Microsoft Level 3 Support Can't Help - can AT? **STILL UNSOLVED AS OF 2/6/12 - help**

[tami]

the memory you have installed. was it a set of 4 sticks matched or was it 2 sets of 2? you will be surprised the instability it adds.

And for your ssd, do you have the latest firmware.

When you bluescreen, are you infront of the computer, or is it when you are away?

How do you have your ssd setup? do you have trim enabled?

1. 4x4GB set, so it shouldn't be an instability issue. Again, I did try each piece of RAM individually to no avail. I tried 7 pieces of RAM in total, all individually. All bluescreened. No exceptions.

2. I replied about SSD firmware earlier - it's been a few months but it was at the latest firmware when I essentially eliminated the possibility that it was an SSD issue. Plus, when I RMA'd the SSD, the newer drive had the latest firmware. (The newest version is 2.15; I have 2.11 - but I've determined this to be a non-issue.)

3. Answered this earlier too 95% away, 5% in front.

4. Yes, TRIM is enabled and working properly.

>fsutil behavior query DisableDeleteNotify
DisableDeleteNotify = 0

Right now, RAM voltage is keeping me hanging on. If it fails, I'll try aforementioned options and maybe upgrade my SSD firmware, but I'm not sure that's the problem. Before I BSOD, if I am physically there, I always get memory errors.

 

[SaDiZTiKStyLeZ]

Yea. I definitely would have worked on ram settings before going any further with this situation.
9/10 times, BSOD's were caused by incorrect ram settings on my custom builds.
I always consult board or ram manufacturer's forums directly, on recommended settings.

 

[tami]

so far, so good. No bluescreens that I've detected (last one was 1/28), though I did have to reboot once during the time I'd normally have one. I think we solved this issue but will update if the problem recurs.

 

[tami]

Okay, apparently I spoke too soon. I had a BSOD within the last 3 hours - I was idle when it happened. I raised my RAM voltage from 1.601 to 1.623 - hope that's not overkill (what is the upper recommended limit without breaking things here?)

So remember when I say that I have to reset my RAM or CMOS to get Windows to work? I'm back there again - otherwise, my system doesn't find the boot disk. Seems like a weird issue altogether... has anyone even ever heard of this?

 

[tami]

Crashed again at 1.623v. Now up to 1.652v. How much is too much?

 

[Lifted]

Surge protectors won't help if voltage on the line is dipping, that's what a UPS is for. Some power supplies are more susceptible to this than others, so you may not have noticed this in the past.

A UPS would be able log events such as input voltage drops, but there must be something cheaper out there that can monitor it for you.

 

[tami]

Hmm, so I'm just playing with the RAM voltage. It did okay at 1.603 for awhile, but then it crashed.

I went up to 1.623 and then 1.652, but my computer froze (no BSOD) about 2 hours later (3 minutes ago), and I'm down to 1.630.

I actually think RAM voltage (perhaps coupled with CPU core voltage or other settings) may help here. What do you think I should change these to if at all? What's my upper limit?

 

[tami]

anyone?

 

[oynaz]

Somebody mentioned problems with your power as a possible cause, and in my opinion this should be upgraded from Possible to Probable. Problems with your power would fit all your symptoms, so that is where I would look next.
I believe you can buy some cheap power testers, which you plug into your wall sockets, and it will report if it detects problems.

 

[fralexandr]

i concur with oynaz and some of the other posters, do you happen to know when the electrical was last updated in your place of residence?
http://www.amazon.com/Gardner-Bende...PC/ref=sr_1_16?ie=UTF8&qid=1328613290&sr=8-16
that can test if your outlet is properly grounded (there are cheaper ones if you have or are eligible for a free trial of amazon prime). you can also probably buy a tester at osh/home depot/most hardware stores which would belay the time necessary for shipping

if this is the problem you'll probably have to either call an electrician (if you own the property) or spend a bit on a continuous or line filtering ups (if you don't, and can't get the management to do something about it; as oynaz says below this isn't a great alternative)

 

[oynaz]

Yes, that tester you are linking to is exactly what I meant. But if it detects anything, call an electrician. Faulty electrical installations causes shocks, but worse, fires, and an UPS won't help with that.

 

[tami]

Thanks. I live in an apartment and maybe someone here has an electrical tester.

Just seems a bit strange for this to start happening on this computer and not previous computers that were plugged into the same outlets...

 

[fralexandr]

sometimes rodents like to chew through cables our house's internet kept becoming unstable during and after rains, and the technician said our cables had been chewed open

 

[Binky]

Does it ever crash under load?

If it generally BSOD's when idle, have you disabled sleep to see if the crashes are related to sleep mode(s)?

Have you tried booting into another OS? Maybe you could boot into Linux Live CD and run a stress test there.

 

[Jaepheth]

I once had a similar issue that was solved by underclocking the RAM a bit.

 

[ViRGE]

I have to admit, even though you have the latest version of the firmware for you SSD I'm still eying it as a culprit. Particularly since AT managed to get a reproducible F4 error on a 240GB SSD using the same SandForce chipset.

http://www.anandtech.com/show/5508/intel-ssd-520-review-cherryville-brings-reliability-to-sandforce

I'm assuming your OS is on it. I would strongly consider repartitioning the HDD, cloning your OS over to that partition, and removing the SSD. If it is a SandForce problem, that will make it go away.

 

[daveybrat]

Download and run Bluescreenview. Then scroll over towards the right hand side and look for the driver or file that is causing the BSOD's.

Let us know if there is a pattern or what the file is.

http://majorgeeks.com/BlueScreenView_d6200.html

 

[flexy]

You said you replaced all hardware (really,EVERYTHING?) but did you do actual tests for stability?

