Debugging new build

[cpbldr]

I recently build a new system about a week ago with the following parts:
GA-EP45-UD3R
OCZ2RPR10664GK
Core 2 Quad Q6600
PC Power & Cooling S61EPS
PNY 9600GSO 768MB PCI-Express Video Card
case has 6 80mm fans

For the most part part the system seems to be stable however theirs a few things it sporadically does that I don't know whats causing it. Here is some of the quarks that its done.

• Consistent cold boot failure. (computer off for 3+ hours then start)
  • Started to boot, blue screen in windows start up and rebooted
  • Powered on came back 5 min later both screens where blank
    • Mobo lights here all lit
  • Second boot screen was scrolling by, restarted
  • Bios usually come up on just first monitor, one time it came up on the second restarted in bios boot screen, came up on first monitor restarted again, second monitor signal led indicator turned blue but nothing after that.
• out of 4 fresh installs of windows 64bit, twice after running the Intel inf updater, all the usb ports were dead,used keyboard and device manager said drivers corrupt,a uninstall and reinstall fixed it, but the other 2 installs no problem after inf update.
• Took me over a hour to install Google earth
  • found out it was msiexec.exe, any installer that used that was just hanging, if i replaced the 64bit version with the 32bit version it worked. later after a restart or two the 64bit msiexec.exe works again and does not hang.
• System shows Norton ant programs not responding on shutdown temperately ( never had that before on my system with same Norton version, anti process termination?)
• Winamp AVIS visualizations earthier don't work or crash winamp.
• One point Firefox 3.0.7 would randomly hang
  • any web page in any window would just continue in loading status as if the server was not responding.
  • would have to close Firefox (terminate process that was not ending) then restart Firefox and everything was working again.
  • Have had it do that after reinstall though

I know a lot of the things i listed sound like software issues but with them being so random and combined with the cold boot restarts makes me think something flaky.

The only things i have change in the bios is to set it to optimal setting, disable the full screen logo, change ram timing to 5-5-5-18 and 2.1v as the ram specs say. No errors after some memtest86+ testing ( never ran a complet test but it got up to test 7-8 before i decided the ram was ok)

some short burn in testing and benchmarking come up fine(one crash of PerformanceTest).

on a side note using 64bit of burnintest or PerformanceTest none of the 3d
graphics run but there fine in 32 bit. Is that because the graphics card is 32bit? (installed the latest dirext x 9 drivers and 64 bit forceware).

Does anyone know what might be causing some of theses issues more so the reboots it seems like something intermittently failing and causing different issues? I have seen many postings about GA-EP45-UD3R with OCZ2RPR10664GK but the seem to be working together, could that be causing the boot issues?

My guess is its something that fails when only warming up from a cold start. Ambeant room temp is ~20c. after a 2-3 reboots the system boots up fine but i have to reset the ram voltage and timing since it thinks it caused the boot fail, however the settings work fine after everything been on for a few min. with speedfan everything idles in the 30's, Everest reports all core temps at high 30's on stock cooler. I just bought a non stock cooler even though the temps are decent but i don't want to put it in till i know the systems stable at stock settings.

[Lsdmeasap]

Re: Debugging new build

Well, I do see possible software issues and x64 XP is buggy, but if you normally use it then it may not totally be software. I can say for 2x2GB there is normally a few changes that need to be made other then what you have mentioned.

Memtest, are you using the latest version? Just checking to be sure, if not here is the latest >>
Memtest86+ - Advanced Memory Diagnostic Tool

Not even a full test? That is for sure not a sign of any memory stability, a full 5-10 passes is recommended. And since some of your settings changes you mentioned do not include the normal 2x2GB ones I would suspect Ram instability may be partial if not all of the cause of your issues here

I have not used this "Performance Test" you mention, so I am not sure what you are asking about it. Intel Burn Test will run in x64 mode if you run the x64 exe, or have the latest that includes both so you could choose the x64 exe to start. Latest is available here >>

http://forums.tweaktown.com/f69/latest-overclocking-programs-system-info-benchmarking-stability-tools-30530/

But, you need to Fully test your ram and get it stable before you even think about passing Intel Burn Test, or you will only have failures.

