Help! Impossible hardware problem

The magic smoke.

Moderators: phlip, Moderators General, Prelates

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Help! Impossible hardware problem

Postby ryak2002 » Thu Apr 07, 2011 12:42 am UTC

I've had this hardware problem for over a month now and no-one can figure out how to proceed further on troubleshooting it. Basically it will randomly reboot itself while it is running and the frequency of the reboots started out very rare but it has reached the point where it happens within 5 minutes of boot-up.

Steps i have tried so far:
(1) Obvious troubleshooting steps - reseated the memory, etc.
(2) Wiped the hard drive and reinstalled windows (several times)
(2) Bought a new power supply
(3) Took it to a professional place who tested and told me that the motherboard was failing.
(4) RMA'd the motherboard and i got back what was obviously a different board, but after putting everything together it is still having the same problem.

Symptoms: The random crash i said. There is always an error report associated with the same message:

Component: AMD Northbridge
Error Source: Machine Check Exception
Error Type: 11
Processor ID: 0

<System>
<Provider Name="Microsoft-Windows-WHEA-Logger" Guid="{C26C4F3C-3F66-4E99-8F8A-39405CFED220}" />
<EventID>20</EventID>
<Version>0</Version>
<Level>2</Level>
<Task>0</Task>
<Opcode>0</Opcode>
<Keywords>0x8000000000000000</Keywords>
<TimeCreated SystemTime="2011-03-09T05:43:19.614381800Z" />
<EventRecordID>140205</EventRecordID>
<Correlation ActivityID="{68376C21-46EA-4EAD-BD73-667A01794BBF}" />
<Execution ProcessID="1696" ThreadID="2724" />
<Channel>System</Channel>
<Computer>Ryan-PC</Computer>
<Security UserID="S-1-5-19" />
</System>
- <EventData>
<Data Name="ErrorSource">3</Data>
<Data Name="ApicId">0</Data>
<Data Name="MCABank">4</Data>
<Data Name="MciStat">0xfa000010000b0c0f</Data>
<Data Name="MciAddr">0x0</Data>
<Data Name="MciMisc">0xc00a0fff01000000</Data>
<Data Name="ErrorType">11</Data>
<Data Name="Length">928</Data>
<Data Name="RawData">435045521002FFFFFFFF03000100000002000000A0030000322A05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</Data>
</EventData>
</Event>


If i take out 3 of the 4 ram sticks, the problem goes away, but it doesn't seem to matter which ram stick i leave in, so it doesn't look like bad RAM.

System specs:
Motherboard: ASUS M4A77-TD (RMA'd, so presumably good)
Memory: 8 Gig (4 x 2Gig) OCZ PC3-12800 Gold RAM - 4GB, 1600MHz, DDR3
CPU: AMD Phenom II X4 965 Black Edition Quad Core Processor - 3.40GHz
Cooling: fans
PSU: Corsair 650W (new)
GPU: Sapphire Radeon HD 5770 Video Card
OS: Windows 7 Ultimate 64

User avatar
Solt
Posts: 1912
Joined: Tue Mar 27, 2007 5:08 am UTC
Location: California

Re: Help! Impossible hardware problem

Postby Solt » Thu Apr 07, 2011 6:20 am UTC

Willing to bet it's your CPU.

You say it's related to RAM and it says right there that it's caused by the Northbridge. The Northbridge is the memory controller and it's located on the CPU: http://en.wikipedia.org/wiki/Northbridg ... mputing%29

What do I win?


Edit: A hunch. Wiki says northbridge plays a role in overclocking. The spec page for your motherboard says 1600(O.C.) for supported RAM speeds. What does the O.C. mean? Can all this be fixed with a setting? Highly suggest you read through your motherboard manual before replacing the CPU.

http://www.asus.com/product.aspx?P_ID=0AvsBb7WBZe2i9zK

Edit2: O.C. is probably a red herring. Make sure you have the latest BIOS. Follow the link for your processor under the 'CPU Support List' tab (yes, their BIOS updates are processor specific, which should be a promising sign!) Note that there are two versions of your processor, make sure the entire model number matches.

