Is this a mainboard issue?

Highwind

Highwind

[H]ard|Gawd
Joined
Sep 9, 2010
Messages
1,436
It's a doozy, and the cliffs are "I think it's the motherboard." You've been warned.

Corsair TX650W
Asus M3N-HT Deluxe
Phenom II X4 945 (C2 Stepping)
GSkill 2x2GB DDR2 1066 (F2-8500CL5D-4GBPI)
GTX 460 1GB (MSI N460GTX Hawk)
500GB Sata OS/Program Files (WD5000AAKS)
400GB Sata Storage (WD4000AAJS)
LiteOn IDE DVDRW (iHAP122 9)
Windows 7 Pro 64
1920x1080x32 (Samsung P2770FH)


For some weeks now now I have had frequent and sporadic issues with poor performance. The problem arose out of no discernible cause and is most evident during two activities:

1. Opening programs such as Outlook 2007, Steam, or just using browsing hard drives via Windows Explorer will cause the window of the program in question to open and sit non-responsive for several seconds. Additionally, when mousing over my Steam game list or the taskbar, highlights will lag behind the cursor substantially.

2. While playing 3D games, the framerate will plummet from the typical steady 30-60 (depending on the game) to 15 and under. While I have been using Fraps to confirm this, it is painfully evident even with no framerate counter. Looking at the ground or similar 'nothing' places have no effect on this anomaly.

Game examples include BFBC2, GTA4, The Witcher 2, Fallout 3/NV, HL2 and Source games, and even Halo 1.


Troubleshooting:

-I first suspected my previous GPU (GTX 260) was the cause, so I put in an RMA call and was pleasantly sent my current 460, but the symptoms persisted.
-Tested RAM via Memtest86+ v4.00 and Windows Memory Diag
-Moved page file to other drives
-Replaced my OS drive with one of the same model, reimaged, even reinstalled Windows
-Tested my rig with another ample power supply
-Tested with current (latest 3401) and previous two BIOS


Notes:

-Stock clocks or mild OC (CPU 3.3GHz, RAM 880, FSB 220, DRAM/FSB 2:1) make no difference.
-CPU, MCP, GPU, MB temps read well within safe ranges (each struggles to break 50 C under load) regardless of the above OC.
-The only component I've not been able to swap out with an equivalent to test is my CPU, but logging my CPU usage shows it is not fully loaded during this anomaly.
-MB has tall aftermarket chipset heatsinks.

I seem to have run a process of elimination that leaves the mainboard as the lone suspect. Getting a new MB is within my means, but it is a very low priority unless I am certain it will solve my problems and fulfill my hopes and dreams.

Questions/Comments/Input/Good morning
 
Last edited:
Any errors in the Event Viewer?

Have you opened the Task Manager and watched what process(s) may be hogging resources when these slowdowns happen?
 
Try using a different sata cable, try different sata ports.

I would also try with one stick of ram to see if it persists.

I would try dropping your ram to 800 and see if it persists.

If you can, I would also try manually setting the voltage to the CPU and RAM. Then if it persists try bumping up the voltage just slightly and see if it changes anything.

IWhat case are you using?
 
@Old Hippie:

I hadn't thought to look there. There are a couple of errors regarding PhysX installations, along with the following:
capturetwe.jpg


Under Microsoft Antimalware
Session "Microsoft Security Client OOBE" stopped due to the following error: 0xC000000D

Under HAL
The platform firmware has corrupted memory across the previous system power transition. Please check for updated firmware for your system.

Under NVNET
NVIDIA nForce Networking Controller : Has encountered an invalid network address
Click to expand...

As for taskman, I keep it open almost 100% of the time my PC is on. CPU time is predominantly hogged by the game that I am playing, particularly a few of the aforementioned games that use 4 cores. Next to that, the occasional MS antimalware app (msmpeng) usually comes in a distant second. During situation "1." in the OP, nothing is using more than single-digit percentages. I always have some overhead left over when things start bugging out in either case. I have never seen all physical memory utilized either.

I suppose I forgot to mention the following, but for what it's worth, the board has other problems I've been aware of long before the stuff described in the OP cropped up. The nForce ethernet NIC will not work for me with any drivers (I use 802.11g anyway), I can't for the life of me find a driver that will get the eSATA jack to work, and the computer counts time slow (e.g., if I do not auto-sync my time, it will lose minutes each hour). I did change the battery on the board for one that is known working, but that changed nothing.



@qbanb8582

I have tried keeping the ram at 800 and the cpu at stock 3ghz. The last time I attempted to change the cpu voltage, CPU-z read the voltage as unchanged. I will attempt to notch the volts up on both again, likely one at a time. I will try with one RAM stick and see what happens, but booting from different sata ports and/or with different cables did nothing different.