Intel Inf Drivers, are you using all the latest not what is on your CD? If not, that could be the issue with that. Latest Intel Inf Drivers are here >>

http://downloadcenter.intel.com/confirm.aspx?httpDown=http://downloadmirror.intel.com/16023/a08/infinst_autol.exe&agr=&ProductID=816&DwnldId=16023&strOSs=&OSFullName=&lang=eng

And all the latest drivers for your board are here, Intel one as well I just always get mine direct from Intel >>>

http://www.gigabyte.com.tw/Support/Motherboard/Driver_Model.aspx?ProductID=2921

Sounds like you like Norton, so I will not leave my thoughts on that

Please fill this out with the settings you are using now and I will advise back, I think you may just need a few adjustments settings wise and you may be fine >>>

This is Just a template I use to ask people their settings, pay no attention to the numbers in it, replace with yours

[b][u]MB Intelligent Tweaker(M.I.T.)[/u][/b]
Robust Graphics Booster ...............: Auto or Fast
CPU Clock Ratio ..........................: 
Fine CPU Clock Ratio.....................:.0
CPU Frequency ...........................: 


[b][u]Clock Chip Control [/u][/b]
[b][u]Standard Clock Control[/u][/b]
CPU Host Clock Control..................: [Enabled] 
CPU Host Frequency (Mhz) ............: Mhz
[B]PCI Express Frequency (Mhz) .........:Set to 100[/B]

C.I.A.2 .................................:[Disabled]


[b][u]Advanced Clock Control [Press Enter][/u][/b]
CPU Clock Drive...........................: <<< 800
PCI Express Clock Drive.................: <<< 900
CPU Clock Skew (ps)....................: <<< 
MCH Clock Skew (ps)...................: <<< 


[b][u]DRAM Performance Control[/u][/b][u][/u]
[B]Performance Enhance...................: [STANDARD][/B]
Extreme Memory Profile (X.M.P.).....: Disabled
(G)MCH Frequency Latch...............: [Auto]
System Memory Multiplier ..............: 
Memory Frequency (Mhz) ..............: 
DRAM Timing Selectable ................: [Manual]

[b]Standard Timing Control[/b]
CAS Latency Time........................
tRCD .........................................
tRP'...........................................
tRAS..........................................

Set all Suggested below to highest Values first.  Then later you can lower, one at a time, testing as you go

[b][u]Advanced Timing Control[/u][/b]
tRRD...........................................
tWTR..........................................
tWR............................................
tRFC...........................................
tRTP...........................................
Command Rate (CMD) ....................:2

[u][b]Driving Strength Profiles[/b][/u][b][/b]
Driving Strength ............................

[b]Channel A[/b]
Static tRead Value.........................:6/7-10 333/B MCH - 8/9-12 400/D MCH
 tRD Phase0 Adjustment...................:Auto <<< Leave for all below as well
tRD Phase1 Adjustment...................:Auto
 tRD Phase2 Adjustment .................:Auto
 tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)....................:Auto 
Twr2wr(Different Rank)...................:Auto 
Twr2rd(Different Rank)...................:Auto 
Trd2wr(Same/Diff Rank)..................:Auto 
Dimm1 Clock Skew Control...............:Auto  ps
Dimm2 Clock Skew Control...............:Auto  ps

[b]Channel B[/b]
Static tRead Value.........................:6/7-10 333/B MCH - 8/9-12 400/D MCH
 tRD Phase0 Adjustment...................:Auto <<< Leave for all below as well
tRD Phase1 Adjustment...................:Auto
 tRD Phase2 Adjustment .................:Auto
 tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)....................:Auto 
Twr2wr(Different Rank)...................:Auto 
Twr2rd(Different Rank)...................:Auto 
Trd2wr(Same/Diff Rank)..................:Auto 
Dimm1 Clock Skew Control...............:Auto  ps
Dimm2 Clock Skew Control...............:Auto  ps

[b][U]Motherboard Voltage Control[/U][/b][U][/U]
Voltage Type.………... Manual
[b]CPU[/b]
CPU Vcore….……….........................:Auto if not overclocking, if you are overclocking please do find and set your own voltages

CPU Termination…..... 1.200V*.......: << If overclocking Please set to stock, Auto overvolts this way to much.  Optimal range for overclock would be 1.2-1.5 depending on Vcore, try to keep at least .1 under vcore.  Auto is fine if not overclocking

CPU PLL…………………....1.500V*.......: << If overclocking Please set to stock, Auto overvolts this way to much.  Optimal range for overclock would be 1.5-1.65.  Auto is fine if not overclocking