Also download all the latest drivers under the "Download" tab.
"Welding was faster, cheaper and, in theory,
produced a more reliable product. But sailors do
not float on theory, and the welded tankers had a
most annoying habit of splitting in two."
-J.W. Morris

Game_boy
Posts: 1314
Joined: Tue Mar 04, 2008 7:33 pm UTC

Re: Help! Impossible hardware problem

Postby Game_boy » Thu Apr 07, 2011 1:09 pm UTC

Confusingly, there is also a northbridge on the motherboard.

CPUs hardly ever fail unless you're overclocking, I'd say it's more likely to still be the MB or RAM than the CPU.

You've done the right things so far, it's bad luck that it hasn't been fixed yet.
The Reaper wrote:Evolution is a really really really long run-on sentence.

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Thu Apr 07, 2011 3:03 pm UTC

Update:

Bios was the latest on the motherboard i sent in. I'm not sure about this one they sent back to me, but it didn't fix it before, so it probably wouldn't have any effect assuming it isn't already the latest on this new one.
The CPU is not overclocked.
I ran an entire cycle of MemTest86+ with all 4 memory sticks in and it reported no errors in my memory. I just ran one test which took an hour. Does this definitively clear the memory or should i run the test additional times?

I know i've been following the right steps but i'm running out of ideas now. What else could possibly be wrong that I haven't already eliminated? :shock:

Ram - Tested (passed)
Motherboard - RMA'd
Power Supply - new

User avatar
Kromix
Posts: 288
Joined: Thu Mar 25, 2010 7:26 pm UTC
Location: DFW, Great State of Texas

Re: Help! Impossible hardware problem

Postby Kromix » Thu Apr 07, 2011 6:58 pm UTC

I also Motion to CPU,

Nowadays CPUs (AMD and Sandy bridge intel) talk directly to the RAM bypassing Northbridge.
(but Windows doesnt know that :) )

Cooling: fans <- What kind o fans? (the one the CPU Heatsink)


Spoiler:
Image
Image
Click Me! <---- Click There! :)
Spoiler:
Image

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Fri Apr 08, 2011 6:27 am UTC

Have you tried an open air test? Bring the everything out of the case and put it on the anti-static wrap. Boot it up and see how things work.

Also, what are your temperatures?

Moose Hole
Posts: 398
Joined: Fri Jul 09, 2010 1:34 pm UTC

Re: Help! Impossible hardware problem

Postby Moose Hole » Fri Apr 08, 2011 4:58 pm UTC

Did you try turning it off and on?

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Fri Apr 08, 2011 5:36 pm UTC

It's almost certainly not a heat problem. I've already tried running the PC with both sides off the case which is pretty much open air. It didn't have any effect. I've also tried running speedfan to monitor the core temperature and it never gets even remotely hot enough to indicate an overheating problem.

The cooling system for the CPU is the stock AMD one. I'm aware that the stock fan/heat sink isn't very good but it certainly shouldn't be so bad it causes overheating problems.

I'm probably going to get a better cooling setup but i'd like to make sure the computer is working before replacing something like that.

User avatar
mosc
Doesn't care what you think.
Posts: 5403
Joined: Fri May 11, 2007 3:03 pm UTC

Re: Help! Impossible hardware problem

Postby mosc » Fri Apr 08, 2011 7:40 pm UTC

memory controllers in this type of system often have trouble with multiple sticks in the same channel. When you put multiple memory sticks on the same channel, the latency of one stick vs the other constantly puts strain on the memory controller. Commercial systems generally use ECC memory, which among other things has more standardization of memory latency and is not "over-spec'd" to speeds that it thermally might be able to hit, but with latency issues. These latency issues are not show stoppers in any setup where the sticks are on their own channel, but you're talking about 4 sticks.