This is my case (I added a 120mm intake fan on the side window over the PCIE slots): http://www.newegg.com/Product/Product.aspx?Item=N82E16811144045




Looks like I've got quite a day ahead of me.
Also, an afterthought, I have the nForce NIC disabled in device manager since it has not worked for me anyway.
 
NVIDIA nForce Networking Controller
Click to expand...
nVidia has the worst drivers around.

Whatever the hell that is....get rid of it.
Edit...I just read you disabled it but do yourself a favor an uninstall the drivers. ;)

I've always read a complete uninstall helps but you have a multitude of other problems and they aren't connected to the lack of eSATA or the nForce controller.

If you're serious about TSing your machine I would suggest clearing the BIOS and disabling everything not needed, disconnect the storage drive and go about your business with the task manager open.

My suggestion doesn't really cover any new ground and you said you reinstalled Windows (which I hope was a new install and not the same old OS image) but other than that I'm pretty much outta suggestions.

Did this slowdown come after any patches/updates?
 
I attempted with 1 stick of ram to no avail. I was also unable, as I suspected, to bump the vcore up. I did set the memory voltage to 2.1 (it is rated for that voltage) and left it at 800MHz, but this made no difference either.

HOWEVER. I think I've run across useful info, and it is dumb that I haven't noticed this. It seems whenever the slowdown begins, the fsb multi reads 4x (for a CPU speed of 800MHz as opposed to 15x/3000MHz). I did some digging, and my suspicion now lies with the CPU being throttled as the reason for my fps drops.

Assuming that is correct, that begs the question, what is causing it to be throttled? I disabled CoolNQuiet and C1E in the BIOS and changed my power plan to High Performance in Windows, but this still occurs.

I have watched my temps. I watch my CPU temp all the time, and if it breaks 50, it's getting loaded HARD. I can't in good faith say that the CPU is overheating. I'll accept the possibility, but I find it difficult to swallow.

I did cut the heatpipe between the mosfets and north chip to install the North/south heatsinks. But I left the remainder of the heatpipe over the mosfets since that is the section of the pipe that has the copper fins.

Is it possible that the mosfet area or something else on the MB are overheating? I know that the chipset is getting better cooling than before, and I want to believe that disconnecting those two chips from the pipe improved the mosfet cooling (with the pipe intact, the north-south chips were hot to the touch.

Does this sound like the mosfets are getting fried? It would make sense to me that this would be linked to my inability to force a vcore hike.

Also, I have found an app that allegedly forces the CPU to maintain its high multiplier and voltage and prevent throttling. But I'm a little scared to use it if something is going to melt.

Consider this my troubleshooting journal for now. Input is still welcome, as I am by no means an expert.
 
Old Hippie said:
nVidia has the worst drivers around.

Whatever the hell that is....get rid of it.
Edit...I just read you disabled it but do yourself a favor an uninstall the drivers. ;)

...
...
...

Did this slowdown come after any patches/updates?
Click to expand...

I've tried to uninstall that driver in the past. W7 immediately re-recognizes a new device and installs generic Microsoft drivers, which are equally worthless. This is an RMA replacement from many months ago, and I don't remember if the ethernet jack worked on the original board. Regardless, ASUS probably won't accept an RMA with a cut or missing heatpipe.

As far as the last question, no, I don't recall patching, updating, installing or changing anything to trigger the onslaught of this nightmare. If I had even a remote guess I'd let you know, but it seems to have legitimately come out of the blue.

Also, please pardon my comprehension--lack of sleep is getting to me--but what do you mean by TSing my machine?



--An update edit.
I've been leaving CPUz up throughout this morning's tests, and as I sit typing with only that, the task manager, and this browser open (pretty near idle) the multi/volts have returned to 15/1.42, as if something cooled off and backed off the throttle. And I just got out of another BFBC2 test run where the multi did indeed drop to 4x during the slowdown.

And in the event I lean over and KO for awhile, let me say while I still can, thank you guys for the help.
 
Last edited:
What settings do you have in ur bios for power management? Is there any thermal protection enabled?
 
This probably won't make difference but download the latest nforce drivers for your chipset and install them if you haven't done so.
 
qbanb8582 said:
What settings do you have in ur bios for power management? Is there any thermal protection enabled?
Click to expand...

All bios settings related to power management have been disabled.

qbanb8582 said:
This probably won't make difference but download the latest nforce drivers for your chipset and install them if you haven't done so.
Click to expand...

I have the most recent installed and have, over the past weeks of this mess, tried several prior versions. No cigar.
 