CPU Referen.…………....0.755V*.......: <<< .64% Of Vtt (Termination above) Auto Will set this correctly 

CPU Reference2………,,0.800V*.......:<<< .67% Of Vtt (Termination above) Auto Will set this correctly 

[b]MCH/ICH[/b]
MCH Core…………….....1.100V...........: << 
MCH Reference….…….0.800V...........; <<< Leave
MCH/DRAM Ref.…......0.900V...........: <<< Leave
ICH I/O……………….....1.500V............: <<< Leave
ICH Core…………...……1.100V............: <<< Leave

[b]DRAM[/b]
DRAM Voltage ……....1.800V............: 
DRAM Termination .…0.900V............: <<< Leave
Channel A Reference 0.900V............: <<< Leave
Channel B Reference 0.900V............: <<< Leave

[b][U]Advanced Settings[/U][/b][U][/U]
Limit CPUID Max. to 3.....................: [Disabled]
No-Execute Memory Protect............: [Enabled]
CPU Enhanced Halt (C1E)................: [Disabled]
C2/C2E State Support....................: [Disabled]
x C4/C4E State Support..................: [Disabled]
CPU Thermal Monitor 2(TM2) ...........: [Enabled]
CPU EIST Function.........................: [Disabled]
Virtualization Technology................: [Disabled] Enabled if you use Vmware/Virtual PC

[b][U]Integrated Peripherals[/U][/b][U][/U]
Legacy USB Storage Detect ..............[Disabled] *Note* Must be enabled to flash from USB

[cpbldr]

Re: Debugging new build

Thanks for your response. I will let memtests86+ run overnight and see what I get.I think it may be failing on cold boots but don't know for sure because the system reboots 2-3 times before it can get to memtest86+, by then the systems heated p and whatever caused the reboots is not showing.

I spent a day before the build making sure i have all the latest drivers and apps and burned a fresh memtest86+ boot CD 2.11, saw ozc has there own mod memtest86+ version but older. Always get the latest drivers for everything from the hardware manufactures not the board unless theirs significant differences besides versions.

I have seen all though tweak settings posted around theses forms, is here some app that you guys use to dump the information when booted or do you just remember and fill them out manually?

[Lsdmeasap]

Re: Debugging new build

Ok, well let us know how you get along, and post your settings if you would like some advice.

Nope, sorry it all has to be manually written down then filled in and posted here. It would be nice though for sure! I think if you have a OLD Printer you can use the Prt Scn Button, but that has to be a 1985 printer

[cpbldr]

Re: Debugging new build

Ok ran Memtest86+ over night, no errors whatsoever.


I'm wondering a few things about memtest86+ though:
Does it test the L1 and L2 cache first?
Whys it show 4096k of L2 cache and 32k L1?

CPU specs are L1-128KB+128KB, L2 Cache-2 x 4MB

Newegg.com - Intel Core 2 Quad Q6600 Kentsfield 2.4GHz 2 x 4MB L2 Cache LGA 775 Quad-Core Processor - Processors - Desktops

Does it ever tests 64 bit patterns? (x32 win seems more stable than x64 to me)

Can something be failing under 64bit but not 32bit?

Hows memtests86+ testing 4096M - 4864M when there is only 4096M?

[Chucko]

Re: Debugging new build

I'm wondering a few things about memtest86+ though:
Does it test the L1 and L2 cache first?
Whys it show 4096k of L2 cache and 32k L1?
CPU specs are L1-128KB+128KB, L2 Cache-2 x 4MB

L1/L2 caches are tested indirectly as part of the memory system as a whole. Because all Core 2 Quad CPUs are really 2 separate Core 2 Duo chips in a single package, and Memtest86+ runs on only one core, the L2 of 4096KB represents the active chip. Each core has only 32KB of L1 data cache.

Does it ever tests 64 bit patterns? (x32 win seems more stable than x64 to me)

Can something be failing under 64bit but not 32bit?

Memory is physically accessed as 8 byte (64 bit) doublewords, the caching system and memory controller take care of the difference. Memtest86+ is written using 32 bit instructions for the sake of portability.

Based on your test results it seems highly possible that your memory is OK but your CPU is defective.

Hows memtests86+ testing 4096M - 4864M when there is only 4096M?

BIOS memory mapping. The BIOS and IO are mapped into the region just below 4096MB, the physical memory in that location is mapped above that space.

