so i have had my kraken liquid cooler for a good amount of years and my temps are creeping higher and higher and now my system has frequent freezes and bluescreens and the tamps are quite high like idle at 60's and it only bluescreens in heavy games it has to be said i do have an i9 9900k which i have heard is especially toasty, so what should i do?
#Liquid CPU cooler dying?
1 messages · Page 1 of 1 (latest)
idle temps don't mean anything, what are the temps when you game?
set up a monitoring program and record it
have not checked in quite a while any program you recommend for that? and what sorta game is cpu intensive?
and can temps even cause bluescreens?
HWINFO64 is the gold standard I think, for system monitoring in windows
i do have that installed tho i dont know how to setup monitoring
i have also had to underclock my cpu to 4.4
It just prompts you for a file to save log data to
i dont see that button
found it
okay so now i have started it, should i then just start up a game?
Intel XTU is another decent monitoring tool since you have an intel cpu
Do you have a second screen?
technically yes, but not really as i dont have space for it
So, it exists, but it's just not connected, ok
yes
It'd just make monitoring easier, to put the temp sensors on one screen, game on the other
You can, but you won't see them until later when you look at them, so there's no real time data
I'm sure some software does this with an overlay, but I don't know about that
Anyway, leave the log running and start a game that causes crashes.
well its intermitten had it happen yesterday and today with two different games
so i dont really know when
and if it crashes wil i just not loose the log?
the log gets saved constantly, it'll record things right up until a crash
well i went into the bios and changed the speed from first 4.7 which is started at to first 4.6 for a while and now 4.4
ah thats nice
yeah.. that's how logs work :)
Intel says base clock for 9900K is 3.4, max boost is 5.0
i guess then im gonna go back to bios and set it back up where it was and then just start playing, see if it crashes, not really much else i can do then
yes but it has always ran at 4.7 since i got it
Well if you had it for a few years, it might be time to replace it? 🤔
Since it might get clogged up after a while
Ah yep, it'll only boost as high as it has room for, since it's measuring the temperature
but can temps even cause a bluescreen?
usually no, if a cpu overheats it just shuts down, no bluescreen
should i maybe set the boost clock to auto if that exists so it maybe can underclock itself if it needs to?
I'd try auto, yep
is there a way i can see why it did a bluescreen?
im not super technical
i have heard about a eventviewer but i dont know if thats it
Every bluescreen has a code on it, it's a vague error reference that you can look up
It's vague, but it can point you to the right direction
damnit, im gonna try setting it on auto to begin with, and if it keeps bluescreening then ill take more drastic meassures i guess
cause only happens when i game in intensive games
Event viewer is another tool that you can look at, to figure out what's going on.
yeah i just have no idea how to read that xD
in event viewer specifically you would want to go to windows logs > system
Basically it's a windows internal tool that lets you view events. And events are programs or system misbehaving.
hmm i will look into it then
There's usually not much info in event viewer, but if you have a critical error with code 41, that points to a sudden power loss. So either the psu failed to deliver enough amps or the cpu shut itself down.
you can sort the events by date and time, and look at what happened when you crashed.
I wouldn't worry about the event viewer now. First check your temps while you have a game running.
You can use that to view minidump files
or upload the minidump here (it's tiny), for us to have a look at it
We're mainly interested in the error code in the bluescreen
plus the circumstances when it happened
okay i got the program what now
open it, the default location should work to view your minidumps
these are the two i had in same file one today and one yesterday
fail wrong file
there we go xD
you've tested the ram with memtest86?
no? i mean it cant be the ram they are practically brand new
so, I had rma'd ram that lasted two days before it started throwing errors.
and that was the second rma of that set.
if its the ram then that would be after the rma ended
Windows threw up an error about memory, so that's your clue. It doesn't matter if it was bought yesterday.
will the memtest program show if its that
ram is technically lifetime rma
It might, memtest takes a long time to run though, so do it overnight
how do i even open the program
download it unzip it use the program to "burn" it to usb run the usb like a windows install usb
could it be that its ceuse i have 2 sticks in that are there only to add rgb, like they have no actualy ram, its something i got with my old pair of ram purely for astetics could that be screwing with it
nope
There's a readme.txt as usual, with full instructions.
i dont have a usb to burn
they are cheap ¯_(ツ)_/¯
how did you even install windows without a usb..
well, it's indicative of a memory issue
well i had one, it was a sd card into a sd card reader usb but i cant find it and have not been able to for about 2 years
running memtest would confirm or deny the assumption
But your issue sounds like a memory problem, random crashes in apps that are memory intensive.
got damn it all, i guess ill look for it, but if not them im kinda screwed
usb sticks are like <$10 no?
yeah that was my they are cheap comment. heh
you should have a couple at hand anyway, just in case your pc shits itself and you need to reinstall
first of all im in denmark so + 25% second its more the principle of getting something i already should have somewhere and it maybe not even being it
And I'm in Finland, don't talk to me about expensive electronics :P
it's still under 10€
Anyhoo, my suggestion is, run the memtest, you'll know more after it's done.
Random crashes are tricky to pinpoint, but that's our only clue right now.
If you crash again, look up the hwinfo log as well, and see what the temps were.
yea