System Issues

[TheTranquilEye]

Not often I get stumped but I need the forumers opinions.

My rig keeps freezing (IE no mouse, K/B, nothing) and a reboot into either of my OS just gets to the point of the XP or Vista loading screen and then goes to a black screen and no more. I've tried the recovery console in XP but it just sits at 'examining drive'. But, if I leave the PC until stone cold, it'll boot up. So, it's a heat issue? I'm not sure. My case has a 12cm fan at the front blowing air over my 3 HDDs and an 8cm at the back blowing out. I fired up Core Temp (I hope that's the one to use?) and it shows both cores at 28-35 even under load. The weird thing is, I leave my PC on 24/7 and I can get on it first thing in the morning and it'll be OK for a couple of minutes and then freeze. There's no load on it as I usually just look at the morning's news on t'internet.

Also, on reboot, the S.M.A.R.T. status of both OS drives (but strangely not my storage drive) now shows as bad.

WTF is going on?

TTE

Edit: Just as a precaution, I've reset the BIOS to default values, so I'm not overclocking.

[mattk]

unusual that both drives would die at the same time. but it happens occasionally.

i would backup everythign you can and then replace or RMA the drives.

[system7]

I never trust error messages to mean much beyond a fault somewhere...and that board has two IDE/SATA controllers, which complicates things to near-headache level. Try booting a Ubuntu CD, see if it behaves itself better with hard drives disconnected. You might be able to diagnose a SATA fault separate from IDE fault, but it could even be a memory problem and two sticks would be worth trying.

3 drives will pull 75 Watts at spinup and some power supplies will struggle. Maybe a single drive offers more hope of a clean boot. What is the power supply anyway?

A lot of drives are repairable with downloaded makers utilities, zero-fill being the most certain total wipe if recovery console fixmbr and fixboot c: fails.

[TheTranquilEye]

Well, I would RMA them but they were replacement drives from NT after I had this happen before. The only common denominator, in my view, is the PSU which is a 600W from Overclockers. I have no idea what make it is as I'm not at home. I think I'm going to swing by the Reading branch and pick up one of these:

http://www.novatech.co.uk/novatech/specpage.html?HYP-4M880

I'll disable the SMART status monitor in the BIOS and keep my fingers crossed the drives aren't really damaged. My data is safe as houses as I backup regularly.

[mattk]

before forking out on things find out what make your drives are and run the diagnostic utility. if they are samsung then ignore the SMART warning but if maxtor/wd/seagate run their diagnostics software. in some instances they can repair them and you will get an actual reading from the drive.

[system7]

before forking out on things find out what make your drives are and run the diagnostic utility. if they are samsung then ignore the SMART warning but if maxtor/wd/seagate run their diagnostics software. in some instances they can repair them and you will get an actual reading from the drive.
Heh, Heh, tis' true. Only get another supply if you fancy one anyway. This could be anything! Don't forget that unplugging the supply at the mains is the most thorough reset with a fault condition. But I'd definitely lighten the load and try that Ubuntu thing with less memory to get a feel for this problem. I think you can read HDs from a Live CD, but not quite sure. You definitely can with Knoppix.

Hiper 880W is a bit OTT anyway, isn't it? :P

[TheTranquilEye]

They are Samsung. Why do you say to ignore the warning? Is there something I should know?

TBH, I've been looking at getting another PSU anyway 'cos I have plans for my rig that will need some extra oomph.... :wink:

[mattk]

samsung drives dont report smart status normally. i don't know why they don't. maybe they didn't integrate the SMART chip on their drives, or have used a different one. best get the samsung diagnostics tool.

[system7]

Going off-topic for a minute, as we often do, you certainly push your spec hard, Matt.

TT 420W PSU is not what I would guess for all that hardware of yours... :P

And what brand of DDR2-667 will clock at 843? How did you set that up? :?

[TheTranquilEye]

The last drives I had and RMA'd were Samsung. When SMART stated they were bad, I ran the Samsung diagnostic on them and they both failed all tests.

[mattk]

yeah my PSU be a bit iffy. its really old now. but the amps are sufficient to run on my system.

rendition memory. which was (and could still be) one of our standard NOV codes. still stable except the random powering off. still not decided what is causing that maybe where the PSU is so old. volts at 2.15 i think. But thats the highest and most stable as i could get it. 843 is the average mhz coz mb changes fsb betwen 328, 329 & 330

didnt upload results to orb, will have to get all the info from BIOS when i get home

[Nox]

i was getting an occasional smart failure on one of my drives, and other occasional oddities resembling a possible low power warning, shoved a corsair hx1000 in there and it all went away! was running my rig right on the limts though :)

Nox

[TheTranquilEye]

i was getting an occasional smart failure on one of my drives, and other occasional oddities resembling a possible low power warning, shoved a corsair hx1000 in there and it all went away! was running my rig right on the limts though :)

Nox

Yeah, I'm pretty sure my rig's straining my current (excuse the pun) PSU. I've Fastracked the Hiper 880W at the Reading branch, so I'll pick that up on the way home from work in a while. Might give me the chance to tidy up the rat's nest cabling too.... :roll:

[system7]

Just a minute TTE. Are you still running an overclock?

Pushing a E8200 to 3.4 GHz which is 28% must be pushing DDR2-667 memory up to DDR2-850 in sync mode.