I'd say replace the memory with a 2 stick solution (if this is a dual channel board which I suspect it is) rather than 4. It's not really your memory's fault, nor is it the CPU, nor it's the motherboard. It's more like all three of them together not getting along.

The CPU's memory controller is unlikely to be the issue. It's possible, but this would be hard to brake on it's own without more damage to the chip. If you can't run a dual channel setup easily, I'd blame the motherboard. Could be too much crosstalk due to improper groundplane density between the memory channels. Just because the memory controller has been integrated into the CPU doesn't mean the motherboard still doesn't have the physical connections between the memory DIMM and the controller.
Last edited by mosc on Fri Apr 08, 2011 7:43 pm UTC, edited 1 time in total.
Title: It was given by the XKCD moderators to me because they didn't care what I thought (I made some rantings, etc). I care what YOU think, the joke is forums.xkcd doesn't care what I think.

GeorgeH
Posts: 527
Joined: Mon Aug 17, 2009 6:36 am UTC

Re: Help! Impossible hardware problem

Postby GeorgeH » Fri Apr 08, 2011 7:43 pm UTC

ryak2002 wrote:I just ran one test which took an hour. Does this definitively clear the memory or should i run the test additional times?

Memtest is a more advanced memory checker than built-in BIOS checkers, but a pass doesn't mean everything is fine; a pass only means there's nothing obviously wrong. Running it more than once is generally useless.

ryak2002 wrote:If i take out 3 of the 4 ram sticks, the problem goes away, but it doesn't seem to matter which ram stick i leave in, so it doesn't look like bad RAM.

Memory: 8 Gig (4 x 2Gig) OCZ PC3-12800 Gold RAM - 4GB, 1600MHz, DDR3

I'd bet that's your problem. AMD CPUs can run 2 sticks of RAM faster than 4. IIRC they also can run out of dual-channel mode (i.e. an odd number of sticks) faster than dual channel mode. I believe that 1600MHz is beyond rated spec for 4 sticks in dual channel, so if your BIOS is setting your timings automagically using the RAM's SPD data it's essentially overclocking your PC beyond stability.

I'd loosen your timings a bit and see what happens. I'd tell you what speeds to shoot for, but my hurried Google-fu is failing to quickly find me AMD's spec. Someone else might be able to find it for you, and/or OCZ might be able to give you timings to try.

User avatar
mosc
Doesn't care what you think.
Posts: 5403
Joined: Fri May 11, 2007 3:03 pm UTC

Re: Help! Impossible hardware problem

Postby mosc » Fri Apr 08, 2011 7:45 pm UTC

I'd also agree to try tuning down your memory. OCZ often over-spec's it's memory. They'll sell you DDR1066 that they were able to overclock to DDR1600 as DDR1600. Latency though, will bite you in the ass.
Title: It was given by the XKCD moderators to me because they didn't care what I thought (I made some rantings, etc). I care what YOU think, the joke is forums.xkcd doesn't care what I think.

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Fri Apr 08, 2011 9:35 pm UTC

The RAM will not run at 1600 MHz if there's no overclocking. By default it's probably running at 1333 or 1066 MHz.

As for the heating issue, if you forgot to apply the thermal paste or if it's applied incorrectly then it won't matter if the case is open. If you're using the stock cooler then I'm assuming you're also using the stock thermal paste but there's only enough thermal paste for one application and you must have needed two since you had to install the chip in a new motherboard.

Also the point of an open air test is to ensure that the the case isn't shorting out your motherboard.

User avatar
naschilling
Posts: 142
Joined: Wed Apr 06, 2011 2:52 pm UTC
Contact:

Re: Help! Impossible hardware problem

Postby naschilling » Sat Apr 09, 2011 2:16 pm UTC

In about 20 years of working on computer hardware, I've seen ONE (1) processor that died of "natural" causes. I've seen plenty that have had broken pins or were cracked (I still don't know how they did that), but only one that looked like it should work and didn't. Most often, it is an issue with memory or the interface to the memory.

