PDA

View Full Version : Multiple BSOD Errors...Please Help



IFlyGC
03-28-2004, 11:08 PM
Hi folks,

First time visitor to these boards but you all come highly recommended so I'm hoping you can help me out with this issue. Firstly, the setup:

P4 1.7Ghz
Intel 850GBAL Mobo
1024 Ram (PC800)
GEForce 3 GRaphics
SoundBlaster Live Platinum 5.1
Generic Modem
Western Digital IDE Card (w/200 GB HD Attached to it)
Win XP (SP1 NOT installed)

Ok, here's the problem. I get random BSOD messages at random times. I cannot duplicate the issue although they seem to happen more frequently after the PC is warm (or I'm aggrivating it more, I really don't know). Either way, the cover is off and there is an open window blowing cold air right into it. Here are the latest 4 errors:

Stop 0x8e (0xc00000005, 0x00000002, 0xf468091c, 0x000000000)

Stop 0xd1 (0xf8425228, 0x00000002, 0x00000000, 0xf80225cf)
NDIS.SYS (Date Stamp 3b7dc561)
USBPORT.SYS (Date Stamp 3b7d868b)

Page Fault in Non-Paged Area
Stop 0x50 (0xee481212, 0x00000000, ox8060c183, 0x00000000)

Driver IRQL Not less Or Equal
Stop 0xd1 (0x0000dfae, 0x00000007, 0x00000000, 0xf859b037)
ULTRA.SYS

There is no "trigger" for these. After I restart the machine it will occasionally hang on the reboot, other times it works fine. I've double checked the drivers (all current, including the BIOS) and the memory and mobo have recently been replaced in an attempt to fix the problem. (yes, it's been going on for a while). Virus check is clean, I've also tried wiping the HD's clean and reinstalling everything with both w2k and XP....both OS's have similar issues.

Couple of other things....

Tried replacing the PSU - No change
Tried replacing the Vid Card - No Change
Took out ALL the PCI cards - Happened a lot less, but still gave problems on a clean install.

There is a 6-pack of your favourite brew for anyone who can fix this!

Thanks for your help!

~Fly~

Whyzman
03-28-2004, 11:41 PM
<marquee behavior="alternate" scrollamount="20" loop="3" width="50%">Hello IFlyGC,</marquee>

<marquee behavior="slide" scrollamount="20" loop="3" width="50%">Welcome tohttp://www.pcguide.com/ubb/pcgubb.gif Forums!</marquee>

My first suggestion would be to try a RAM tester:

http://www.memtest86.com/

You might also want to try one stick of RAM at a time if nothing shows up with the tester...

I would download SP1 and put it in the works...

Quantax
03-29-2004, 01:46 AM
So when you go into the Device Manager and look next to all devices, there's no yellow exclamation or question marks?

IFlyGC
03-29-2004, 07:44 AM
I ran MemTest on the old memory and it gave hundreds of errors (hence why I replaced the memory) and with factory fresh Samsung memory (recommended by Intel) it gave just as many errors....hence why I replaced the mobo.

As for SP1, I read on other forums that SP1 causes more problems than it helps...is this not true? There seemed like to many people complaining about it for it to be not true, ya know?

As for device manager, there are no flags. There was one thing flagged a few months ago (a USB TV Tuner) which has been unplugged and the driver uninstalled. It didn't change the situation.

As for trying one stick of Ram, my mobo manual says that PC800 needs to be used in matched pairs. I'll gladly go try it but I'm not sure it'll work. I'll go check and report back.

Thanks for the help!

~Fly~

Whyzman
03-29-2004, 08:51 AM
Originally posted by IFlyGC
factory fresh Samsung memory (recommended by Intel) it gave just as many errors....hence why I replaced the mobo.And??? Are you still getting errors using the RAM tester??

Troubleshooting your problem needs to be addressed on two different levels (i.e. Hardware and Software). If you're still getting RAM errors it doesn't make any sense to go any further until the problem is resolved...issues at this level are long before software gets involved...

