Hard Light Productions Forums
Off-Topic Discussion => General Discussion => Topic started by: redsniper on June 16, 2005, 10:52:13 am
-
It won't boot at all. When I turned it on the HD activity light blinked and then the cd and floppy drives made some noise and then it restarted before the monitor even had a chance to come out of sleep mode. After resetting the BIOS it continued this loop of restarting for about 5 minutes after which it made it to POST but no farther. Now when I restart it often fails to show the cpu speed or RAM count and whether it does or not I can't access the BIOS at all. It's worth mentioning that during the restarting loop I heard a kind of clicking/clunking kind of noise about once every two seconds which may have been a HD click o' death, but I don't know for sure. So what does this sound like? Is the cpu fried? Is the mobo toast? Or do I just need a new HD?
EDIT:
Asus A7N8X-E Deluxe
Athlon 2800
Maxtor 80GB
-
No beep codes?
-
nope
-
Hmm... my first thought is a PSU problem maybe. But that wouldn't really tally if you've been able to reset the BIOS (may be worth trying to reset the CCMOS with the jumper in case, though; maybe it got corrupted and didn't save correctly).
You should have a beep code if its memory, AGP card, HD, or chip problems IIRC.
-
CMOS, that's what I meant, I did the whole 'remove battery switch jumper replace battery' thing, that's what got it to stop the restart loop. I don't think this mobo has beep codes since there's no guide for them in the manual, it has a speech synthesis thing instead.
-
a7n8x-e does have beep codes, as well as speech synth.
-
well the only beep it makes is on the rare occassion that it displays the cpu info and counts the RAM it makes a single beep, but it always did that back when it worked.
-
The clunking sound you heard every two seconds was that of your harddrive re-starting several times. I wouldn't be too concerned about it. As for the failing to start, I can only see the CPU or Northbridge as the culprit at this point, and even then it's a little iffy. Anyway, you may not know this, but after a CMOS reset, you are required to set the jumper to force a 100MHz FSB before it can boot on most nForce2 based boards. That may very well be where you are.
BTW, now may be a good time to clean your thermal paste on your CPU and Northbridge coolers, and replace them with Arctic Silver. Also check your northbridge cooling fan. If it's dead you have to replace the whole heat sync if running at 200MHz FSB.
Good luck :).
-
I don't have thermal paste, my cpu uses phase-changing thermal pads and my Northbridge doesn't have a fan, just a big heatsink. I've had to reset CMOS before when I accidentaly overclocked my cpu and all I had to do was take out the battery and move a jumper over for a few seconds and it worked fine, but I'll see if I can find the jumper you're talking about and try setting it to 100MHz.
-
I haven't had a chance to work on my computer until now and upon checking my mobo manual, I've discovered that my FSB can either run normally or be limited to 200MHz, which is meant for Duron CPUs. My CPU is meant to run at 333 MHz FSB so I want to make sure that it won't be damaged by limiting it to 200 MHz. So is it safe to do this?
-
Originally posted by redsniper
I did the whole 'remove battery switch jumper replace battery' thing, that's what got it to stop the restart loop.
You did remember to put the jumper back in the normal position afterward, right? This can cause bootup funkiness too.
Originally posted by redsniper
well the only beep it makes is on the rare occassion that it displays the cpu info and counts the RAM it makes a single beep, but it always did that back when it worked.
The single beep typically means "All is well". You'll know a beep code when you hear it.
Originally posted by Scuddie
The clunking sound you heard every two seconds was that of your harddrive re-starting several times. I wouldn't be too concerned about it. As for the failing to start, I can only see the CPU or Northbridge as the culprit at this point, and even then it's a little iffy. Anyway, you may not know this, but after a CMOS reset, you are required to set the jumper to force a 100MHz FSB before it can boot on most nForce2 based boards. That may very well be where you are.
BTW, now may be a good time to clean your thermal paste on your CPU and Northbridge coolers, and replace them with Arctic Silver. Also check your northbridge cooling fan. If it's dead you have to replace the whole heat sync if running at 200MHz FSB.
Good luck :).
Actually hard drives clicking, particularly if it's with any kind of regularity, is almost never a good thing. redsniper, try unplugging the drive from the IDE controller and see if it still makes the sound. It could very well be the hard drive failing and the POST refusing to continue until it can spin the drive up.
Also, I don't know about the Asus crap but I've never had to touch the FSB jumper on either of the two nForce2 boards I've owned - even after a CMOS reset. The FSB does get reset, for safety reasons more than anything else I suspect, but it's been automatic and hasn't required any intervention from me aside from setting it to what it's supposed to be.