When using dual-channel memory, if one stick deviates from the specs of the others even slightly, it messes up the whole system, often exactly how you describe. If your computer works fine with only one memory chip installed, you should try every combination of two chips to find a stable configuration. If none exists, I would suspect your memory controller, which on modern processors, is on your processor. My gut still says that one or two of your memory chips don't meet their claimed performance.
If you don't have walls, why would you need Windows?

User avatar
Solt
Posts: 1912
Joined: Tue Mar 27, 2007 5:08 am UTC
Location: California

Re: Help! Impossible hardware problem

Postby Solt » Sun Apr 10, 2011 4:17 am UTC

GeorgeH wrote:I believe that 1600MHz is beyond rated spec for 4 sticks in dual channel, so if your BIOS is setting your timings automagically using the RAM's SPD data it's essentially overclocking your PC beyond stability.


Huh. So that's what the 1600(O.C.) on the motherboard specs means?
"Welding was faster, cheaper and, in theory,

produced a more reliable product. But sailors do

not float on theory, and the welded tankers had a

most annoying habit of splitting in two."

-J.W. Morris

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Mon Apr 11, 2011 9:07 pm UTC

What you're saying sounds reasonable. So far i have tried with 2,3, and 4 sticks of RAM in and seen the problem but if what you are saying is correct, i suspect i was putting them in the slots in such a way that it would still cause a problem. I was putting them in (stick) (stick) | (empty) (empty) when i tried 2 sticks.

You're suggesting i should try one memory stick in each pair? (stick) (empty) | (stick) (empty) ? I can try that when i get home.

Reducing the clock rate on the RAM is something i already tried. It did seem to make the problem occur less often, but it didn't make it go away, even when i turned it down to the lowest possible setting.

EDIT: That didn't help. I still have the problem whether i put 2 sticks on the same channel or one in each channel. Also upon further review of my motherboard book, it appears I was putting one in each channel in my original tests since the channels are as follows: (A2)(B2) | (A1)(B1)

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Mon Apr 11, 2011 10:41 pm UTC

RAM speed should not have been the issue. The motherboard auto-configures it to either 1066 or 1333. Unless there was overclocking, the RAM should not have been running at 1600.

Modern computer hardware have all sorts of features built into them to prevent the kinds of problems that computers used to have.

Most problems with newly built computers today is either due to defective hardware, loose connections, inappropriately applied thermal paste, the case shorting out the motherboard likely due to motherboard standoffs not being used, and RAM not being correctly seated.

It is likely that the RAM is bad since the brand you have doesn't have a very good reputation. If you don't think the RAM is the problem then as I said before, it could be either your temperature or the case shorting out the motherboard. The maximum temperature of a Phenom II is around 60º C. If the temperature approaches that during load then that's the problem. The case could be shorting out the motherboard if you did not use all the motherboard standoffs.

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Tue Apr 12, 2011 1:53 am UTC

I think I'm going to try that open air test next as well as seeing if i can borrow some ddr3 ram from a friend to see if it works with different ram.

Temp did get to 52C in one test, but it was running a bit cooler than that before i took it apart and put it back together. And it was still experiencing the problem. If i remember right i was seeing it happen with temperatures in the 40s.

User avatar
mosc
Doesn't care what you think.
Posts: 5403
Joined: Fri May 11, 2007 3:03 pm UTC

Re: Help! Impossible hardware problem

Postby mosc » Tue Apr 12, 2011 1:00 pm UTC

naschilling wrote:In about 20 years of working on computer hardware, I've seen ONE (1) processor that died of "natural" causes. I've seen plenty that have had broken pins or were cracked (I still don't know how they did that), but only one that looked like it should work and didn't. Most often, it is an issue with memory or the interface to the memory.

I've seen several CPU's burnt out from overclocking, but more importantly, I've seen several CPU's not work correctly with several motherboards. It's not so much the CPU as it is the motherboard not supporting that particular chip's unique flavorings. Motherboards are incredibly complex and interwoven pieces of hardware that don't often work perfectly.
Title: It was given by the XKCD moderators to me because they didn't care what I thought (I made some rantings, etc). I care what YOU think, the joke is forums.xkcd doesn't care what I think.