I would also suggest running the Harddrive diagnostics...let's take that out of the hardware equation also...

As far as the RAM goes, I think you need two matched for speed purposes, but one bank at a time will work for troubleshooting with the RAM tester (somebody please jump in and correct if I'm in error here! ;) )

As far as when we get the troubleshooting focused on Windoze...Critical Updates are necessary for more reasons than just security holes...they are necessary for programming flaws also...

IFlyGC
03-29-2004, 09:16 AM
Ok, I just tried your version of the Ram tester. Mine was a windows version. I passed with flying colours on the first 2 complete passes. I'm now going to let the beast warm up and see if I can make it crash, then re-run the test and see if it IS a heat issue.

As for the HD's, they are fine...no problems found there.

~Fly~

Whyzman
03-29-2004, 09:31 AM
Originally posted by IFlyGC
As for the HD's, they are fine...no problems found there.And??? The tests were run using the manufacturer's diagnostics (these are run in DOS), not a Windoze software program??

ski
03-29-2004, 09:45 AM
Agree with Whyzman. The RAM problem has to be resolved if memtest is still showing errors.
If it is, then also make sure that only Default settings are selected for the RAM in BIOS.

IFlyGC
03-29-2004, 09:58 AM
Whyzman....you really don't like my windoze utilities, do you? ;)

Yes, I re-did the test using the manufacturer's diag. program and all 3 came up clean as a whistle!

So, as it stands right now.... both the memory and the HD's are free of defects and I'm waiting for it to crash again so I can run the memory check again with a warm PC.

~Fly~

Whyzman
03-29-2004, 10:21 AM
Originally posted by IFlyGC
Whyzman....you really don't like my windoze utilities, do you?Well...It's nothing personal mind you! However, it just goes back to troubleshooting basics. You want to remove variables and isolate.

Windoze is the epitome of a variable because it contains a host of variables, and could be "hosting" variables (i.e., viri, trojans, worms) all of which could mess with any of the utilities and produce tainted results.

DOS utilities come in "under the radar" of Windoze ;)

BTW...what kind of temps are you getting in the BIOS??? I would let it idle and monitor from there and see if you get any serious fluctuations...

IFlyGC
03-29-2004, 10:45 AM
Well, it just crashed again. Well, twice....the second time while I was trying to post this message! The first time blue-screened with:

Driver IRQL Not Less Or Equal
STOP 0xD1 0x00000329, 0x00000005, 0x00000000, 0xf826bbe
NV4_MINI.SYS - Address f826b13e, Base At f814a000, Date 3f820145

The second time just hung.

There is no temp reading directly from the BIOS, but Intel provides a monitor from windows (sorry, I don't have a DOS temp gauge) and the temps are:

CPU - 43C
System Area 1 - 30C (I believe this is near the center of the mobo)
System Area 2 - 38C (I believe this is near the RAM)
Fan Speed 3413RPM

There are 2 more fans, but they do not give RPM settings. They ARE spinning however.

~Fly~

Whyzman
03-29-2004, 12:59 PM
It appears that you have been on line with this sans SP1...is that correct??

I would suggest running HiJackThis (in my signature below) and paste the log in this thread...

Do so after saving the program to its own folder...that way any alterations can be undone if needs be

Your temps are well within normal readings...

Before you run the HiJackThis...you might want to see if you get any crashes in SAFEMODE...

Also, look in DEVICE MANAGER whilst there and see if there might be any anomalies !, or ? there. Look for double entries of devices...if you find any, REMOVE "all" entries and then reboot allowing Windoze to find and reinstall. These are known as "Ghost" drivers...don't attempt to guess which one is legit...no way to tell!

Double entries won't show up in normal Windoze...only SAFEMODE...

After playing with that for a while ;) I would run HiJackThis...

ski, do you have that troubleshooting link handy using selective startup in case he needs to go there once we've looked at HJT??

IFlyGC
03-29-2004, 01:27 PM
Yes, I am still running without adding SP1 to the mix. I tried updating earlier but it BS'd before it even started the install process.