Originally posted by redsniper
I haven't had a chance to work on my computer until now and upon checking my mobo manual, I've discovered that my FSB can either run normally or be limited to 200MHz, which is meant for Duron CPUs. My CPU is meant to run at 333 MHz FSB so I want to make sure that it won't be damaged by limiting it to 200 MHz. So is it safe to do this?
It won't be damaged, the chip just won't run at its full speed that's all.
-
Alright, I had some time to mess with this on my lunchbreak. The click/clunk is coming from my smaller backup hard drive. It clicks whether the IDE cable is plugged in or not. Unplugging both the IDE and power cables doesn't help my computer to boot up. I still haven't tried limiting the FSB but I will.
-
I tried switching the FSB. No effect.
I'm really at a loss as to what's causing this and how to fix it. I mean, it's not even getting to POST... I don't know what to do.
-
Last time I got something like this (about 3 days ago), it was because my Hard-Drive was dying and taking my Boot Sector with it, though, if it's your second drive I have non clue how this could be making your system refuse to boot up, I actually managed to boot mine from my Windows CD and perform a Scandisk. I actually managed to get my crashed Hard-drive working enough to transfer everything off it, and I've just bought myself 2 160Gb Western Digitals to replace it ;)
My computer refused to make it to POST and also refused to give beep messages, and I also got the click-click noise. Sounds like something similar to me.
-
Have you tried an absolute bare startup, using nothing except CPU and memory? Unplug everything from their slots, and I mean everything. No video, sound, drives, or anything. Just run it totally bare, and if your PC speaker gives you a beep, then your problem is with a component. If it doesn't, you've got a break in your BUS. Could be bad CPU, bad memory, or bad motherboard, It matters not. If your memory is bad (which I doubt), your CPU and motherboard can be saved. Otherwise, it's best to replace every system component, including the PSU, as the CPU, Motherboard, and PSU can all contribute to a global hardware failure.
Anyway, before you do that, make sure the jumper/switch is not set at the reset position. That would definately prevent it from booting at all. It's a simple thing to check, and trust me, I have been foolish enough to overlook it more than once :D.
-
disconect both hard drives and see if you get to post
-
You know, I had a duron that starting doing this. The reason was that the heat sink fan had died, and it was stopping to prevent the system from frying itself. Have you checked the CPU fan?
-
all the fans work. It manages to start POST when I first turn it on after it's been off for a long time, but it's weird. When it counts the RAM it keeps counting higher and higher and once it gets high enough it just starts over from zero and keeps counting. If I press Del to get into the BIOS it goes into its restart loop and nothing happens. I'll try a bare startup and booting from the Windows cd.
-
Originally posted by redsniper
all the fans work. It manages to start POST when I first turn it on after it's been off for a long time, but it's weird. When it counts the RAM it keeps counting higher and higher and once it gets high enough it just starts over from zero and keeps counting. If I press Del to get into the BIOS it goes into its restart loop and nothing happens. I'll try a bare startup and booting from the Windows cd.
what kinda ram in it?.. those boards only take dual channel ram(meaning two sticks of the same type)..
oh, and check your cables, a floppy/IDE cable in around the wrong way can really wreak havok..
you have the exact same model mobo as i do.
-
offbrand RAM in all 3 slots, which the manual says is okay. Two 256 sticks and one 512. It hasn't caused me any problems for the past year.
-
Originally posted by Turnsky
what kinda ram in it?.. those boards only take dual channel ram(meaning two sticks of the same type)..
They don't force you to employ dual channel RAM, I ran a single 256Mb stick on my old 8RDA+ for months before ponying up the cash for a pair of 512s (which weren't an expensive matched pair kit either, just a relatively cheap pair of Hynix 512Mb PC3200 DIMMs) . If you only have one stick, utilise different size sticks or employ the third DIMM slot then it should fall back to single channel mode.
-
I tried a bare startup. It incorrectly displayed the cpu info (as 1250 Mhz) and counted the RAM, the count didn't start over though, it went to a blank screen and started the restart loop again but faster since it wasn't checking all the drives, the only way I could tell it was restarting was because the lights on the front of my case flickered each time it restarted.
-
The common cause of incorrect CPU speed is an incorrect FSB setting. Did you put it back to 133/166 after trying 100Mhz earlier?
-
no, but it displayed it incorrectly a few times before I switched the FSB.
-
Let your memory count up three times. It is common practice for the boards to identify the memory over three passes for verification purposes. Also, your 1250MHz reading is definately correct. You have a 2800+, which uses an internal multiplier of 12.5, and you are currently running at 100MHz FSB. 100MHz * 12.5 = 1250MHz.
