Instability issues.

[hazeman]

1. Full description of the problem and symptoms
   Big picture is-- random periods and durations of memory intensive computing-- I'll receive a BSOD (which flashes too fast to read what the error is, or simply the program with crash and I'm back to the desktop. This is not exactly a problem that I can recreate. Various CPU and System temps, random times.
   
   
2. Did it work normally at one time, or has the problem always existed?Not sure how to answer this. Fairly new build and the majority of the time this system has been sitting on/idle. I did Flash the BIOS with the included Gigabyte app, @BIOS. Flash went smooth, however there is no indication that updating the BIOS had an adverse affect.
   
   
3. Is the problem consistent and repeatable, or entirely random, or semi-random?As stated above not really. It can happen in 2 hours. It can happen in 2 minutes. I cannot recreate the exact scenario that is causing the crash or BSOD.
   
   
4. I already tried these steps:Ran Memtest86+ and memory checked out.
   
   
   
5. My software:
   • Windows XP Pro SP3
   • avast! (usually disabled)
   • Windows firewall DISABLED, Linksys Router
   
   
   
6. My hardware
   • Gigabyte GA-MA770-UD3
   • AMD Kuma 2.7GHz AM2 (stock clock) 7750
   • EVGA GeForce 9800GT
   • OCZ Reaper HPC 2x2GB PC1066
   • PSU: Thermaltake 430W

 

[RadiclDreamer]

From the memory specs

1066MHz DDR2
EPP 5-5-5-15
(CAS-TRCD-TRP-TRAS)
Available in 2GB (2x1024MB) Dual Channel Kits
Unbuffered
OCZ Lifetime Warranty
2.1 - 2.3** Volts
240 Pin DIMM

Are you running it @ this voltge?

 

[hazeman]

Voltage and Timings are controlled by the BIOS. The timings are 5-5-5-15 and the voltage is 2.1

I'd like to rule out the BIOS update. I have yet to read any issues regarding problems with the BIOS update that are anything close to this problem I am having. I am however considering clearing the CMOS and reloading the factory defaults.

 

[hazeman]

I Ran Memtest and no errors reported.

Also, the 4GB kit that I have is listed as 5-5-5-18 (which is NOT how my BIOS has this memory configured).
Memtest however, has these modules at 5-3-3-3-26...

 

[hazeman]

In addition to above: It appears that there are several "potential" IRQ conflicts. These are reported from within Windows. However, there are no apparent issues with either the Sound driver or the Video driver (which both are on IRQ 18).

I'm going to do some inspections. Computer has been off all night since my last post and all day while I was at work. From past experiences I've known that faulty memory can trigger behaviors similar to this. I'm going to try each DIMM by itself one at a time-- after I do some intermediate inspections.

I was really anticipating someone replying that could help me out with this. As I am really dreading a possible RMA process. (Isolation of the problem is my biggest concern at this point).

Thanks.

 

[RadiclDreamer]

5-3-3-3-26 is out of spec if its the real speed. But then again memtest has it fine so I doubt thats it. Are you letting it create a dump file? If so have you ran it through the ms debug tools?

 

[hazeman]

I just setup the MS Debug utility. I haven't parsed it out yet to see what it says. However, while I was checking into to things, I noticed a few things...

1) Not really an issue, but the "exhaust" fan on the rear of the case is actually an induction fan and it is blowing into my CPU fan. (Two fans blowing against each other). that explains the mild CPU temp variations and the slightly lowered fan RPM.

2) My video card is emitting a subtle 'ticking' noise. This may or may not be an issue of relevance or perhaps leading to a future problem.

3) I checked my NewEgg info and I actually did a very noobish thing. Several weeks before purchasing I had selected a different mobo and an actual matched set of DDR1066. When I was reviewing my wish list I had made some modifications and the Crucial memory I wanted was out of stock so I browsed through and picked a "similar" set of OCZ Reaper memory. I had not realized that the Voltages differed. So, I set the EPP to manual and changed the DDRII voltage settings from 1.8v to 2.1v and so far, so good.

--- I'm now waiting for any possible side-effects to this. I'm safely assuming there won't be any because this is the rated voltage for these DIMMS.