The machine will crash in Safe Mode just like regular mode. The only thing I can reliably reproduce is the more I use the machine, the more it crashes...hence why I was thinking it's a heat issue. It BS'd just as it finished saving the HJT log but it gave no fault name or file, just a stop code:

STOP 0x23 0x00080110, 0xfffff000, 0x00001000, 0x00000000

Here is the HJT log whilst in safe mode.

Logfile of HijackThis v1.97.7
Scan saved at 1:11:57 PM, on 3/29/2004
Platform: Windows XP (WinNT 5.01.2600)
MSIE: Internet Explorer v6.00 (6.00.2600.0000)

Running processes:
C:\WINDOWS\System32\smss.exe
C:\WINDOWS\system32\winlogon.exe
C:\WINDOWS\system32\services.exe
C:\WINDOWS\system32\lsass.exe
C:\WINDOWS\system32\svchost.exe
C:\WINDOWS\System32\svchost.exe
C:\WINDOWS\Explorer.EXE
C:\Documents and Settings\Dave\Desktop\high\HijackThis.exe

O2 - BHO: (no name) - {06849E9F-C8D7-4D59-B87D-784B7D6BE0B3} - C:\Program Files\Adobe\Acrobat 5.0\Acrobat\ActiveX\AcroIEHelper.ocx
O3 - Toolbar: &Radio - {8E718888-423F-11D2-876E-00A0C9082467} - C:\WINDOWS\System32\msdxm.ocx
O4 - HKLM\..\Run: [IMONTRAY] C:\Program Files\Intel\Intel(R) Active Monitor\imontray.exe
O4 - HKLM\..\Run: [NvCplDaemon] RUNDLL32.EXE C:\WINDOWS\System32\NvCpl.dll,NvStartup
O4 - HKLM\..\Run: [nwiz] nwiz.exe /install
O4 - HKLM\..\Run: [HP SchedIndexer] C:\Program Files\Hewlett-Packard\LaserJet 33xx\hppschedindexer.exe
O4 - HKLM\..\Run: [HP AutoIndexer] C:\Program Files\Hewlett-Packard\LaserJet 33xx\hppautoindexer.exe
O4 - HKLM\..\Run: [LVCOMS] C:\Program Files\Common Files\Logitech\QCDriver3\LVCOMS.EXE
O4 - HKLM\..\Run: [LogitechGalleryRepair] C:\Program Files\Logitech\ImageStudio\ISStart.exe
O4 - HKLM\..\Run: [LogitechImageStudioTray] C:\Program Files\Logitech\ImageStudio\LogiTray.exe
O4 - HKLM\..\Run: [KernelFaultCheck] %systemroot%\system32\dumprep 0 -k
O4 - HKLM\..\Run: [AtiPTA] atiptaxx.exe
O4 - HKLM\..\Run: [Mirabilis ICQ] C:\PROGRA~1\ICQ\ICQNet.exe
O4 - HKLM\..\Run: [QuickTime Task] "C:\Program Files\QuickTime\qttask.exe" -atboottime
O4 - HKLM\..\Run: [WINDVDPatch] CTHELPER.EXE
O4 - HKLM\..\Run: [UpdReg] C:\WINDOWS\UpdReg.EXE
O4 - HKLM\..\Run: [Jet Detection] "C:\Program Files\Creative\SBLive\PROGRAM\ADGJDet.exe"
O4 - HKCU\..\Run: [MSMSGS] "C:\Program Files\Messenger\msmsgs.exe" /background
O4 - HKCU\..\Run: [NvMediaCenter] RUNDLL32.EXE C:\WINDOWS\System32\NVMCTRAY.DLL,NvTaskbarInit
O4 - HKCU\..\Run: [Yahoo! Pager] C:\Program Files\Yahoo!\Messenger\ypager.exe -quiet
O4 - Global Startup: HP LaserJet Director.lnk = C:\Program Files\Hewlett-Packard\LaserJet 33xx\hppdirector.exe
O8 - Extra context menu item: E&xport to Microsoft Excel - res://C:\PROGRA~1\MICROS~2\Office10\EXCEL.EXE/3000
O9 - Extra button: ATI TV (HKLM)
O9 - Extra button: Messenger (HKLM)
O9 - Extra 'Tools' menuitem: Yahoo! Messenger (HKLM)
O9 - Extra button: ICQ Pro (HKLM)
O9 - Extra 'Tools' menuitem: ICQ (HKLM)
O9 - Extra button: AIM (HKLM)
O9 - Extra button: Related (HKLM)
O9 - Extra 'Tools' menuitem: Show &Related Links (HKLM)
O9 - Extra button: Messenger (HKLM)
O9 - Extra 'Tools' menuitem: Windows Messenger (HKLM)
O12 - Plugin for .spop: C:\Program Files\Internet Explorer\Plugins\NPDocBox.dll
O15 - Trusted Zone: *.smartforce.com
O16 - DPF: {02BF25D5-8C17-4B23-BC80-D3488ABDDC6B} (QuickTime Object) - http://www.apple.com/qtactivex/qtplugin.cab
O16 - DPF: {2C52AF58-B9B1-11D5-9DF6-00508B755B44} (AXClientUtil2 Control) - http://www.smartforce.com/v2.1/applications/liveplay/Activex/AXClientUtil.cab
O16 - DPF: {30528230-99F7-4BB4-88D8-FA1D4F56A2AB} (YInstStarter Class) - http://download.yahoo.com/dl/installs/yinst0309.cab
O16 - DPF: {62475759-9E84-458E-A1AB-5D2C442ADFDE} - http://a1540.g.akamai.net/7/1540/52/20031216/qtinstall.info.apple.com/mickey/us/win/QuickTimeInstaller.exe
O16 - DPF: {D27CDB6E-AE6D-11CF-96B8-444553540000} (Shockwave Flash Object) - http://download.macromedia.com/pub/shockwave/cabs/flash/swflash.cab

