W7 computer freezing (screen is like old analog television, frozen sound and HDD led lighting)

postcd

Weaksauce
Joined
Nov 24, 2016
Messages
94
Hello, on my desktop computer (detail) i turn it on sometimes to play for an hour or two game, but in maybe 40% cases (every other day or so) happen that computer freeze during gameplay (seen no freeze during game not running, though it is minimum time computer running without game):

- screen displays noise like old analog TV when no signal is found (same thing happen with this particular screen when certain device is turned on/off - i mean it is some electricity phenomenon/over-current spike or such, the screen have to be turn off/on)
- sound is frozen (not in a loop but single sound is heard)
- HDD led is on, constant lighting (is it creating dump or such?)

Windows event log shown:
event 7011 - Service Control Manager - A timeout (30000 milliseconds) was reached while waiting for a transaction response from the AMD External Events Utility service.
event 1001 - BugCheck - The computer has rebooted from a bugcheck. The bugcheck was: 0x0000001e (0xffffffffc0000005, 0xfffff880055ca4e8, 0x0000000000000000, 0xffffffffffffffff). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: . (NOTE: MEMORY.DMP is 588MB big, dumpchk.exe not found)
event 1001 - BugCheck - Unable to produce a minidump file from the full dump file.
event 41 - Kernel-Power - The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

on previous Windows install it not happened, i had different system HDD. Maybe connected via different cable to different SATA socket.

i have already tried to:
- use different power cable connecting to the different wall socket, not to the UPS
- update GPU drivers using SDIO utility (i think this happen across 2 or even 3 such updates in past weeks/months)
- run memory check (i think it was OK if i remember correctly)
- run disk check (chkdsk /f /r), i think it was ok, SMART is all good, no uncorrectable
- sfc /scannow

i see that the GPU is powered through one power connector to which goes two cable branches from PSU (PSU has only two, remaining goes to MB), one is also powering small system 3.5" HDD and other goes to a PC status leds panel (just two cables, others from panel goes to MB not to PSU).

ideas: temperature? i have not yet tried to find a way to record it in time so i see if something is overheated during crash
 

pendragon1

Extremely [H]
Joined
Oct 7, 2000
Messages
43,888
use hwinfo to monitor your temps. make sure the guts/fans are clean. try running DDU and reinstalling the newest drivers from amd, not whatever sdio is.
 

GiGaBiTe

2[H]4U
Joined
Apr 26, 2013
Messages
2,406
I would go with the PSU failing. While FSP isn't terrible, they aren't great either, and 350W is just barely cutting it for your system. The machine crashing under load is a very common symptom of a failing power supply.

Change out the PSU with something in the 400-500W range from a reputable manufacturer.

FSP has a habit of using shitty capacitors, I bet if you looked in the fan grille on the PSU, you'd see some capacitors that are inflated/leaking.
 
Top