User avatar
alexriehl
Posts: 127
Joined: Wed Sep 22, 2010 1:19 am UTC
Location: Auburn, AL
Contact:

Re: Help! Impossible hardware problem

Postby alexriehl » Tue Apr 12, 2011 1:50 pm UTC

I had an Acer Aspire mini-PC that fried 3 times with symptoms exactly like the ones you're describing. The failed component: Northgate. Reason: Overheat. Reason for that: The cooling fins on the Northgate were 90 degrees to the cooling fins for the CPU. There was only one fan in the entire case, connected directly between the CPU fins and the outside world. (And it was only a 2-inch fan. Which almost fried itself trying to keep the CPU cool.) As a result, there was NO airflow over the northgate. The PC was so compact that there was no room for the heat to dissipate. Towards the end, I would be on the floor, pushing the power button every 5 seconds for about 2 hours before it would start up so I could save my data. Basically, poor design on Acer AMD compact desktops. Last time I buy one of those.
[EDIT] Clarification: In this idiotic design, the northgate was built into the motherboard, and should have had a separate cooling fan. [/EDIT]
AMA about suitcases
Spoiler:
Image

User avatar
Meteorswarm
Posts: 979
Joined: Sun Dec 27, 2009 12:28 am UTC
Location: Ithaca, NY

Re: Help! Impossible hardware problem

Postby Meteorswarm » Sat Apr 16, 2011 6:37 am UTC

Come to think of it, your symptoms are _very_ similar to what I observed when my southbridge in my Eeepc lost its thermal transfer pad (and connection to heatsink), except that mine was clearly heat related - a recent crash meant the next would be soon, and a fan could act as life support. If the mobo had been higher wattage, this might not have happened.

Maybe take an infrared thermometer to your case and see if anything is out of the ordinary? It's possible some component is overheating or has otherwise failed in both of your boards, or maybe your case is preventing airflow to some cooling device.
The same as the old Meteorswarm, now with fewer posts!

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Tue May 10, 2011 12:17 am UTC

So i ran the table test and it does appear to be working perfectly. I did have one reset but when i brought it back up, the error message was not the same, and it looks like it was just a bluescreen unrelated to my original problem.

So what does this mean about my problem? I'm using the spacers and for the life of me i can't figure out what could be wrong about the way they are in. It seems pretty straightforward. Would a picture help?

User avatar
alexriehl
Posts: 127
Joined: Wed Sep 22, 2010 1:19 am UTC
Location: Auburn, AL
Contact:

Re: Help! Impossible hardware problem

Postby alexriehl » Tue May 10, 2011 12:45 am UTC

ryak2002 wrote:So i ran the table test and it does appear to be working perfectly. I did have one reset but when i brought it back up, the error message was not the same, and it looks like it was just a bluescreen unrelated to my original problem.

So what does this mean about my problem? I'm using the spacers and for the life of me i can't figure out what could be wrong about the way they are in. It seems pretty straightforward. Would a picture help?


If the spacers are solid, try cutting/machining vents into them. Might help air flow a little better.
AMA about suitcases
Spoiler:
Image

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Tue May 10, 2011 3:15 am UTC

What kind of case are you using? A picture would be helpful too.

One possible cause is that the standoffs are too short, causing the motherboard's heatsink backplate to come in contact with the case.

User avatar
Kromix
Posts: 288
Joined: Thu Mar 25, 2010 7:26 pm UTC
Location: DFW, Great State of Texas

Re: Help! Impossible hardware problem

Postby Kromix » Tue May 10, 2011 12:58 pm UTC

KingofMadCows wrote:What kind of case are you using? A picture would be helpful too.