I've tried to uninstall that driver in the past. W7 immediately re-recognizes a new device and installs generic Microsoft drivers, which are equally worthless
Click to expand...
Have you tried that Atheros F1 Gigabit port and disable/remove the other?

It probably really doesn't matter with your main problem but I've read a lotta bad things about that nForce driver and I'd just get rid of it.

If not, get a nice Intel PCI network adaptor for 25.00...it's worth it.

I'm outta ideas.

Good luck and next time buy Intel. :)

PS....TSing = Trouble Shooting
 
I would guess you try a different mainboard. You could check to see if the heatpipes are making contact with the mosfets. You could put fans pointing at them and see if it changes that would let you know its a heat issue.
 
Sounds like over heating the cpu to me.

Have you tried taking the heat sink off and re-seating it? A pain in the ass I know. Did you build the system? Or was it a shop somewhere? Last time I had a shop put together a box for me they put the thermal paste on extra thick -- caused all sorts of problems till I found it.
 
Answers first, then updates:

@Elf_boy
I did indeed reseat both the CPU sink and both chipset sinks. The system is 100% self-built, I wouldn't trust a shop to get it all right unless I was the shop guy (an ironic statement in this thread, I know).

I agree at this point that the problem is overheating, but I am very hesitant to say it's the CPU toasting. This CPU has never breached 60 C that I've seen (I monitor a lot) and it takes prolonged heavy load to get it into the 50s. It sits idle at around 35 C. That is the main reason I want to blame the MB heatng over the CPU heating, the second reason being that I've altered the MB cooling solution and may have worsened part of it by doing so. Mind you, I am not ruling out that the CPU is indeed what is overheating, but I'm skeptical about that conclusion.

Anyway --the update for what it's worth--

Yesterday evening, I grabbed a program that modifies the CPU power states such that the processor's multiplier and voltage do not get throttled. Basically it forces full power at all times, no backing off to save energy and heat when the juice is not needed.

Since then I have spent hours load-testing in some CPU-heavy games I mentioned above, in particular BFBC2 and The Witcher 2, and performance has been back up exactly where it was before all this began--with a little extra boost from the 460 I got out of this. Doubtlessly this is not great for whatever component is having heat issues, but for the moment, all systems are go. As much as I'd love an excuse to get a new board, I have my rig performing to spec.

I'll leave this thread open since I advocate sharing reports and empirical info should something go awry in the future, but for now consider the case closed... until something finally goes pop. Thanks again to those who chimed in.
 
I am slightly confused by your motherboard. You said it has tall aftermarket chipset heatsinks? I am assuming that the original piped heatsinks have been replaced? If so, why? And if not I suspect it is your motherboard that is failing.

Those heatsinks that Asus and many other manufactures use are not very good quality wise. It would be my guess at this juncture that you have had this system for 2-2.5 years and while they were putting solid cap capictators on those boards the whole chipset cooling solution that had shared the heatpipes between the NB-SB and voltage regulators next to the CPU socket were doing nothing more than sharing heat amongst all three areas of the board. It is my opinion without seeing your rig nor being able to test anything that you were right from the start. Your motherboard is dying a slow but eventual death. Oh how I hate those gimmicky things that OEM's will do to lure people in.
 
@woodscomp
This is on the north chip: http://www.thermalright.com/new_a_p.../hr05_ifx/product_chitset_cooler_hr05_ifx.htm
This is on the south chip: http://www.thermalright.com/new_a_p...i_ifx/product_chitset_cooler_hr05_sli_ifx.htm
My reason for doing so being that the MCP sensor was reading in the 70s even on cold boot with the original heatpipe solution (now it is in the 40s). That just made me super nervous... and I can reuse these in future upgrades.

Additionally, you're right more or less about how long I've had it. The rest of the system is newer, but I got the M3N-HT in early 2009.
 
Good choice to get rid of the "heatpipe" solution that came stock on those boards. I do not know how long you ran the board with the stock cooling solution but it could very well be heat degradation over time. Also how soon after changing the solution did you notice the problems? If it was right away you may have accidently damaged one of the chips on the board or the soldered channels leading from them. The channels are exposed and if you slipped with a screw driver or other instrument that could very well be the source of your problem. If you have had that solution on there for a long time then I would still guess it is a motherboard failure before a CPU failure. Which brings me to the CPU, monitor your CPU fan speed. If by chance it is running super slow you may just have a bad fan on it and it is time to replace the heatsink fan unit. I have seen this from time to time as well. You stated that your temps are all within normal range, but what you are describing as the problem sounds like it is heat related. The simplest way to check your fans is with a separate program or you could remove the side panel and observe it manually. Hard drives also fail a lot within the 2-4 year period. But this does not sound like a hard drive issue.