The fact that you are able to view POST status for a while, then looping reboots, and finally no POST indicates that you definately have a heating problem. I suggest checking all the contacts and definately replace the stock heatsync on the northbridge with a real chipset cooler. Copper and a ballbearing fan are a must. And put some arctic silver on it too. Also replacing your heatsync fan on your CPU is a good idea, it may not even be transfering heat effectively right now.
Anyway, keep us posted.
EDIT: Take out the DIMM in the Dual channel slot. Dual channel setup cannot exist with an odd number of DIMMs, except in mixed mode, and mixed mode is very unreliable. That could also be your problem.
-
I'll certainly agree with the CPU Fan thing, it's actually made a world of difference by putting an Akasa BB Fan/heatsink to replace the generic cheapy one that came with the computer ;)
-
well, I got the AMD Processor in a Box package deal, where you get the chip itself and a heatsink with a fan attached. Screwing around with thermal paste isn't an option though. My particular cpu has a ceramic thing poking out the top which lines up with a pad of phase-changeing material on the bottom of the heatsink. When it gets hot, the stuff sort of melts to seal with the ceramic die. It's explained in some document I managed to scare up on AMD's site. I'll see if I can find it. Anyway, this means I can't just go buy any heatsink and slap it on there. Besides if all these problems are being caused by heat damage isn't it already to late for my cpu/mobo? I don't see how better cooling will fix anything if they're already damaged. :sigh:
EDIT: Oh, and it wouldn't boot from the Windows CD. Also strangely enough, the HD activity light was blinking even though both HDs were unplugged.
-
Yep, sounds like the stock HSF to me...
You're not forced to use those, it's quite possible to use aftermarket heatsinks - I have them on two of the three Athlon boxen I've built/I'm building right now. AMD don't exactly recommend it because aftermarket HSFs typically mean thermal paste and they're worried the stuff will "pump out" from under the HSF due to constant heating and cooling. However, with Arctic Silver 5 the swing has to be pretty extreme before it'll hapen.
-
One of us isn't understanding, probably me. Anyway, my cpu connects to the heatsink in a weird way. Here's a picture to illustrate:
(http://img.photobucket.com/albums/v458/redsniper7/cpu.jpg)
Here's the official AMD stuff as well:
http://search.amd.com/query.html?qt=thermal+paste&charset=utf-8&qp=&qs=%7C+language%3Aen&la=en&lap=en&qm=1&tqmhak=0&ct=1325960345&oq=phase+change&rq=1
1st result on this link^
but maybe I'm mistaken and you're just saying I should put thermal paste on really really thick?
-
To my knowledge, all Socket A Athlons are manufactured like what you describe with the CPU core directly underneath a piece of ceramic. They're built this way to get the core as close as possible the the heatsink surface to make the transfer of heat from the core to the heatsink more efficient.
All Athlon heatsinks, regardless of where they came from or who made them, only make contact with that ceramic block. When replacing the stock heatsink with an aftermarket one, you only have to put thermal grease on the CPU core. If you try and put it anywhere else not only will it not make contact with the hsf unless you lather it on really thick but if you're using something like Arctic Silver, which is slightly conductive, then you'll wreak havoc with all those little bridges on the top of the chip.
This is why AMD only recommend you use the phase change pads they put on the bottom of the stock HSFs: If you use something like Arctic Silver and subject the chip to massive swings in temperature, there's a chance the grease can be "pumped out" from under the HSF, possibly shorting out the bridges in the process. According to Arctic Silver, this only happens with temperature regularly gets above 100 degrees C making the chances of this occuring pretty slim for most people - at least according to them. I must say though that I've been using Arctic Silver for a couple of years now on various CPUs and I haven't noticed anything amiss yet.
If you're going to try replacing the HSF then there's one thing you need to know: due to the design of the chip packaging and the way the HSFs mount onto the socket a lot of pressure get exerted onto that piece of ceramic. If you're not careful in fitting the new HSF then it's possible to break the ceramic and effectively destroying your chip. There are a number of ways to avoid this but the simplest and cheapest is just to be careful. Take your time and if it doesn't fit right or you need to put a large amount of pressure on it (it's normal to have to put a bit of pressure on) then stop, take it apart and check you're not doing something stupid like putting the hsf on the wrong way (most have a lip that fits a part of the socket meaning they only go on one way). I've installed/replaced the HSFs on my Athlons here over half a dozen times and haven't broken a core yet.
-
man, i hate those stock heatskinks and the way you have to push that clip down with a screwdriver.