4) Until just within the last 2 hours have I had to use the Reset switch. Upon doing so, my system fans sped waaaay up. At least 2000 RPM each. So now this is something else I'm needing to inquire.


At least for now the instability issue seems to have ceased. The noise emitting from my video card has me slightly concerned, the system fans speeding up when I use the reset switch, and any possible side-effects from increasing the voltage. (The increase of the voltage from the default to 2.1v has me concerned only because this board was designed for voltage-matched DIMMS, which these are not.)

 

[RadiclDreamer]

If your fans are system controlled then a rev up on reboot is fine, they should very quickly normalize. The ticking from your video card could be any number of things, most likely just cheap fan noise. Stop it for a sec and see if it goes away.

 

[KGB]

Originally posted by: hazeman

1. Full description of the problem and symptoms
   Big picture is-- random periods and durations of memory intensive computing-- I'll receive a BSOD (which flashes too fast to read what the error is, or simply the program with crash and I'm back to the desktop. This is not exactly a problem that I can recreate. Various CPU and System temps, random times.
   
   
2. Did it work normally at one time, or has the problem always existed?Not sure how to answer this. Fairly new build and the majority of the time this system has been sitting on/idle. I did Flash the BIOS with the included Gigabyte app, @BIOS. Flash went smooth, however there is no indication that updating the BIOS had an adverse affect.
   
   
3. Is the problem consistent and repeatable, or entirely random, or semi-random?As stated above not really. It can happen in 2 hours. It can happen in 2 minutes. I cannot recreate the exact scenario that is causing the crash or BSOD.
   
   
4. I already tried these steps:Ran Memtest86+ and memory checked out.
   
   
   
5. My software:
   • Windows XP Pro SP3
   • avast! (usually disabled)
   • Windows firewall DISABLED, Linksys Router
   
   
   
6. My hardware
   • Gigabyte GA-MA770-UD3
   • AMD Kuma 2.7GHz AM2 (stock clock) 7750
   • EVGA GeForce 9800GT
   • OCZ Reaper HPC 2x2GB PC1066
   • PSU: Thermaltake 430W

That ram is NOT on the QVL list for your MB.

HERE is the QVL.

 

[hazeman]

KGB, I know it's not... my post above explains the self-error. However, even bumping the DDRII voltage up from 1.8 to 2.1 there is still a bit of instability.



The BSOD shows:

PAGE_FAULT_IN_NONPAGED_AREA

TECH INFO:
STOP: 0X00000050 (0x911A915C, 0X00000008, 0X911A915C, 0X00000000)

 

[KGB]

Originally posted by: hazeman
KGB, I know it's not... my post above explains the self-error. However, even bumping the DDRII voltage up from 1.8 to 2.1 there is still a bit of instability.



The BSOD shows:

PAGE_FAULT_IN_NONPAGED_AREA

TECH INFO:
STOP: 0X00000050 (0x911A915C, 0X00000008, 0X911A915C, 0X00000000)

Yes, I read your post above.

What I'm sayin' is: that RAM may never work in that MB and Gigabyte will only support issues with supported RAM.

Good luck and let us know what happens.

 

[Intexity]

just curious as to if a 430 watt psu leaves enough breathing room for your components...

 

[hazeman]

KGB: As that is understood as well. I'm in the process of deciding on either an Exchange of the memory or keeping it for later on and just purchasing another set. The mobo so far is great and the BIOS features on this board are equally nice. Vista is also another controversial topic with this build. Being 3GB+ RAM and a DirectX 10 capable graphics card.

Intexity: The manual that came with my EVGA 9800GT recommends a 500W+ PSU. However, the processor I've chosen is the Kuma (95W~1.25v) instead of a 125/145W chip (which is supported on this board). I'm seeking the best method for testing the power consumption of this build at a moderate-high load. http://www.journeysystems.com/?power_supply_calculator suggests that this build is around 280W at load. I know you never want to have "just enough" and erring on the side of caution, having "too much" is typically a good thing.

 

[Intexity]

hmmmm..... it was the video card i was thinking of. would be helpful if a program could gauge the watt usage per compononent as the computer is running. but i am probably wrong. i have no experience using lower wattage psu's so i will step out. just food for thought.