That's pushing it, even if MattK's memory will do it... :?

[TheTranquilEye]

I've reset the BIOS to default, so it's all running at stock. Besides, before I had all the trouble before, I was running at 3.4 for 5 moths with no issues.

[mattk]

you must also remeber when it comes to overclocking its all dependant on the actual parts you have. just because someone has the same memory and board i have it doesn't mean they can push it as far as i have. every chip has a different tolerance. as far as yet with a q6600 G0 step i have not been able to get it past 3Ghz on my fathers machine.

[system7]

Oh Dear, sounds like Memtest86+ might have to be deployed here if the PSU doesn't fix it. RAM isn't indestructable. :shock:

Good Ol' Ubuntu has it as a boot option... :P

[Nox]

you must also remeber when it comes to overclocking its all dependant on the actual parts you have. just because someone has the same memory and board i have it doesn't mean they can push it as far as i have. every chip has a different tolerance. as far as yet with a q6600 G0 step i have not been able to get it past 3Ghz on my fathers machine.

also is very dependant on how clean the electricty supply into your house is, and through to your computer and ambient temp of the room.

Nox

[TheTranquilEye]

Right, first update.

Stuck the new PSU in last night and PC ran up fine. Started running the XTC3 benchmark and it just rebooted. It went 1/2 way through POST and rebooted again. Then it just looped, getting to the same point in POST each time. I turned it off completely and left it for 1/2 hour during which I checked all cables and connections, which were fine. Switched it back on and it booted into XP. Went into Firefox and it rebooted... :x

This time it wouldn't even POST. Turned it off and back on and then noticed the red LED on the mobo wasn't lit. Tried another power lead and socket, no joy. FFS!

I then took out and reseated the 24 pin power connector. This time the LED came on. Booted up fine and has been on all night running Orthos. No probs this morning, ran the XTC3 demo and posted my results. So far, so good. Waiting now for CityLink to deliver my new RAM.

Why a connection should suddenly become dodgy when I haven't touched it or moved the PC is a mystery.

Watch this space...

[Jonny2Bad]

Yes very odd, but could so easily be a lose fitting somewhere dry solder joint etc etc.

[TheTranquilEye]

Certainly had me to the point of tearing out my remaining hair last night. I may take the mobo out later and give it a thorough inspection. Bit of a pain as I spent ages last night tidying up the cabling... :cry:

[system7]

Well, that's a good bit of work there, TTE. Progress anyway.

I guess you now know there is an issue on the 24 pin connector, either with the new Hiper 880W or on the motherboard.

A visual inspection seems like a plan. Especially on the 5V pins. All power supplies default to 5V standby mode being permanently on.

I'd be tempted to try the old power supply again.

[TheTranquilEye]

All looks OK to me System, well as much as I can see. Anyway, new RAM arrived, stuck that in and booted up fine, no issues. Just tried to up the FSB to 425 and wouldn't boot! Goddamit, my old generic 667 managed it. Oh well, I guess I'll have to start messing with voltages at some point. Can't be *rsed right now but I'm sure with some help from the lovely forumers, I'll be able to push my little E8200 to a decent speed.

Any road up, just in case you're wondering, here's a picture of my rig. The case is a re-used Fujitsu Scaleo from my old PC. It's actually not too bad and is all metal construction. The only p*sser is that I had to cable-tie the orange fan to the front of the hard drive bays to keep the drives cool. I will spend some more time tidying the cabling but TBH that Hiper has a lot of connectors...

http://i14.photobucket.com/albums/a347/spraints/IMG_2689.jpg

[TheTranquilEye]

Having lived with the S.M.A.R.T. status on my 2 replacement Samsung drives for some time by simply turning off S.M.A.R.T. monitoring in BIOS, both drives are now failing badly.

One has my XP 64 bit OS on it, the other XP 32 bit. Looking in the Event Viewer System logs, both drives are reporting disc controller errors. Symptoms include freezing, BSOD and refusal to boot. Both completely fail the Samsung diagnostics.

Clearly, both drives are knacked. Now, I can rule out (pretty much) my other hardware causing the issue as I have a new mobo, new PSU and new RAM. I'm concerned that I returned two identical drives to NT for the same issue. These lasted maybe a couple of days before I started seeing the bad status in BIOS.

I have bought a WD 500GB which I would like to clone my XP64 Samsung on to with a view to RMAing the faulty drives, still leaving me with a working system.

Anyone have any ideas why the drives have failed and what proggy is best to clone to the WD drive?

Edit: Weirdly, throughout all this (and the previous issue with bad drives) I have had a 400GB WD drive connected the whole time which is unscathed. It's a storage drive and has no OS.
This throws up interesting questions. Is there a link between the OS and the drives going down or have I had a bad run of Samsungs etc.... :?

And if someone from NT (maybe Danno?) could tell me if I'm going to have issues returning the Samsungs?

Cheers, TTE

[system7]

You want to get Western Digital's Data Lifeguard utilities here, TTE.

Tools and Diagnostics for DOS (CD). Covers all the bases. It's that Ontrack standard software. You can do most anything with it, including cloning.
http://support.wdc.com/product/download.asp?groupid=606&lang=en

I don't know what's going on here, m8.

But the new WD drive sounds like a plan, at least. :?