-
The phase change pads are designed to be that thick to enforce exactly what they are for, padding. In reality the pads are VERY compressed and that's one of the only reasons they conduct heat. Arctic silver, ofcourse, is not that way. You are supposed to use no more than 0.015 inches thick across the die. In layman's terms, a very small coating. You apply the bare heatsync (a REAL one with no paste or pads) onto the die, and clip it down. The compounds will do the rest. Over the course of a couple hundred hours of normal use, the bond becomes fully stabilized. The end result is equivelant to figure 1 of your MSPaint diagram.
The chipset cooler is an entirely different story. the entire surface of the chip heats up. There is no die, per se, but the surface area is where heat is dissipated. Most of the time, replacing the chipset cooler requires removing your motherboard from the case, and undoing the mount devices on the back. Fairly simple if you know what you're doing. Once you have a copper cooler with a fan, paste or padding (doesn't matter which) is applied to the entire northbridge chip face, and you attach the cooler the same way you detached the passive sync. It would be equivalent to figure 2 of your MSPaint diagram.
And that's really all there is. Read up more on it here.
http://www.arcticsilver.com/as5.htm
EDIT: Damn you admiral, you beat me to it.
-
Originally posted by FireCrack
man, i hate those stock heatskinks and the way you have to push that clip down with a screwdriver.
It's not just the stockies, a lot of aftermarket ones do that too. It's a bit unsettling knowing that one slip and you'll give your motherboard a new mounting hole but if you take your time and do it slowly then there's little to worry about.
-
Aha! thank you very much for enlightening me. I didn't realize the paste would only go on the die. However, if my cpu and/or northbridge is already heat damaged isn't it too late to put better cooling on them? Shouldn't I have to replace them now?
-
If your northbridge is permanently damaged (VERY unlikely), then you have no choice but to get a new motherboard for Athlon CPUs. A replacement A7N8X-e will do well.
As for the CPU, it is very likely that it is permanently damaged if heating was the cause. The extent of that damage depends on how it may have been overheated to begin with. For instance, if the processor was caught in an infinite loop, and the safety precautions did not take hold, it may never operate reliably again. Whereas if something caught the fan blade, and heat wasn't dissipated fast enough due to no airflow, and the CPU wasn't stuck, the damage may be negligible. It really all depends on the circumstance.
Anyway, before doing anything, did you remove the dual channel DIMM as I told you before? It may be something as simple as that.
-
:doh: forgot to check the DIMM but I'll do that. As for how things could have overheated, our air conditioner has been broken, it's summer in Texas, and my computer has always run a little hot.
-
Well... I just got off the phone with a professional computer repair guy who I got to investigate this. It turns out my mobo is totally buggered and my video card might be messed up as well. I suppose this might be a good excuse to upgrade though. :)
-
Hmmmmmmmm... Careful of phone repairs, he might mean the best, but unless someone is there actually hands-on looking at it, I'd be prepared to keep bashing at it for a little while.
Slightly OT, but I replaced the CPU fan a couple of days ago, only to find that so much thermal grease had been put on the CPU it had formed a suction seal which meant that as I lifted the heatsink, the CPU came with it, seeing a 3Ghz CPU stuck to the bottom of your heatsink withj a couple of bent pins is not very good for the heart. Fortunately, I managed to fix it with a wooden toothpick , some brute force and a lot of expletives ;)
-
I wasn't very clear about that. My computer is at the guy's shop and he's been checking it out. I just called him to ask what he had found out.
-
Originally posted by Flipside
Slightly OT, but I replaced the CPU fan a couple of days ago, only to find that so much thermal grease had been put on the CPU it had formed a suction seal which meant that as I lifted the heatsink, the CPU came with it, seeing a 3Ghz CPU stuck to the bottom of your heatsink withj a couple of bent pins is not very good for the heart. Fortunately, I managed to fix it with a wooden toothpick , some brute force and a lot of expletives ;)
I did that once (thankfully it was on a company machine that was due for other repairs). It actually isn't at all uncommon to have the thermal grease bind to the chip more tightly than the housing will hold on to the pins; a lot of service techs expect that to happen.
-
Ah right, understood Redsniper, sorry to hear that then :(
And it's good to know that at least I'm not the first person to lift the heatsink and think 'Wheres the damn CPU?' ;)
-
I was pondering what killed my comp the other day and I think I've got a half decent theory. Whenever I felt the air blowing out the back from the fans the air from the main part of the case felt a little warm but not too hot. The air coming from the PSU however alway felt hot. What I think happened is that with the air conditioner broken and summer temperatures the power supply overheated, malfunctioned somehow, and fried my mobo. Does that sound plausible?