I have my sister's computer here and I am having some issues determining the exact problem. I suspect the PSU voltages to be the issue of 2 recent damaged HDDs. Both of them died when put into this system in about 9hrs of operation. Here are the voltages reported from my digital PSU tester when just the 20pin molex is connected to it (other are plugged into other hardware): +5v :: 5.3 +12v1 :: 12.2 +3.3 :: 3.4 -12v :: 11.9 +12V2 :: LL (from not 24pin molex I assume) +5VSB :: 5.2 PG :: 280ms My question is the +5v rail. Because the limit is 5.25, would 5.3 likely destroy hardware? (like the two HDDs) Also, what is the 5VSB, and PG for?
I think the +5VSB is for the +5V "Stand By" to provide power to certain devices even when off. Things that might be used to turn on the computer via something other than the power switch: i.e. LAN or keyboard. It could be bad luck, but if you do suspect the PSU to be a problem, swap it out with a known good unit for awhile and see if the bad luck continues. I do find it odd that the hard drives are the problem and not the optical. Do you have a Y power adapter in use by chance? I'm just trying to see if there's something else that might come up as an issue.
"I think the +5VSB is for the +5V "Stand By" to provide power to certain devices even when off. Things that might be used to turn on the computer via something other than the power switch: i.e. LAN or keyboard." Ah, ok. That's probably it then. Thanks. "Do you have a Y power adapter in use by chance?" No, its directly to the PSU. It's only a 250w (not sure what brand); fixing up an older computer. Trying to get a replacement is gonna take some time :x: Yea, its weird that only the HDD's are being affected. Optical is fine. Every time I install any HDD into this system, mechanical failure occurs . Installed 100s of drives so I know its not me. Edit: Just in case... If not the PSU, what else could it be?
If it's not the PSU, you have extremely bad luck with the HDD or some very bizzare IDE issues. It also might be the particular lead going to the hard drive, but it seems to be an issue between the PSU and HDD only, and based on multiple drive failures, the suspect list is pretty narrow.
Alright. Well, Thanks for the help. I will post back if I get this resolved so I can let all of you know what it was.
I agree, the chances of the PSU failing a HDD are pretty small IMHO. The 5v supply may be tad over because it will drop through the wiriing. Also you don't say how you measured these voltages - was the PSU loaded....I hope so.....otherwise you'd get strange figures I believe. I suspect it was connected to the working PC...yes? How are you defining a failed HDD? Sure, if it doesn't work on another PC, but I wonder it's just this PSU at fault.....particularly as you say the Optical drive is still OK. =after considerable running time. EDIT:However, on rethinking this, it could be that the Optical drive controller is particularly able to switch off (so any overvolatge doesn't get through).......yet your V-measurement shows no unusual voltages - - - - - I wonder if there's a heat-issue with these HDDs - they can run quite hot normally. Any unusual ventilation, hot room, sun-on-the-computer etc can reduce the life of parts. I wonder "why" you are focussed on the PSU. . . . . . . HDD do fail and a "spare one" may already be close to failure, but it won't tell you. . . . . . . Do you know what started this "failure problem" - was the PC knocked off its desk, perhaps? Presumably the PC is not new, so it's been OK for a long time - - - then "sufddenly" there is this HDD failure issue, . . . and you offered to help. If the replacement HDD was "new" (You don't say).....just how good was it? It surprises me how reliable they are, what with vendors handling them like slabs of cheese (can I say that?). I will never buy one through the post, by choice. . . . . It's just asking for a short-life. Buy them out a large box of drives you believe are untouched. Manufacturers go to great lengths to make their HDD reliable. IMHO they run "longer" than the software we put on them!!! Good luck.
I measured the voltages on proper load (HDDs and such) with a digital PSU tester. they failed mechanically i assume because both drives could not be read on a known working system (set as a slave on the working system). The BIOS did not even register it. Both drives that failed were 4 years old, bought new at the time, and slightly to moderately used; no benchmark software or other high stress tasks were used on them except the occasional defrag. I do have to say that one of them was a 40GB IBM; the other was Western Digital 60GB. As for the obvious signs of damage causing events -- the computer was used by another PC savvy person as a workstation and was converted into a play computer when the drives were bought. I highly doubt they were subject to falling. Well, I was focused on the voltages because no other indications were presenting it self to a sudden death of a HDD. EDIT: what about improper voltage grounding?
I think I'm running out of ideas - however, it occurs to me to ask: This "tester" - do you mean a digital multimeter? ((Your reply implies a dedicated power loader. In which case you would not have the HDDs connected. The "tester" would go through a range of loadings and give you a time/performance log....I suspect this is not the case....)) + Were these (now dead) HDD's acting as Slaves on yr sister's PC - Or did you install new OS and some working software...? The disc-sizes are quite small, even for 4years old - do the discs have a date-stamp? (maybe ID from the chips, like 22-96 indicates week 22 of 1996 for example.). I wonder they are somewhat older than you say...? Not sure this is going to be resolved, as several others (here) have tried also. I'm wondering why you haven't replaced the PSU? - they are cheaper than HDDs - Surely with a decent HDD and new PSU the chances of another failure has to be very small - probably enough for an old PC. If you want belt+braces, then maybe replace the mbd also, provided it will take the RAM and Video card. (it would be too much to expect it to take the old processor....). It sound extreme, but ...."family".......keep 'em happy. Re grounding - yes, it's a possible source of trouble - those 4-pin power-connectors to HDDs (with IDE, I recall) were somewhat unreliable, but my experience was they just stopped the drive working (no power, no go!). I have only had HDD failure due to LT "wear-out" - ie there was no other culprit to blame. I hope you've tried changing the ribbon connector leads...? Is your sister sure no-one had done something to the old PC....what was her first "reported fault"? I suspect it was the primary HDD Finally I wonder if it was excess voltage that failed the drives? This might be from Storms, or local welding plant, even a badly supressed 1HP motor. It could affect the HDD, simply because that is operating at most-times, whereas the Optical is only on rarely.....in effect it avoids the spikes. Thse spikes would not show on a DMM but it's spikes that kill semiconductors.....like the motor controller. Fit a generous supressor, not one of those "Computer" types, I mean something with decent iron in the chokes.....My old win98SE pc has a PostOffice supressor on it, with a CVT in addition.....practically nothing can get through that. However, my Dell is straight on the mains....with nothing untoward.....I suppose I've softened my fears...and "here", there are no obvious spikes. Hope that helps. Good Luck.
One HDD was a master, the other a slave. Yes, when I converted it to a play computer, I wiped both drives with quick format, then installed OS on larger drive. I checked both drives, and I think she was guessing at the age because the IBM said 1999 and the WD said 2003. I replaced both ribbon cables when the first HDD failed. The first symptoms on the IBM drive were BSODs reporting memory alloc failure -- then loss of data, then it just died. WD was BSOD almost every boot during windows explorer usage with same error, then just died. However, I am just going to replace both mobo and PSU and get this over with, getting a headache over this. I think your right in which this is probably not going to get resolved so thanks for all the help. Everyone's posts drew me closer to just replacing those two components. Someone can close this thread now...
