Diskette drive 0 seek failure--sign of bigger woes?

All: The Dell Dimension 8100 that has been the second PC in our home network for about 5 years has frozen up from time to time, forcing a reboot of the system to bring it back to life. Two weeks ago, it failed to get through the BIOS startup, and gave me the following message: Diskette drive 0 seek failure Diskette drive 0 seek failure Alert! Hard drive thermal probe not detected. Alert! Cardcage fan not detected. Alert! System fan not detected. Alert! Chipset heat sink not detected. System halted! I figured something was heating up in the system, and unplugged it for about 10-15 minutes, plugged it back in, and it rebooted. The machine was left running until yesterday, when I removed the Microsoft Internet Zone games from the system (Internet Hearts appeared to be corrupted, and wouldn't shut down, leaving an area in the middle of the desktop screen all screwed up). I had to reboot to complete the job, and I got the same message set during BIOS startup. I unplugged the machine, waited about an hour (I was doing something else besides this downstairs), plugged it back in, and tried to reboot again–no luck. I left it alone, and tried it again a few minutes ago–still no joy. I figure there's a connection somewhere on the motherboard that's intermittent, and sometimes lets me reboot, and sometimes not. I just don't know what to do or where to start to see if I can do something about it short of taking it to a repair shop or saving up to get another machine–even a refurb, which this one was… What do I do now? Thanks–Tom vonHatten

Could be one of a few things, the fact you left it this late to get help complicates things, and could make it a more expensive repair, depending on whats going on that is, currently I have 3 thoughts, dying\dead hard drive, heatsink needs re-seating, or worst case bad power, and who knows what can go when that happens, but lets start with the simple checks, boot into the BIOS, and look for "System Health" or similar, see what the temp's are, leave it running for a short while and see if the temp's increase, if it hit's 70 degrees shut down, and report back with this info.

Next identify the make of you're hard drive, then go to the makers site and download they're utility and scan the drive for errors, use the extended scan, see what this shows, quick point to make, if system temp's previous are up at 70 + then do not scan the drive, we need to address the temp's before scanning or it will get too hot.

Bad power we will cover once we have this info, as it is the worst case, hopefully one or both of the other tests will be the cure, to identify the drive look at the listing in the BIOS, and search google for what it shows, and you will see the make, if you have the UBCD that has all the makers utility's on it already, otherwise just get it from the makers, let us know the results and we can go from there.

Thanks for the suggestion. Unfortunately, I tried to get to BIOS setup, and couldn't. I get the message litany I described earlier before I can get into BIOS setup, apparently. Now what? –Tom vonHatten

Well I think a 2 step attack is required, to begin remove the heatsink, clean it off (I recommend Arctic Clean highly) with a lint free cloth, then apply some new thermal compound, Arctic Silver 5 is highly thought of as the best, if heat is the major issue stopping you from getting to even the bios, then this should cure that, if it works then you can proceed as suggested above.

The other thing you can do is remove the drive from the problem system, add it as a slave (check the jumper if ide) in a working system, then scan the drive in that system, and if able\needed save any files you don't want to lose just to be safe, once you have done these tasks we can see what the results are and see what way to go next.

I'll gather up the cleaner and the thermal compound tomorrow (I can get them at a shop in the St. Louis area), and then go after it tomorrow night, i guess. Will I need anything else besides these items at this point? –Tom vonHatten

Well my hope is that this will be all you need, if this is purely temp related then it will be all you need, and are very useful things to have if you mess around with a lot of systems, but even this is the only sort if time you need it, the ease of cleaning with this, and the benefits of a good thermal compound are well worth it, and will last for quite a while, you only need a few drops of cleaner, and mines lasted a few years with occasional use, best things I ever bought maintenance side, as to any other issue, until we can scan the drive, and access the system we would be simply guessing, but currently this is all you need assuming the existing is\has failed now.

Oh–am I just removing gunk from the heatsink, or do I need to do the same to the processor? –Tom vonHatten

You need to remove ALL existing\residual thermal compound, otherwise the new compound can be compromised, I usually give it 3 goes with the cleaner, 1st gets the worst, 2nd gets the stuff that got left behind, and the 3rd gets the residual that you can't even see by this stage, then the purifier gets a bit more and prepares the surfaces for the new compound, it ensures the fresh compound can fill the voids and make good contact to both surfaces to give the best performance. Are you used to applying new compound ? Or would you like a brief guide on what I do ? *edit To clarify you need to clean BOTH the heatsink AND the CPU, just incase I wasn't clear enough here.

I've never done this before–so some directions might be a good thing (Martha Stewart's overuse of that phrase notwithstanding)…. I've looked at some other information online, and just want to be sure I'm removing only what's needed. For example: in some of the pictures, I see a thermal pad on top of the processor. Does this need to be taken off? The other sources say to just put the thermal conductor on top of it and smoosh it out. It's things like that I wan tto be sure I'm not screwing up. Thanks for the help so far–TvH