[cpbldr]

Re: Debugging new build

Here is my Mombo settings with the changes i made after optimal settings in bold:

MB Intelligent Tweaker(M.I.T.)
Robust Graphics Booster .................: Auto
CPU Clock Ratio .........................: 9x
CPU Frequency ...........................: 2.40Ghz( 266x9)


Clock Chip Control 
Standard Clock Control
CPU Host Clock Control...................: [Disabled]
CPU Host Frequency (Mhz) ................: 266Mhz
PCI Express Frequency (Mhz) .............: [Auto]
C.I.A.2 .................................: [Disabled]


Advanced Clock Control [Press Enter]
CPU Clock Drive..........................: 800mV
PCI Express Clock Drive..................: 900Mv
CPU Clock Skew (ps)......................: 0ps
MCH Clock Skew (ps)......................: 0ps


DRAM Performance Control
Performance Enhance......................: [Turbo]
Extreme Memory Profile (X.M.P.)..........: [Auto]
(G)MCH Frequency Latch...................: [Auto]
System Memory Multiplier ................: [Auto] 
Memory Frequency (Mhz) ..................: 1066
DRAM Timing Selectable ..................: [Manual]

Standard Timing Control
CAS Latency Time.........................: 5 [Auto]
tRCD ....................................: 5 [Auto]
tRP'.....................................: 5 [Auto]
[B]tRAS.....................................: 15 [18][/B]


Advanced Timing Control
tRRD.....................................: 4
tWTR.....................................: 4
tWR......................................: 6
tRFC.....................................: 72
tRTP.....................................: 4
Command Rate (CMD) ......................: 0

Channel A/B Timing
Static tRead Value.......................: 4 [Auto]
 tRD Phase0 Adjustment...................: 0 [Auto]
 tRD Phase1 Adjustment...................: 0 [Auto]
 tRD Phase2 Adjustment ..................: 1 [Auto]
 tRD Phase3 Adjustment...................: 1 [Auto]
Trd2rd(Different Rank)...................: 8 [Auto] 
Twr2wr(Different Rank)...................: 8 [Auto] 
Twr2rd(Different Rank)...................: 7 [Auto]
Trd2wr(Same/Diff Rank)...................: 9 [Auto]
Dimm1 Clock Skew Control.................: Auto
Dimm2 Clock Skew Control.................: Auto

Driving Strength Channel A/B.............: All Auto


Motherboard Voltage Control
CPU
Load-Line Calibration....................: [Disabled]
CPU Vcore................................: 1.26250V [Auto]
CPU Termination..........................: 1.200V   [Auto]
CPU PLL..................................: 1.500V   [Auto]
CPU Reference............................: 0.760V   [Auto]

MCH/ICH
MCH Core.................................: 1.100V   [Auto]
MCH Reference............................: 0.760V   [Auto]
MCH/DRAM Ref.............................: 0.900V   [Auto]
ICH I/O..................................: 1.500V   [Auto]
ICH Core.................................: 1.100V   [Auto]

DRAM
[B]DRAM Voltage.............................: 1.800V   [2.100V][/B]
DRAM Termination.........................: 0.900V   [Auto]
Channel A Reference......................: 0.900V   [Auto]
Channel B Reference......................: 0.900V   [Auto]

[cpbldr]

Re: Debugging new build

I will do some CPU testing once I get a chance to put on the new heatsink. With IntelBurnTest the CPU starts to climb past 60c witch makes me nervous.

When i cold boot and it restarts a few times and then memtest86+ comes up it defaults back to 1.8v and memtests86+ shows 5-7-7-20 timings until i change the voltage back.
considering i usually leave my computer on 24/7 cold boot may not be much of an issue but more of a concern. just have to remember to reset everythign back in bios after any long off time.

Thanks for all the help.

[cpbldr]

Re: Debugging new build

OK after just having my computer off for a while then cold starting it restarted 3 time and then got into memtest86+ here is what i got:


An then shortly after once its warmed up and i get into windows have stuff like this:


So is it bad ram? or just need to warm up? will it more likely fail later?

[Chucko]

Re: Debugging new build

The Memtest failure patterns do suggest intermittent RAM chips. Do you have any other RAM you can test with? Can you isolate it to one DIMM or the other?