I also checked the device list for ghosts and there were none shown. It looked just fine with no repeat devices. I did note a couple of devices were sharing IRQ's (17 & 21 I think) but from what I recall, XP assigns them automatically doesn't it?

It's days like this I want to go back to using DOS!

Thanks for your continuing help Whyzman!

~Fly~

ski
03-29-2004, 03:16 PM
Here's the link for doing a clean boot with XP:
http://support.microsoft.com/default.aspx?scid=kb;en-us;310353
But if the system is also crashing in Safe mode, then it appears that the problem is being caused by faulty hardware, rather than by a corrupt or missing driver, and a clean boot may not help in troubleshooting it. But I could be wrong(it will not be the 1st time).

IFlyGC's initial post said that removing the PCI cards reduced the number of errors. I'm wondering if one or more of them is either bad or incompatible with XP, or if the MB is faulty, or if the MB's BIOS is the following version:
http://downloadfinder.intel.com/scripts-df/filter_results.asp?strOSs=45&strTypes=BIO%2CDRV%2CUTL&ProductID=486&OSFullName=Windows*+XP+Home+Edition&submit=Go%21


I also found the following info on Google:

0x000000D1: DRIVER_IRQL_NOT_LESS_OR_EQUAL
The system attempted to access pageable memory using a kernel process IRQL that was too high. The most typical cause is a bad device driver (one that uses improper addresses). It can also be caused by caused by faulty or mismatched RAM, or a damaged pagefile.

0x00000050: PAGE_FAULT_IN_NONPAGED_AREA
Requested data was not in memory. An invalid system memory address was referenced. Defective memory (including main memory, L2 RAM cache, video RAM) or incompatible software (including remote control and antivirus software) might cause this Stop message, as may other hardware problems (e.g., incorrect SCSI termination or a flawed PCI card).

