New
#1
Possible faulty HDD cables causing BSODs. Opinions?
Hi guys
I thought it was about time I looked for help on these forums directly, instead of just lurking around attempting to glean information off other people’s issues for the last year and a bit. Sorry for the long post in advance, i just want to get as much information out there as possible.
Long story short, early last year i purchased a new computer from a custom builder for almost $4000. Ever since i received it, i’ve been getting BSOD’s. The builders never found an issue with it (something i doubt) and because of their lack of interest in following through with my problems, the manufacturers warranty has run out. I’ve re-installed windows multiple times to no avail, updated drivers, run memtests, HDD tests, the lot.
The BSOD’s that i could pinpoint, i found fixes for, which didnt seem to help because the errors just came from somewhere else.
Excessive (30 gigs) data transfer used to cause BSOD’s, which was fixed by turning off USB3 support.
Saving large files (such as 700+meg photoshop images and 1.5+gig Zbrush files) would often cause BSOD’s, and corrupt them in the process. I dont think thats been fixed, but its much more infrequent than it used to be.
Putting the computer to sleep and attempting to resume it would cause BSOD’s, i think i hotfixed that a while ago, but i’ve been to reluctant to use sleep mode since, so i haven't extensively tested it. Something to do with HDD spinup latency, and not being able to read\write data fast enough.
Plugging in more than three external HDD’s at a time would often cause a crash.
Running Checkdisk scans on my hard drives would report hundreds of corrupted files and sectors, which were all fixable, but kept appearing after subsequent crashes corrupted new ones. The hard drives are perfectly healthy, with no bad sectors on multiple assessment programs. I’ve also had parts of windows corrupt that were active during a crash - notably the .NET 2.0 framework which was notorious to repair. Windows once bricked itself (i couldnt even use the Installation DVD to repair it) and i had to reformat the C:\ drive outside of the machine because it wouldnt even recognise the partition the OS was on.
I’ve also started having these odd issues with drive letters and paths - i run most of my music and video libraries off external drives, and windows seems to consistently change the paths from upper-alphabet letters (H:\, I:\, K:\, etc) to lower-alphabet ones (W:\, X:\, Y:\, Z:\, etc). Changing them back in Disk Management doesn't help - windows always switches them back on reboot.
My internal secondary hard drive often disappears from my computer too, prompting me to scan for hardware changes in order to re-find it. This has been recently happening to the local C:\ drive as well, but windows typically locks up and displays per-program file access errors before i can do anything about it.
Each BSOD (when they generate minidump files, which is now a rarity), traced through Bluescreen viewer, seems to point to a random array of drivers and files in the system folders. I’m beginning to think these have less to do with the causing the crash, and more to do with just being the files that were accessed when the error occurred.
I’ve tested my hardware for errors for going on a year and a half now, and they all still report fine. I ran a memtest pass post-boot for 15 hours, getting through 3 passes with no errors. My CPU and GPU are often put under strain and never seem to cause issues (oddly preventing them - rendering scenes for a short film took four days non-stop, and it didn't BSOD once). The HDD’s have perfect health in short tests, longer tests don't seem to complete before a BSOD interrupts it, which recently made me think.
I ran a SMART test on both internal HDDs (via HDD scan), and the only issues it brought up were alert signs next to “UltraDMA Cyclic Redundancy Check Errors”, which Google said represented “The count of errors in data transfer via the interface cable”.
Seeing as my PC is quite high-end, and and such all the cabling has been pleated and hidden away - would all these issues be pointing to damaged cables connecting the hard drives to the motherboard? I gather being stretched and tugged wouldn't have been good for them during the initial build.
Thoughts? Sorry if this is phrased wrong, doesnt have enough info, or posted in the wrong place, i’m new :)
I've attached as much info as i could from the SF Diag tool, it couldn't find any minidumps for some reason, but i have a locked folder in my secondary hard drive filled with 139 megs of large .dmp files, if that would help.