Asking for help with WIERD BUZZING SOUND and hard Win7 Lockup during flight

[Robert McDonald]

Request for help on diagnosis / troubleshooting the following weird issue-

 

During routine flights, usually after about 30 mins or so, my Windows 7 box will suddenly freeze up during FSX and a LOUD BUZZING SOUND will come from the speakers (not the PC). Onscreen, the FSX in-flight image is frozen in time... (NO BLUE SCREEN CTD)

 

Here is the sound, in MP4 format. It gets a bit louder after a few seconds because I moved the MIC closer to the speakers.

 

Clues:

 

GTX690 dual-GPU vid card with WHQL drivers (latest) from nVidia card is now running in SLI mode (possible suspect) 3-D is OFF

Onboard Sound Card (ASUS P8P67 Pro mobo)

All latest Win7 updates/patches/SP's 64bit

 

Suspects:

 

REX ESSENTIAL

Note: After this hard crash occurs (no blue screen - system is totally froze up and the loud buzz continues until I do a hard reset), REX won't work properly following this phenomenon and must be reloaded. This suggests to me that something within REX gets corrupted during this entire process.

 

EFB (Electronic Flight Bag by AivlaSoft)

I doubt this is the culprit, since it's running, I'm listing it. I did download the revised Navigraph AIRACC version -2- (specifically for EFB) Still have the problem.

 

Squawkbox 4 running in Windows 7 Compatibility Mode.

 

CORSAIR AX-850 Power Supply, apprx 11 months old. While I have never had a problem with this PSU, I have read of others who experienced the hard system lockups who eventually ran it down to a power supply. I have sent a copy of the recorded noise to Corsair support. This noise comes from the computer's audio system, not the little midget speaker on the motherboard. I must cycle the AC power to clear this lockup. NOT a crash to desktop. I am almost wondering if the SLI mode is sucking more power from the power supply unit - possibly this PSU can't handle the extra draw of BOTH gpu chips under heavy load. I am just throwing thoughts into the wind. I only recently turned on SLI-mode (multi-GPU mode), formerly it was running problem free in single GPU mode.

 

What I've tried with no success:

 

Turned down overclocking from usual 4.8 Ghz to 4.7. No difference.

Tried running SB4 in Windows 95 compatibility mode No difference.

 

If anyone could kindly listen to this loud buzz and put forth a path to troubleshooting this I would be grateful for the assist. This is a someone new issue. I don't remember having the problem before REX ESSENTIAL, but that's my old-man memory, might be something else.

 

I would almost suspect a driver issue, but darned if I can figure out which device is the problem child.

 

Thank you all in advance.

[Swofford]

Try running SB4 in Windows XP SP 3 compatability as well as administrator. I has to do that to cure all sorts of issues with it before moving to FSINN.

[Robert McDonald]

Try running SB4 in Windows XP SP 3 compatability as well as administrator. I has to do that to cure all sorts of issues with it before moving to FSINN.

 

Thanks, I'll certainly try it. I'm trying to think of "what's new" on my system as I don't recall this being a problem for many months of flying with a ton of adds. I of course was running SB4 before this trouble ... so it's 'unlikely' to be the cause...

 

If I had to guess, it MIGHT be something to do with SLI and my video card. The GTX690 when I first got it, I ran it solely in single-GPU mode, and only used multi-GPU very briefly when I connected 3 surround monitors. Lately I put the card into SLI model, and I'm suspecting that 'might' be involved. I prowled around the Corsair PSU forums and read of a person who discovered the culprit was a bad power supply.

 

I -might- experiment with some of the nVidia beta drivers to see if the buzz lockup goes away... but before that I will fly the plane naked without the SB4, EFB and REX Essential running. If NO CRASHING, I will gradually add those 3 back 1 at a time until I once again encounter the lockup.

 

I can add that this is NOT an 'overheating' issue, my Coolermaster HAF-X with supplemental GPU fans, extra roof fan, and Noctua DH14 is keeping both the CPU and the GTX690 well within acceptable heat limits during flight, verified with coretemp monitors and nVidia GPU temp/fan monitor software as well.

 

Thank you for your suggestion.

[Robert McDonald]

I have been given a laundry list of things to check by EVGA support (stellar, by the way)

1. Boot up system, while in BIOS check 12volt Rail power s/b between 11.6 and 12.4 and fluctuate no more than .05. Wild up and down spikes on the 12-volt rail are probable indicators of BAD POWER SUPPLY. Mine was within spec no fluctuation at all.
   
2. Visit treexy.com and download DRIVER-FUSION (free driver cleanout utility) Check!
   
3. Uninstall existing nVidia drivers and particularly nVidia HD Surround Sound drivers, a known probable cause of issues with the GTX690 card (according to EVGA support) CHECK.
   