Since the 0xD1 errors were the most frequent ones, I focused on them.
Each of the 3 0xD1 errors had 4 different drivers named(ndis.sys, usbport.sys, ultra.sys, and nv4_mini.sys).
ndis.sys is a network card driver, usbport.sys is a PCI USB card driver, ultra.sys is a PCI IDE card driver, and nv4_mini.sys is a video card driver.
Is it possible that all 4 devices are bad, or all of their drivers are corrupt or incompatible with XP, or the MB or RAM is bad, or the pagefile is damaged?

Also, the 0x50 error message description points to bad RAM or a bad PCI card.

IFlyGC
03-29-2004, 03:41 PM
Ski,

I don't know if this narrows the problem down further but I did notice that occasionally on a reboot it will fail and the screen will show random coloured blocks, sometimes with pretty symbols, sometimes flashing.

If I leave the system alone (like it has been for the past 2 hours) it will quite happily idle and be just fine. It's when I start using it crashes in the variety of flavours we've seen above.

As for items being incompatible, every device installed has a manafactuers driver specifically for XP....am I safe in assuming that makes it compatible?

As for the 0xD1's...the USB "card" and Network "card" are on the mobo. The cards I have attached are

Generic Modem (CompUSA cheapie)
Western Digital IDE Driver w/200GB HD
Soundblaster Live Platinum 5.1
NVidia GEForce 3

They are the only real "cards" attached to the system. All USB devices have been disconnected incase that's the trigger.

I appreciate you joining the "battle"!

~Fly~

Whyzman
03-29-2004, 07:23 PM
One possible course of action would be to strip back to barebones and begin adding back one component at a time...barebones would be video, RAM, Keyboard and Mouse...

Have you tried using different PCI slots for the cards??

If it's crashing in SAFEMODE, then as ski pointed out it is more likely hardware related...

Heat could indeed be an issue; faulty interface between processor and heatsink (however, your temps appear normal); could be Power Unit...

You might want to have a closer look at the motherboard and check for expanded or leaking capacitors...

IFlyGC
03-29-2004, 07:58 PM
I've tried different slots for the PCI cards but now that I think about it, I never cleaned the contacts so I'll go ahead and try that.

As for the heat sink, I'm using the one provided by Intel with the CPU and I refreshed the grease when I replaced the mobo (with Intel provided grease).

I have tommorow off work as well. I'll bite the bullet, re-format the C drive and do a fresh install with barebones stuff. I'm going to go back to w2k as I really don't like XP all that much.

Thanks for all your help and I'll let you know how it goes tommorow!

~Fly~

IFlyGC
03-29-2004, 10:00 PM
Hmmmmm...this is wierd.

I just tried to shutdown the machine and well, it didn't. You remember how Win 98 used to display the "windows is shutting down" even after it was shutdown? My XP just did it. The HD's powered down but the fans kept running and the screen stayed on. I KNOW it used to powerdown completely by itself.

Is this connected in anyway?

~Fly~

Paleo Pete
03-30-2004, 01:21 AM
[quoote]occasionally on a reboot it will fail and the screen will show random coloured blocks, sometimes with pretty symbols, sometimes flashing[/quote]

This sounds like either a virus or bad video memory. I don't see anything that resembles an antivirus in your Hijack This log either. Also looks like a pretty short list of running processes too, especially for XP with ICQ, Yahoo Messenger, AIM, Office...

Run a virus scan on that thing, the odd colors and symbols at bootup sound pretty suspicious. Try Housecall (http://www.housecall.antivirus.com/) it seems to work pretty well, and definitely get a good antivirus installed. Soon.

IFlyGC
03-30-2004, 07:57 AM
Just an update...

Left the machine running overnight with just the barebones installed and a ran a CPU burn to keep the machine active. Yep, it crashed, although this time the screen just went black and wouldn't return.

One other thing I've noticed, when the screen goes black, or it fails on start-up or anything else like that, I can press the num-lock key off/on 6 times and then it will quit responding. Is 6 keystrokes the buffer for the keyboard?

As for A/V, I had an A/V installed but removed that along with the firewall and some other stuff when all these problems started.