Well there are a couple of ways depending on where you get info from, there doesn't appear to be a way set in stone, but IMO part of one way of doing it comes from the old AMD days, and thats the "smoosh" way, over clockers go to the extreme as they need the max performance from it, and this involves creating a mirror finish on the parts prior to applying the paste, but for the average user this is overkill, so the method I use is what I will explain, but there are other just as valid variations, but from doing this over the years I believe this to be the best way to maximise the result without going crazy. Cleaning the old stuff off is essential, what you find will depend on how it was built, some use a "sticky" pad, if this is the case here you need to remove ALL of it, the solution I suggested makes easy work, but will take a few goes depending how thick it is, but repeat as much as required to clean it all off, if it was purely a paste (as what will be used when done) then it should be easy going, but make sure you get both the heatsink and the CPU, depending on how things are fixed. You may find it easier to clean the CPU in place, it minimises the risk to it, but take care not to spread mess or solution off of it, I use 2 or 3 drops at a time as needed, it minimises the risk of spreading it too far, the heatsink can be done in you're hand, but do take note of which way it comes off, it must go back the same way or you wont be able to secure it again. Once the old stuff is cleaned (I suspect there wont be much there given the issues so far), if you use the solution I suggested use the purifier, then you are ready to apply the new paste, this part is optional, but what I do is get some cling film over the end of my finger, it's very important not to touch the surfaces with bare fingers, the acids etc in you're skin can have an impact on the job, then apply a small amount of paste and rub it into the CPU, and the heatsink, this helps fill in voids, although the surfaces may look nice and flat, they are far from it, bin the cling film now. Next apply the paste around the CPU, I tend to run a line parallel line to the edge, but say 5mm or so away from it, and a cross from corner to corner, I then use a card to spread the paste evenly across the whole surface, you could use a bank card, and clean it off when done, or what I have done, is occasionally you get card offers in the mail, they send an example card, I keep and use one of these to keep my cards clean lol, but they allow you to get a nice even covering. Now the heatsink is a little trickier, my guess is even cleaned, you will be able to see where it makes contact with the CPU, because you do NOT cover the whole heatsink, you only cover the area where it will touch the CPU, go a tiny bit over to make sure you cover any risk of being slightly out, but don't go crazy if you follow, and don't use too much paste on either of the parts, you don't want it squeezing out as you apply the heatsink and getting everywhere, so you don't need lots of paste, it's easier to add a bit more if you don't quite have an even covering than it is to take it off. The last part is putting the heatsink back in place and fastening it again, this is where the most care is needed, you need to place it as flat as possible on the CPU, otherwise it defeats the purpose of preparing the surfaces nice and flat, and I find it very hard to do it perfectly, the fixings often pull it about in my experience, this can't be avoided, all you can do is minimise how much that happens, and when fixed in place you are good to go, fire it up and see what happens, if heat was the only issue you should get straight into the bios, at which point you can see the temp's (bios having this option), which is redundant by now as we would have sorted that lol. I hope this makes sense, feel free to ask if you need further explanation, hope it helps.

I think as I take things apart and look at them it will make more sense. Oh–The motherboard, etc. is mounted vertically in the tower–will the paste stay in place once it gets hot, or will it ooze out and cause gravity-induced mayhem? –TvH

just don't use to much, especially with artic silver unlike ceramic pastes it is conductive so can create shorts that can break things when power is turned on, cpu, motherboard, etc.

here's a couple of links that might help.

http://benchmarkreviews.com/index.php?opti…mp;limitstart=4

http://www.arcticsilver.com/instructions.htm

As above, used the way directed it will be fine, it's a thick grease type paste that sticks to anything it touches, hence taking care, it wont change as such when the system is upright, but for ease of working it is best to lay the case (mains removed) on it's side to do this, being flat means it wont run anywhere (the solution that is), and this is why I say 2 or 3 drops (on the modern big square CPU's anyway) is all you should use each cycle max, or it could run off the edge just remember you can add another drop if needed, but you can't remove it if you start losing control of where it goes, so slow and steady wins the day as they say. Nice reads terry1966, hadn't seen the 2nd one, never realised they had so many ways depending on the processor, I've always known of the single line method, this came from the old AMD units, at least thats where I learnt it from, and even though as they are the makers, so should know the best, it just doesn't feel right using the line method, as I can't see how it would fill voids as well as the way I use does, not saying they are wrong, I just can't see it being as effective, but it must be good enough or they wouldn't advise that method, it just doesn't feel right with me lol.

Cleaned it, gooped it, reassembled it and….still no luck. I'm getting th3e same messages, and not out of the BIOS startup. What next? –TvH

That is not sounding good then, if it was a temp issue that would if cured it, fears now turn to more serious damage, eg power spikes frying the motherboard and the like, but lets not panic just yet, pop the hard drive into another system as a slave as suggested earlier, and scan that to see if it is working or not, I fear from the errors mentioned it will probably fail the scans, in which case a new hard drive is the case. The only good news here is you can get 1TB drives dirt cheap these days, so any replacement you choose should be affordable, if this is required I suggest going for WD (Western Digital) as they are solid performers. Then if you have a spare drive for testing, or end up having to buy a new drive due to this one failing, then try that in the problem system and see if it does the same, my fear is it will, because I don't see how a hard drive can block accessing the bios, but I have seen some crazy things happen so don't want to rule anything out. When you did the thermal paste did you look at the motherboard ? I am wondering if the are any obvious signs of damage, eg scorch marks or the ends of the capacitors etc being blown or the like, if there are any such signs then this confirms a fried motherboard, but the problem one faces is if there are no such signs, there is no conclusive way to test if the board is faulty or not, all you can do is test all that can be tested, and when all that proves good, whats left has to be the issue, but you do have to be sure all else is OK before taking such a leap of faith. Also do you have a PSU tester or meter ? If you did it can't hurt to check if the power is good or not, also it can't hurt to try booting with a single stick of ram (assuming you have more than 1 stick), alternate what you do have, try different slots on the board, if the current working system has the same spec ram, try the ram for the problem system in the working system, it will help to prove the ram is good or not, but to scan the ram it needs to be in a system that can at least boot to the optical drive, but all these things will need to be checked to narrow down where the issue is, but it may take you a while getting to a stage where you can do things like test the PSU, hope this gets you started anyway.