Years ago I had an abit board that came with a very poor cooling solution, it had a fan and the fan died. I opted to go with a fanless zalman solution instead and had a hell of a time removing the puny heatsink with the bad fan on it. Somehow I damaged the southbridge chipset and ended up with a $90 paperweight. Although it was a cheap solution they had on that board it was glued on there, replacing the fan was near impossible, I don't speak mandarin and abit support was sub par and super slow.
 
Quasi-necro. OP here.

For what it's worth, I thought I'd explain that the all the issues above have been resolved. Where I left off, I decided my mainboard was throttling the CPU multiplier and voltage, thus being the cause of my problems. Since then, I bought replaced the M3N with an ASUS M5A97 (AM3+) and put in 2x4GB DDR3 and same Phenom II 945. Problem solved--nothing is throttling the CPU now.















However, despite my original problem being solved, the M5A97 would not post with GPU in the primary PCIEx16 slot. 1 long beep, 3 short beeps; VGA problem. I used the secondary slot (x4 speed), and the PC posted, booted, and ran/gamed stable (though with a little setback from the lower slot bandwidth).

I returned that board, then bought a M5A97 EVO, almost the same board, but better. But this board also fails to detect my GPU in the primary PCIEx16 slot, giving me the same beep code. In the secondary slot, the GPU was fine again, and the PC was stable as before.

I used my GPU in this scenario, GTX 460, in a buddy's machine and it worked fine. I used his HD4870 in my 'weird' primary slot, and it worked fine as well!

So currently I'm running my sig rig (stock clocks at all times) with the gpu in slot 2, and am posting out of boredom, befuzzlement, and just for the sake of empirical documentation. Thank you for reading.
 
Blog update:

I just finished installing a replacement board. I returned my M5A97 EVO in exchange for another NIB of the same board. Same problem. I've now gone M5a97 --> m5a97 evo --> m5a97 evo and my 460 doesn't get recognized in the top slot.

Maybe it's just my particular model? As I menioned, any other mb/gpu combination seems to work fine. Maybe I'll be replacing the gpu, cause this board is too good to get rid of...
 
I've tried to uninstall that driver in the past. W7 immediately re-recognizes a new device and installs generic Microsoft drivers, which are equally worthless
Click to expand...
Next time uninstall the device from the Device Manager and it won't need a driver. ;)

It can be 'reinstalled' thru the BIOS.

I Googled around and it looks like your card + MB shouldn't have any problems.

I saw one other thread at the Asus Forum but no solutions were given......For all I know it may have been posted by you. :)
 
Hah, I did not make the thread, but while looking for a solution I did find and make a post in it. Having just checked back, it seems the guy RMAd his board and solved his problem. Given that I'm on my 3rd of this board, I have low hopes for that route.

For now i'm twiddling my thumbs and using the card in the secondary slot till I can figure something out
 
but while looking for a solution I did find and make a post in it
Click to expand...
OK Fowle. :)

I'm curious.....why change your ID from site to site?

I've been the Old Hippie forever (12-14 yrs) on every different site.

Are you hiding or ashamed? :confused:
 
I don't even remember when I registered there. That was the first post I made, I didn't even know what my user handle was since you log in with your email. But out of fun of playing along I'll go with A) hiding.
 
But out of fun of playing along I'll go with A) hiding.
Click to expand...
From what? :D

I just mean, I really don't understand the name changing thing.

Looks to me like 95% of people do this but I don't understand why.

I supposed many think they're 'protecting' their internet anonymity but who cares who you are?

I post under the same Nik on aquarium/sports car/music/fireworks/hacks/computer and other boards under Old Hippie IF it hasn't already been taken.

Wayyyyyy off topic so let's get back to our regular programming! LOL!
 
Hey I said I'll play along, I haven't had time to come up with what I'm hiding from yet. (In all serious, I didn't even know what my forum handle there was till I posted)

The only on-topic news is that I got up from twiddling my thumbs to get my weekly shower. Now I'm back to twiddling and pondering whether I should try yet another replacement, get a refund and buy something else, or somehow switch GPUs. I'm strongly leaning toward a different board, but it kills me because this one is just majestic for its price range (this one quirk notwithstanding...)

(ITT: I ramble on the verge of spamming and get nowhere)
 
y I said I'll play along, I haven't had time to come up with what I'm hiding from yet. (In all serious, I didn't even know what my forum handle there was till I posted)
Click to expand...
LOL!

Forget about it!

I'm just wondering about human nature and since I knew that you had posted under a different Nik, I thought I'd ask. :)

AAR, let's get back to the task at hand.....not an off-the-wall-something I'm wondering about. :)
 
You must log in or register to reply here.
Back
Top