~Fly~

ski
03-30-2004, 09:31 AM
If it crashed with just a barebones configuration as Whyzman had suggested, a virus check as Pete had suggested turns up empty, and a repartition(yes, do this also to remove any virus from the HD's boot sector) and reformat does not fix things as you had suggested, then start swapping out components or have them shop tested to find the bad one(or ones).

IFlyGC
03-30-2004, 11:30 AM
Ok, the machine is clean....no viruses found anywhere (including the boot sector/MBR). I think however I may have found the problem. The lid of the case is off, it is right below an open window with cold air blowing into the room, yet as soon as it crashed I reached in and had a feel to see what the temps really were. The memory was too hot to touch yet Intel's fine little utility was showing a fairly warm 40 degrees.

Now my question for you is this....what would make the memory get THAT hot? The CPU was cold, the vid. card was cold, but you could fry an egg on the Ram. Any suggestions?

~Fly~

ski
03-30-2004, 12:22 PM
RAM is overclocked, running at a voltage that's higher than design, not fully inserted, dirty contacts, bad RAM, or bad slots.

Also, instead of placing the computer below the open window, run an external fan on it, and see if the RAM runs cooler.

IFlyGC
03-30-2004, 02:13 PM
Ski,

The computer isn't overclocked and the RAM is on the Intel recommended list. I just checked the temp with a thermometer and the temp of the heatshield on the RAM was 54 degrees C. Does this sound OK? I'm assuming if the shield is 54 then the ram inside would be hotter, but I really wouldn't know.

I'm off to walmart to get a fan....I'll let you know if it works!

~Fly~

ski
03-30-2004, 02:59 PM
54 deg. C for your RAM's heat sink is a little toasty, especially if the system had been running at idle. And since heat flows like pressure, i.e., from a high point to a low point, then the RAM temp is even higher than 54 deg. C.
And when your RAM's temp is compared to the temp of a properly cooled CPU(less than 50 deg. C) on a system that's running at full load, you can see why this is not good.

What are your RAM's voltage and frequency settings in BIOS?

IFlyGC
03-30-2004, 04:11 PM
In there BIOS there is no memory settings with the exception of turning on or off ECC (which I have set to off). The Intel Active Monitor shows the recommended voltage at 2.5v, with a rare fluctuation to 2.487v....well within limits.

I'm currently running a big ole fan over the CPU and RAM and putting the machine through it's paces to try and overheat it again. I'll let you know the results.

~Fly~

mjc
03-30-2004, 04:16 PM
Yeah, I would say 129 deg F is a little toasty............

ski
03-30-2004, 05:12 PM
Another thing to check is if the PS is the correct one for the MB.
The following are the specs for the PS from the Intel site:

The ATX12V power supply required for the Intel Desktop Board D850GB has the standard 20-pin ATX Main Power Connector, a 4-pin +12V and a 6-pin Aux. Power Connector.
The new 4-pin +12V power connector has been added to enable the delivery of more +12VDC current to the desktop board. This connector is used by the board to power the processor's voltage regulator module (VRM).
The 6-pin Aux. Power connector provides additional current to meet the board's +3.3VDC and +5VDC requirements.


Also, what is the watt rating of the PS?

IFlyGC
03-30-2004, 05:39 PM
It's a P4 approved power supply rated at 400W (3 HD's, 3PCI, 1AGP, CDRW, DVD, Floppy, 4 fans). I've had a 10" fan blowing on the RAM/CPU for the last hour and I've been trying to render 250MB photoshop files while playing mp3's and surfing the net. (not my usual behaviour, but I'm trying to load the ole girl up, ya know?). So far it's been fine and the highest temp I'm getting is 43 on the CPU and 35 on the RAM....much lower!

I'm going to leave it crunching numbers for a while and see what she does!

~Fly~

ski
03-30-2004, 06:02 PM
If that works, then replace the computer's cover and see if the system's existing internal fans provide sufficient cooling.