4. Run DRIVER-FUSION to blow out all remnants of ANY prior nVidia drivers. Select only things that say 'nvidia' within DRIVER FUSION and click REMOVE. CHECK.
   
5. Download the Windows Hardware Quality Lab (WHQL) certified drivers for the GTX690 card from www.nvidia.com CHECK
   
6. Reinstall the drivers but select CUSTOM INSTALL. During the reinstallation, DE-SELECT the HD AUDIO DRIVERS that are offered by nVidia (suspected cause of problems)
   
7. Revisit EVGA website and download their Video Card stress test software "OC SCANNER" Install and run the scanner, watch the performance of the card under load, note if the lockup reoccurs, test GPU1 and then GPU2. PASSED.
   
8. Reload FSX and fly a flight, check for system lockup. FLYING IT NOW, no issues so far. Will report back if the issue remains. I suspected a 'driver' all along. Seems MAYBE this was correct. We shall see. Sharing all of this for other GTX6xx owners in hopes of sparing some pain.
   
9. (NOT DONE) Re-Seat video card in motherboard, check power connectors. If issue remains, try moving your card to a DIFFERENT PC. This would help determine if the CARD is bad or if something else (motherboard, ram, PSU) at fault.
   

[Swofford]

Yeah that's strange haven't had a hiccup with my 670. Hope you got it sorted.

[mr340]

I don't know much about the Nvidea cards But i guest you are not alone with this problem . I was reading this post that might solve your issue . Take a look at this post . It might help with your issue

 

http://www.tomshardware.com/forum/318302-28-computer-freezes-loud-buzzing-sound-speakers

 

Best Regards

[Robert McDonald]

I don't know much about the Nvidea cards But i guest you are not alone with this problem . I was reading this post that might solve your issue . Take a look at this post . It might help with your issue

 

http://www.tomshardw...-sound-speakers

 

Best Regards

 

 

Still PENDING

 

Thank you so much for the help, everyone who wrote comments back to me. This issue IS STILL PENDING.

 

It seems that the HD AUDIO DRIVERS for the nVidia card were conflicting with the built-in audio on my motherboard. This is not 100% confirmed until I do more flights... but so far (fingers crossed) it has not locked up since I uninstalled the nVidia drivers, then ran the Driver Fusion cleaning tool and selected everything in the on-screen list of drivers that said "nVidia" and removed all traces of any nVidia driver from my machine. Then I reinstalled the latest WHQL (Windows Hardware Quality Lab certified) drivers from nVidia.com for the GTX6xx series cards, and here's the important part - DE-SELECTED the HD Audio Drivers from the nVidia installation (select CUSTOM INSTALL). This was suggested by EVGA support.

 

I will re-post if this thing comes back. Flew about 2 hours last night no issues (fingers crossed)

 

OK: it came back flew CYVR to KSEA and crashed as taxiing to my gate.

 

Win7 'found' the HD audio and installed the drivers for it anyway. I just went into device manager and DISABLED the HD audio (there were 2 showing) and will re-fly a test flight. I'm going to be bummed if this doesn't solve it.

 

Thanks to all who wrote in.

[Robert McDonald]

update: BAD RAM CHIPS

 

Corsair support suggested running Memtest86 again, so I downloaded it. I then waited until (sure enough) the machine locked up hard yet again, reran the memory test - failed miserably.

 

Since all 4 sticks are a matched set, I will have to wait for Corsair to send me the replacements. I'm relieved it wasn't the GTX690 or the power supply. Thanks again to those members who tried to lend me a hand on this intermittent (yet ever more-frequent) problem.

[Robert McDonald]

revised update: bad VIDEO CARD - not ram chips?

Replaced the RAM and still was getting memtest86 errors at bootup test - called EVGA and got the GTX690 card replaced via Advance RMA. Popped the new video card into the system and 31 minutes of ram testing later via Memtest86 - all clear... will see what happens will leave memory test running and will fly several test flights to make final determination. Keeping fingers crossed.

 

Aftercare Report Card (Ease of Service/RMA)

 

EVGA 10 out of 10 - they never blinked when I asked for help and an RMA. I had to give a credit card as security for the replacement card. Replacement card is "recertified" - I'm ok with that, it implies it passed their QA testing, which in some ways may be better than brand new. Warranty automatically transfers over to this card (balance of 3 year warranty), they automatically added it to my list of registered equipment on their website.

 

CORSAIR 10 out of 10 - great after sale service, willing to send replacement memory and/or power supply. You can't ask for better than that, like EVGA, Corsair is located in CALIFORNIA, USA. That's a winner for me, since I live in California. They also created an advance replacement for both the AX850 power supply plus the memory sticks. I cancelled the RMA on the PSU and RAM - RAM has been replaced once already (errors still occurred). After changing out the VIDEO CARD, I am hopeful that my issues may be over (crossing fingers) without having to battle the beast of changing out the motherboard.