[cpbldr]

Re: Debugging new build

Update:
Spent a few hours testing different ram positions and slots trying to isolate the problem.
Each time waiting and powering down my system and unplugging it and making sure i was fully grounded before touching anything.
I did not try tweaking with the setting in the bios to much. tried the optimal setting once, never got back into the bios, had to reset the CMOS by taking out the battery under the video card. That was fun. the ram specs on newegg has 2.1v and 2.15V EVP but the chip has a sticker that says 2.2, i tried 2.14 remembering reading 2.15V somewhere and thinking that maybe 2.2 was 2.15 rounded up. ram was still getting errors though. Another thing, it seems to only have problems when i have 2 sticks in in dual channel, does that indicate it may be with the motherboard instead of the ram?

Anyway after much testing I found that the ram failure is directly related to temperature. with my 6 80mm case fans on on a external 2amp power supply i found that the ram would be OK and get no errors if the fans where turned off but shortly after turning the fan on without changing anything else the ram would start failing. My ambient room temp is 19c-20c. I also noticed that as the errors that where going by the bad patterns seemed to always be the previous tests pattern that memtest86+ was running. what could that indicate? a certain cas timing that need to be set higher than advertised specs? to me it reminds me of the stories about ram freezing and recovering information later, like its running to cold. Is the ram rated for temperatures above 19c or something?

[Chucko]

Re: Debugging new build

Wow... you've done your homework!

Chips should not need to warm up to work correctly.

My opinion, from the symptoms you describe, is that there is some component on the DIMMs or mobo that has a weak solder joint, or a hairline break in a PC trace, etc. It's just small enough that things work when warm, but not when cold. Exactly what is irrelevant, but the failures are subtle enough that I would suspect a bypass capacitor or similar passive component.

It would be interesting to find out if the RAM runs at lower speeds, but it's probably not worth the trouble. I suggest you RMA the RAM and try again with another pair.

[cpbldr]

Re: Debugging new build

can someone tell me some cmos settings that should be fail safe with the OCZ2RPR10664GK? Honestly I don't know what half the settings are or do. All I know is somethings causing issues with cold booting that causes the ram to not change between test patterns. I need some idea of some fail safe ram settings so i can tell if its actually the ram or something else. I tried increasing the MCH core voltage from default 1.1 to 1.2 but the system just rebooted twice and reverted back. Would changing the individual channel voltages from Auto 0.9 to 1.05 help? 0.9*2 =1.8 witch is the default value but the rams rated for 2.1=1.05*2

Also the error addresses seem to go 0,8,c,4,0,c,8,4 in the last digit or the address on the one image I took. That's got to be something with the row or column timing or something no?

[profJim]

Re: Debugging new build

To run your memory at 800 MHz, change the following:

PCI Express Frequency (Mhz) ..........: [100] not memory related, always leave this at 100
Performance Enhance...................,...: [Standard] Turbo is for later tweaking
Extreme Memory Profile (X.M.P.)......: [Disabled]

(G)MCH Frequency Latch..................: [266]
System Memory Multiplier ................: [3.00A]
Memory Frequency (Mhz) ..................: 1066 800 <= this is set by the above 2 settings, 1066 is the default setting for your memory

CAS Latency Time..........................: 5 [Auto] these can be left as is or set to:
tRCD ............................................: 5 [Auto] 4-4-4-12
tRP'.............................................. : 5 [Auto]
tRAS............................................: 15 [18]

Advanced Timing Control
tRRD.....................................: 4 => all advanced timing setting should be on AUTO for now
tWTR.....................................: 4
tWR......................................: 6
tRFC.....................................: 72
tRTP.....................................: 4
Command Rate (CMD) ......................: 0

MCH Core.................................: 1.100V [1.20] maybe 1.22 or 1.24

Verify that the CPU Vcore is properly set by comparing your value with the specs on the cpu's retail box.

Good luck.

[Chucko]

Re: Debugging new build

IME you should not need to increase MCH Core Voltage to run at 800MHz DDR. I have run 333 FSB/800 DDR at quite a bit below the normal 1.1V - the P45 MCH is made to run at 333MHz FSB and 800MHz DDR, so there should be no need to overvolt it at these speeds.

I have been unable to get my OCZ Reapers to run at their spec'd 1066MHz in a similar P45 mobo. I manually set the timings and get some great results at 800 and now 833MHz. I hope I can get 1066 someday... but for now it's working great underclocked.

Follow ProfJim's advice and additionally manually set the DRAM Voltage to 1.8V, since the Reapers don't need any extra voltage at 800.