If the problem resurfaces, then it appears that the system requires additional cooling. I believe that the MB has a terminal for a memory fan. And an add'l case fan or 2 may also be needed(recommended locations are upper rear and lower front of the case).

IFlyGC
03-30-2004, 06:17 PM
I'm hoping it IS the heat issue.....it's still crunching away quite nicely. My concern is where to attach the new fans. I have 2 more outlets on the mobo to plug fans in, but nowhere to mount them.

I'm using a 4U rack mount case for this machine and have a 4" fan (intake) on the front panel and a 3" fan (exhaust) on the rear. I'm thinking maybe cutting a hole in the top panel of the case and mounting a fan on that (intake) unless you can think of anoy other ideas?

I've seen a RAM cooler out there but there isn't enough space to mount it because the back of the CD/DVD drives go over the top of the memory.

Any ideas?

~Fly~

mjc
03-30-2004, 07:02 PM
there isn't enough space to mount it because the back of the CD/DVD drives go over the top of the memory.

I think you just may have answered your problem...the drives are most likely preventing/hindering airflow around the RAM.

If possible (I am assuming that the drives will be toward the one side of the case) to mount the fan in the side panel and direct it towards the RAM.

IFlyGC
03-30-2004, 07:02 PM
Ski, Whyzman, I want to thank you both for your help and input, and let you know it appears the problem is solved! :)

I let the machine run for 2.5 hours continuously at 100%CPU usage and using as much memory as possible and it ran flawlessly. I took the extra fan away, replaced the lid and it crashed within 90 seconds. Sounds like the problem doesn't it?

As for the temps, heres the results:

Cold System: CPU-31, Mobo-31, RAM-32
Loaded System w/Fan: CPU-46, Mobo-36, RAM-36
Loaded System w/o Fan :CPU-50, Mobo-36, RAM-41

Thank you!!! I've been battling this machine since September and it's great to have it working properly again.

Take care,
~Fly~

Whyzman
03-30-2004, 08:46 PM
Way Kewl!! :cool: A story with a happy ending! :)

Until a couple of minutes ago I had not explored the innards of a Rackmount case...I found an Antec.

Hindsight being 20/20 I would think we might have zeroed in a bit sooner on the heat problem knowing the case you were using. ;)

From what I could see in the photo it certainly looks like airflow could definitely be an issue. Is there actually some sort of metal devider that sits in front of the PCI cards, behind the drives??

It looked as though it was not completely restricting from side to side but pretty darn near...Here's a solution worthy of note!

http://www.overclockers.com/tips1092/

IFlyGC
03-30-2004, 09:16 PM
Whyzman,

I never even brought the case into the equation. The lid has been off of it from day one...if an 19"x23" hole in the roof wasn't enough then I didn't know what would be.

This is the case I'm using:

http://www.plinkusa.net/web4043x.htm

Thanks again for your help!

~Fly~

Whyzman
03-30-2004, 09:29 PM
There have been ongoing discussions here on correct fan placement etc., and your situation truly emphasises the need for correct air "movement" across components...

What I gather from this is perhaps similar to the "inadequacy" of a passive heatsink for the new CPUs (i.e., the necessity to add a fan in close proximity)!

Even though your case was open on top there was not enough immediate air movement across the RAM to move the heat away...

If this was going on now...I wonder what would have been occurring during the summer with ambient room temps higher?? :(

mjc
03-30-2004, 11:24 PM
Basically, I think the drives were acting as a secondary "lid" and keeping the air around the RAM static.

IFlyGC
03-31-2004, 07:15 AM
Basically, I think the drives were acting as a secondary "lid" and keeping the air around the RAM static.

At least that's better than them keeping static around the RAM! ;)

~Fly~

ski
03-31-2004, 08:16 AM
You're welcome.
Also glad to hear that you found the gremlin.

If you cannot jury rig a RAM cooling setup in the existing case(airflow path inside it reminds me of the Holland tunnel at rush hour), then you can always replace it with a different case that has more efficient ventilation. It's cheap insurance to prevent frying the RAM.