 

ASUS - UNKNOWN - have not requested RMA on P8P67 Pro rev 3.1 Motherboard as yet. Hoping I won't have to, as the NOCTUA DH-14 is no can of corn to remove and reseat. It features RAZOR SHARP cooling fins that can cut fingers (wear VERY HEAVY gloves when installing or uninstalling), and it's VERY heavy and awkward to work with (for best results, you must pull the motherboard out of the case with the Noctua still attached, then remove the Noctua DH14).

 

AMAZON- You can't beat Amazon customer service, period. They have always been good listeners and willing to help when asked

 

DEGREE OF DIFFICULTY of this case (Scale 1=easy to 10= extremely challenging) I gave this an 11. You can dance to it. Plus, it gets you talking to yourself all the time!

What major pain this was. Robbed me of untold number of hours of sleep, crashed any number of flights just prior to landing or shortly after takeoff...Yikes! I still won't know if this is definitely 'it' until I get several continuous hours of flight in with no more sudden lockups. I'm actually almost afraid to test it. If I get another lockup, it will be pretty darned disappointing, let me share that with you!

[Robert McDonald]

Well, I just played Stump the Dummy - and guess what? STILL getting RAM failures... must be either the motherboard, the CPU, or the power supply. I will be ordering a new motherboard (RMA) from ASUS and I'm going to reactivate the RMA from Corsair for the PSU and just for kicks, the RAM one more time.

 

Sometimes, life deals you a 7 and a 2, unsuited.

[Robert McDonald]

Welcome back to Stump the Dummy! Today I awoke to over 6 hours of continuous Memtest86+ 4.2 with ZERO errors. This was accomplished by reducing the overclock multiplier from 48 (4.8 Ghz) to 44 (4.4 Ghz). If that's what it takes to get rid of the hard and unexpected lockup, I'm all for it. My only worry is that the motherboard (?) may be intermittently failing in some way and/or the Power Supply. While I thought it was the Video Card, 3 different video cards all threw errors at anything over 4.4 Ghz, so I'm thinking that I had overclocked my system to a point where it "seemed" stable but was not rock solid (?). Loading up for some test flights... Be sure to tune in tomorrow for another edition of Stump the Dummy!

[Robert McDonald]

Memtest86+ is the way to go to test RAM sticks. The "Plus" version is allegedly more stable with the newer CPU/Mobos. Even though Memtest86 (without the +) has a slightly later build/release date, Memtest86+ works better on my system.

 

The preliminary working theory on my hard lockups is that although I could successfully fly shorter flights with no issues (that I knew about)- when the ambient temps in my office/flight deck climbed during the summer heat, my unstable Overclock started to rear its ugly head. Not realizing what was going on... I never noticed that I encountered such frequent crashes while video editing that I got mad at Sony Creative and thought Vegas Pro 2011 was buggy (I now realize it was my machine causing those crashes, NOT the Sony software!).

 

And of course the occasional and infrequent hard lockups that I had noted before happened with so much frequency I thought some part of my hardware was failing.

 

I now suspect (and am conducting further testing at the new reduced OC of 4.2 Ghz) that my system really is twitchy at 4.8 Ghz on AIR COOLING. This is not defective hardware, it's a limitation of my hot office space coupled with a lot of heat inside my PC case. If I can run 4.4 with out the hard crashes - well,I'll TAKE IT. I was bordering on GIVING this setup away. It was causing WAY too much misery.

 

Tune in tomorrow...

[Robert McDonald]

• Replaced Power Supply - Corsair AX850
  
• Installed 1866 Ram Chips (16gb)
  
• Checked BIOS settings - turned of EPU (Power Saving), set multiplier to 44, re-ran Memtest86+ 6hrs no errors
  
• Bought 2 cans of compressed air - blew out the cooling fins on the Noctua DH-14 plus all the fan intake/exhaust ports and blades on the entire system top to bottom.
  
• Ran Prime95 Blended test CPU temps fine and no lockups after appx 1 hour
  
• Hoping sudden random lockup/freeze/soundloop issue may be resolved - looks like combination of PSU and Bios Settings. Flown several test flights night/daylight no crashes or lockups. Bonus, turning off EPU power-saving in BIOS and enabling internal PLL OverVoltage seems to have greatly reduced the 'stuttering' that sometimes shows up in FSX.
  
• ASUS motherboard overclocking discussion thread - use caution if changing voltages (I did NOT alter any of the default voltages, but I -did- enable XMP memory profile to use the 1866 Corsair Vengeance Lo-Profile ram.
  
• Degree of difficulty this technical support issue (Scale 1=Your grandson could fix it 10=Bill Gates and his army don't know what to do)
  
• score: 11