To be honest, to test a system thoroughly there is only a few things to do..but if you do them you can almost CERTAINLY assume that all hardware is right:

* Download OCCT which does stability tests for the CPU. (It has other tests also, but there is better software). If OCCT runs, say, an hour in "medium data size" mode and doesnt complain you can almost guaranteed say that your CPU, cooling, northbridge etc. is all right

* Get Furmark and stress the **** out of your graphics card. If you run it for some time, with no artifacts and no errors/crashes etc....your graphics card, its cooling etc. are all fine. Furmark ALSO seems to stress your PSU a lot....i once blew a brand new OCZ PSU by running furmark...this was the last time i got an OCZ PSU.

* But the FIRST thing you wanna do is make a bootable stick/CD from memtest86+, boot on the CD and let it run for some time, in particular Test #5 and test #8

If test #5 off the memtest CD runs through several times you can say that there is no issues with memory, timings, voltage etc.

* But you SHOULD do additional memory testing (after the initial testing with the memtest boot CD) with the memtest version which runs in windows.

Open several threads (4 if your CPU has 4 cores etc.) and let it test memory from within windows.

When ALL those programs do not cause errors and run fine i would usually say that your computer hardware is ok.

 

[tami]

sometimes rodents like to chew through cables our house's internet kept becoming unstable during and after rains, and the technician said our cables had been chewed open

I've got a 2 year old and we do have an exterminator come by every so often. Don't think that's the issue here I can see my cables fine.

I once had a similar issue that was solved by underclocking the RAM a bit.

So you're saying to go under my standard RAM voltage? It's actually more stable at ~1.6v versus 1.5ish that I had running earlier.

Particularly since AT managed to get a reproducible F4 error on a 240GB SSD using the same SandForce chipset.

Fair call, but that's reproducable under certain events. Mine just happens randomly, and NEVER under any load. I'll consider a backup drive - I do have a 160GB spare HDD lying around - once I remove other possibilities

Download and run Bluescreenview. Then scroll over towards the right hand side and look for the driver or file that is causing the BSOD's.

The Microsoft tech used WhoCrashed; nothing of use was discovered during his 6 weeks of troubleshooting. It sucked.

 

[tami]

You said you replaced all hardware (really,EVERYTHING?) but did you do actual tests for stability?

To be honest, to test a system thoroughly there is only a few things to do..but if you do them you can almost CERTAINLY assume that all hardware is right:

* Download OCCT which does stability tests for the CPU. (It has other tests also, but there is better software). If OCCT runs, say, an hour in "medium data size" mode and doesnt complain you can almost guaranteed say that your CPU, cooling, northbridge etc. is all right

* Get Furmark and stress the **** out of your graphics card. If you run it for some time, with no artifacts and no errors/crashes etc....your graphics card, its cooling etc. are all fine. Furmark ALSO seems to stress your PSU a lot....i once blew a brand new OCZ PSU by running furmark...this was the last time i got an OCZ PSU.

* But the FIRST thing you wanna do is make a bootable stick/CD from memtest86+, boot on the CD and let it run for some time, in particular Test #5 and test #8

If test #5 off the memtest CD runs through several times you can say that there is no issues with memory, timings, voltage etc.

* But you SHOULD do additional memory testing (after the initial testing with the memtest boot CD) with the memtest version which runs in windows.

Open several threads (4 if your CPU has 4 cores etc.) and let it test memory from within windows.

When ALL those programs do not cause errors and run fine i would usually say that your computer hardware is ok.

Yes, flexy, EVERYTHING.

Well, except the case. I replaced the wiring on the case. I didn't replace the DVD-ROM and Rosewill 3.5 USB drive because the computer bluescreened without those plugged in. I did try a different mouse and keyboard for awhile too. I mean, it's been 6 months and I use this computer at home all the time especially as I work from home, so I need this to be stable.

But you saw the OP. Mobo replaced 4x, SSD replaced 2x, CPU replaced 2x, two different PSUs, 2 different mobos (4 mobos in total), 7 different pieces of RAM at different speeds, two video cards...

I ran Memtest86 already. I don't use my video card for gaming, just for 2x display, so while I could run a bunch of tests, I'm not entirely sure those would solve the issue here. Unless you *really* think I should bother given the knowledge that I truly *did* replace everything. What do you say, and what's left?

 

[Binky]

Gotta be something you installed that runs when idle, or sleep mode itself.

 

[tami]

Gotta be something you installed that runs when idle, or sleep mode itself.

Probably, but the computer isn't always in sleep mode when it crashes

Just want to figure out what it is...

Microsoft had me disable drivers of every piece of hardware - and that's also why I swapped the harware as well, so I still don't know if that's what it is.

 

[ViRGE]

Fair call, but that's reproducable under certain events. Mine just happens randomly, and NEVER under any load. I'll consider a backup drive - I do have a 160GB spare HDD lying around - once I remove other possibilities

The thing about the SandForce BSODs is that they manifested themselves as random F4 errors, which is exactly what you're seeing. It took Anand & crew quite a bit of effort to come up with a reproducible test case for SandForce to actually fix the issue. Ultimately you have some other good suggestions here worth following, but I wouldn't rule out the SandForce BSOD because it's random; you should rule it in because it's random.

 

[THRiLL KiLL]

try disabling hibernation / sleep and see what happens.

also disable the option to turn off hdd after x amount of time

 

[tami]

try disabling hibernation / sleep and see what happens.

also disable the option to turn off hdd after x amount of time

Sleep/hibernation is never enabled. Also, the HDD is never shut down.