definite miscommunication going on here...
While I realize that I should give up on the drive, I really don't like that idea That would make me a quitter. Besides, I wanna save the day here

 

Oman7,

Since it's a single platter drive, I would say there is physical damage to the disk. sounds like it dropped a head!

Russ

 

I very well doubt it is a head crash issue and number of platters, one or three or five, is pretty much irrelevant really. It sounds more like the controller board in the drive is going bad but that is also tough to say exactly. Not a good sign regardless.

 

I'm really wondering if a chip on the PCB went south, causing the burned area...
It was really dusty in there, and this particular Hard drive was on top of the other drive. Ergo, not too much air flow for it. With the HAF932 I have the wonderful front fan cooling the HDD bay I'll now insist on all builds, that there be a hard drive fan.

 

Not to much air flow plus heat rises so that could be it. It is really nice to have a fan in front of the drive bay I have several systems like that. Your WD drive runs pretty cool, I have several of that exact drive, and I'm surprised that it has started to go bad, but it happens.

If the drive was out of warranty and if you had another drive of the same make not in warranty you could swap the controller boards in order to try and recover any data on the drive but this is a risky operation and if you screw up you might wast your other drive.

 

How would it waste the other drive? I have a buddy, that has the other AAKS drive that I bought. I wonder if its PCB is compatible. What I can do, is image his drive, attempt to recover data from the faulty drive with his PCB, and then put it back on his drive. If it's screwed up, I can replace his drive. I'm probably gonna end up paying something for this anyway...

Oh my god!!! My brother has the AAKS(a second one)!!! I can have this fixed tomorrow if they're compatible! Man, I've bought too many drives. Is that possible? LOL!!!

 

Mr-Movies,

With all due respect, the tell-tale is the empty drive, which means the bios chip is intact, but the file allocation isn't. It's just reading the raw drive spec. In 22 years of building computers, I've only see this a few times, and all the indications are that there is physical damage to the drive, more than likely caused by head contact with the platter. It would also explain why the drive vanishes after a few seconds of running. I suspect that a head has gone dead short and caused the burns dead shorting to the platter. It would also explain the power draw from the drive and the burning on the controller board. Even if the controller chip burned out, it wouldn't burn the board where the burn occurred. Generally there's just a small puff of smoke and the chip dies, without any further damage. The fact that the drive runs and identifies itself tells me it's not the controller because when the chip does burn out they are always totally dead, and don't even spin up, because it needs the chip to tell it to spin up and then un-park the heads. That plus the chip itself could never draw enough power to cause the Damage that Oman7 showed me earlier.

Respectfully,
Russ

 

If the head had hit the platters, wouldn't smart statistics say so(Reallocated sector counts)? Because this failure has been progressively getting worse for about 2 weeks now. And last time the drive was working-ish the day before yesterday, the smart statistics checked out ok. I'm not trying to argue. I'm trying to understand In any case, we have another AAKS drive on hand, to attempt to swap PCB's. We just can't throw away 2yrs of pics, without exhausting all options. Short of spending a ridiculous amount of money, to have a professional take it in a clean room, and do their thing

Russ, I entertained the idea, that if a chip/capacitor/diode/etc were bad enough, it could send inappropriate amounts of voltage, causing burns. Granted this would be a rare worse case scenario

 

Oman7,

I'm only supposing that, because it's the most common thing that causes these type of problems. You could get the same symptom with a shorted drive motor, or if the spindle froze. Whatever it is, the drive itself was pulling the power. I'm mildly surprised that it didn't trip your PSU. Then again, the power comes from the PCIe bus, so who knows there. If you plan to recover data, the cheapest I've ever seen was $1400, and that was for a 160GB drive. The particular Lawyer that owned it was too cheap and turned down the option of a RAID mirror, something I recommend for all professional people who can little afford to lose all their data.

Russ

 

The only time I've ever seen a drive show up as empty, as opposed to corrupt data when accessing a file, or not showing up as even formatted, is when a logic problem has occurred, usually with BIOSes. This was a big problem for Seagates a while back. I've seen quotes of around $600 for a 500GB drive, but still very pricey.

RAID is all well and good, but no substitute for a proper backup. If you don't have an off-site backup of the really important data, then it's your own fault when you lose it.

 

Actually not, in 30+ years in the industry if there was a head crash you would still read some drive space typically unless all heads were dead, highly unlikely. The fact that it works then doesn't totally points to the controller. Sorry but I disagree with you and the platter numbers especially, even a single platter drive has multiple heads not all crash at once typically. You don't always have a small puff of smoke when a PCB goes down, they can fade, work and then not being very intermittent and very difficult to trouble shoot some times. There are many things that can go wrong, microchips, caps, diodes, and even resistors amongst other stuff.