One possible cause is that the standoffs are too short, causing the motherboard's heatsink backplate to come in contact with the case.


ooOOO, ive seen this once, it was not pretty to troubleshoot, as longer standoffs meant misalignment of the back plate and PCI slots. i ended up cutting a square on the mobo tray so they would not contact...
Image
Click Me! <---- Click There! :)
Spoiler:
Image

zmatt
Posts: 554
Joined: Mon Feb 07, 2011 3:48 pm UTC

Re: Help! Impossible hardware problem

Postby zmatt » Tue May 10, 2011 1:43 pm UTC

Solt wrote:Willing to bet it's your CPU.

You say it's related to RAM and it says right there that it's caused by the Northbridge. The Northbridge is the memory controller and it's located on the CPU: http://en.wikipedia.org/wiki/Northbridg ... mputing%29

What do I win?


Edit: A hunch. Wiki says northbridge plays a role in overclocking. The spec page for your motherboard says 1600(O.C.) for supported RAM speeds. What does the O.C. mean? Can all this be fixed with a setting? Highly suggest you read through your motherboard manual before replacing the CPU.

http://www.asus.com/product.aspx?P_ID=0AvsBb7WBZe2i9zK

Edit2: O.C. is probably a red herring. Make sure you have the latest BIOS. Follow the link for your processor under the 'CPU Support List' tab (yes, their BIOS updates are processor specific, which should be a promising sign!) Note that there are two versions of your processor, make sure the entire model number matches.

Also download all the latest drivers under the "Download" tab.



No.

This is an AMD system, the Northbridge is not on the cpu. If it were a modern Intel system this would be a mostly true statement as the fsb has been replaced by QPI and the system no longer uses a nrothbridge.

However AMD chips do have on chip memory controllers. If changing the memory is affecting it then it is somehow related. I would go into BIOS and underclock you memory and make sure that it is being fed enough voltage. There should be a sticker on the RAM that says what speed, timing and voltage it should run at. I have seen modern motherboards by default set incorrect default values and cause stability issues before.

Also, if you are running dual channel for your memory make sure you are using identical pairs of sticks. If you aren't and are forcing dual channel somehow this may be an issue.

If you can find another compatible AMD cpu then I would try booting with it and running a stress test program of some form or a benchmark program to make sure it is stable.
clockworkmonk wrote:Except for Warren G. Harding. Fuck that guy.

yewbie
Posts: 208
Joined: Mon Sep 13, 2010 6:35 pm UTC

Re: Help! Impossible hardware problem

Postby yewbie » Tue May 10, 2011 2:32 pm UTC

I would try 3 things:

1) If your using metal standoff's trying switching to plastic ones (if you have them) it may be a strange grounding issue..And if your using plastic use metal ones (you should be using metal ones anyway). Also the suggestion above about standoffs being too small / large / uneven and flexing the board is sold advice.

2) Check your I/O shield and make sure the guides for your ports are not getting put into any of your ports.

3) See if there is a standoff you have in place on your case that is not actually on the board.

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Tue May 10, 2011 8:43 pm UTC

I've only heard of this kind of problem with older cases have. Newer cases either have a hole cut out for the heatsink's backplate or areas slightly bumped out behind the blackplate. So if your case is too old, 4 or 5 years old, then you just might need a new case.

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Wed May 11, 2011 1:36 am UTC

It's a pretty new Coolermaster case. When i bought it ~2 years ago, it had really good reviews, with no complaints about any shorting problems. I took some pictures in case anything obvious sticks out. I'm using the 6 metal standoffs in the position of the 6 screw holes in the motherboard.
Image
Image
Image

I'm not sure if those links will work permenantly, so if those are broken, you should be able to find the same images here:

http://cid-9dc8430d6a9a06b3.office.live ... spx/Public

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Wed May 11, 2011 2:51 am UTC

Doesn't look anything is wrong.

When these things happen, it's often due to the case being bent in some way or the I/O plate touching something.

Also, was the case case dusty? It's possible that there was dust behind the motherboard shorting it out.