It is true though that if the drive motor goes out that you would see this problem too. That is to easy to derive though as the drive would obviuosly not be spinning, no gyro effect or spindle noise.

Sorry Russ,
Steve

 

Steve,

I was thinking more in terms of a head shorted to it's housing, case, or what ever you call it, contacting the platter. I was not made aware that this was an ongoing problem, either. Here's the picture Kevin sent me. Why Windows won't let you open it with Windows Picture and fax viewer, is beyond me. It will only open with Windows Media Player, which sucks for pictures! You can't even select it from the list because it's not there and I cannot locate it in Windows!
http://img163.imageshack.us/img163/7591/burnedpcb.jpg

I had to cut the resolution in half or it wouldn't fit on the screen. Looking at it now, is that Nicotine stains or burnt? I see other places on the board, particularly around the big hole on the left that show the same color. I wish he had taken a picture of the complete board, both sides, preferably a lot smaller than what he sent me. I posted a direct view so it didn't tweak everyone's screen.

My background is electronics, which I decided on when I was 7, back in the days of tubes, so I've been at it for almost 60 years. My experience with computers goes well beyond PCs, having spent the last 30+ years in Medical and Dental electronics. Some of the equipment I work on involves computer power hundreds of times more powerful than the best and fastest computer you can buy, with up to 64 Microprocessors. CT scans, Full motion 3D X-Ray, MRI, you name it, I've probably worked on it!

Russ

 

Those are unusual marks, if they were burn marks, I'd expect to see at least some damage to the PCB itself, but nothing is evident. Windows photo viewer reads the image fine for me (though I did have a problem with it recently)
I must admit though, I'm struggling to see what I'm actually looking at, since I haven't taken a PCB off a disk before.

 

Well that's what I figured too. If those contacts were burnt, wouldn't you see burn marks on the surrounding PCB?

 

Sam,

Recently windows has changed the way some things work with hotmail. A lot of things that you could just click on before and see, now require Silverlite, which I installed at M$'s insistence, which also does not work.

I agree with you 100%, if those were burn marks, there should be some sign of damage, and I don't see any at all! The original picture when I downloaded it was 2592x1944. I had no idea that it was that big looking at it with WMP. I reduced it to 1296x972! I'm betting that it's not burned at all, but rather has been stained by Cigarette smoke! It's even more apparent now than it was in the original view I got of it with WMP. I would be far more comfortable with Kevin changing the controller now that I can see the connection better!

Russ

 

I agree with Sam those are not burn marks. Those are from the conformal coating they use to protect the board assembly and are normal. If you had a high heat issue from a component failure the multilayer board would also show a burnt haze too.

Thank you for the picture that does help.

You won't find Windows Photo Viewer as it runs as a shell file:

C:\Windows\System32\rundll32.exe C:\WINDOWS\System32\shimgvw.dll,ImageView_Fullscreen

Russ, I've been in the business as well since I was around 7 years old, which would put my start around the mid 60's and yes I've worked with tubes as well, remember nixie tubes. My career has been in aerospace, aircraft, ships & subs as well as jet & shaft engine testing. So I've been on top secret bases, foreign military sites, and commercial airline test cells amongst many other things. Most of my career has been in design and development.

Omega, you need to isolate the problem with the drive out and not hooked up plug a live power connector into the drive and verify that the spindle is running as Russ made a good point about the motor possibly being bad. If it is not then I think you might want to try replacing the controller board.

Stevo

 

How fortunate we are to have two such veterans of the industry in one thread

 

Steve,

I have to ask, because everyone always asks me. What was it about electronics that made you decide at 7 thet this is what you wanted to do. BTW, my 7 was in 1952. NRM Radio Electronics opened it's doors on July 4, 1952 in the town next to mine. No one had ever heard of an electronics store before, especially a Wholesale Distributor! I spent every spare moment in that place. My Bicycle was always out front! LOL!! One of my Mentors when I was a kid was the guy who coined the term Audio/Visual

It's interesting that our careers have somewhat Paralleled. I did a lot of work for NASA, in the Mid 70s. I spent about 20 years shuttling back and forth to Europe, South America and the Middle East in the 80s and 90s, because I had developed a reputation for getting things done and done right, and I had the ability to make things work when other people couldn't! I was offered a job with a large Medical Equipment manufacturer, working out of Marseille, France starting in Jan. of 2001, but I had a severe heart attack near the end of 2000, and I've only really worked a regular job for about 8 months since. I had a Heart attack on that job in 2006, and haven't been able to work since! I had originally planned on retiring in France in 2005 at 61! They were real nice about it and kept the job open for me for two years, but I just couldn't do it.

I can't really complain though. I've gotten to see a good part of the world, and somebody else paid for me to be there! LOL!!

Nixie tubes! LOL!! When I first went to Nasa, all of their test equipment used those. I don't think I've seen one since!

Best Regards,
Russ