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Wed May 11, 2011 3:54 pm UTC

Great news! i made some more progress on this. I really didn't see anything to do to the standoffs, but your suggestion about the IO guides caused me to pay close attention to those. There were a few guides bent weirdly, that i bent back to be the way they were supposed to be (As best i could tell, anyway) and with just the bare minimum of things connected, the PC ran for about 2 hours running prime95 without a reset in the case, so this is very promising. I still need to reconnect the rest of the things and try to fix the mess of cords to see if it's really fixed.

But thanks for the tip!

User avatar
alexriehl
Posts: 127
Joined: Wed Sep 22, 2010 1:19 am UTC
Location: Auburn, AL
Contact:

Re: Help! Impossible hardware problem

Postby alexriehl » Wed May 11, 2011 9:08 pm UTC

ryak2002 wrote:Great news! i made some more progress on this. I really didn't see anything to do to the standoffs, but your suggestion about the IO guides caused me to pay close attention to those. There were a few guides bent weirdly, that i bent back to be the way they were supposed to be (As best i could tell, anyway) and with just the bare minimum of things connected, the PC ran for about 2 hours running prime95 without a reset in the case, so this is very promising. I still need to reconnect the rest of the things and try to fix the mess of cords to see if it's really fixed.

But thanks for the tip!

Excellent! Let us know whether it holds up!
AMA about suitcases
Spoiler:
Image

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Thu May 12, 2011 12:15 am UTC

Terrible news! it's not fixed :/

Also i just realized that that reset i had during the table test actually *was* the same error. For some reason i assumed the immediate error message would match the windows event log entry, so i thought it was different. It was happening *significantly* less often on the table, though.

Here's all i did since yesterday and the rate of occurrence is back to constantly happening:

(1) reconnected the power LED and HD LED
(2) reconnected the USB connection to the case's USB ports
(3) reconnected the HD audio connection to the case's HD audio ports
(4) reconnected 1 fan
(5) reconnected my SATA DVD drive
(6) Turned the case vertical instead of horizontal

I guess i'm going to undo these things one at a time and see if any of them in particular is making it worse.

KingofMadCows
Posts: 165
Joined: Mon Mar 21, 2011 5:12 am UTC

Re: Help! Impossible hardware problem

Postby KingofMadCows » Thu May 12, 2011 12:47 am UTC

Taking everything out of the case ensures that it won't get shorted out by the case but it can still be shorted by something else so you have to make sure that nothing touches the motherboard. Isolate it as best as you can. Put it on the anti-static wrap that came with the motherboard.

When you do the test inside the case, try only connecting the things that are necessary. So don't connect things like the power and hard drive LED, the audio connection, USB connection, and DVD.

zmatt
Posts: 554
Joined: Mon Feb 07, 2011 3:48 pm UTC

Re: Help! Impossible hardware problem

Postby zmatt » Thu May 12, 2011 12:59 pm UTC

You are absolutely sure the hard drive isn't failing right?
clockworkmonk wrote:Except for Warren G. Harding. Fuck that guy.

User avatar
Kromix
Posts: 288
Joined: Thu Mar 25, 2010 7:26 pm UTC
Location: DFW, Great State of Texas

Re: Help! Impossible hardware problem

Postby Kromix » Thu May 12, 2011 12:59 pm UTC

i'm still leaning on CPU issue although could be a chipset issue.


Another thing, did you ever update the chipset drivers?


http://www.asus.com/Motherboards/AMD_AM ... /#download
Image
Click Me! <---- Click There! :)
Spoiler:
Image

ryak2002
Posts: 17
Joined: Thu Apr 07, 2011 12:26 am UTC

Re: Help! Impossible hardware problem

Postby ryak2002 » Thu May 12, 2011 5:42 pm UTC

Reconnecting only what was necessary was what i did yesterday, and it ran for several hours. It reset once outside the case and not at all when i put it back into the case. Then i reconnected those things and it returned to happening almost immediately.

I havent tried updating the motherboard firmware on this new motherboard, but i already tried that before i RMA'd it.


Return to “Hardware”

Who is online

Users browsing this forum: No registered users and 5 guests