Question Pc randomly crash cpu and vga light flashes red on motherboard.

Jun 19, 2021
2
0
10
0
Can somenone please help me.


So I built my first pc 2 days ago, everything worked fine both gpu and cpu were sitting on around 70c when gaming.
Then suddenly when i was just browsing the internet the pc just crashed and the vga and cpu light flashed red on the motherboard.
The pc restarted and no damage was done. After that it started crashing frequently and Ihave no idea how to fix it, do you have any suggestions?
ive updated my bios, updated gpu drivers and unisnstalled old one ive tried everything
(crashed 2 times while writing this)



Pc specs
mobo: msi tomahawk b450 max ComboAM4PIV2 1.2.0.2
cpu: r5 3600
gpu: asus matrix r9 290x (bought second hand from facebook marketplace)
ram: corsair vengeance pro 2x8gb 3600mhz xmp enabled
psu; corsair cx650m
 
Last edited:
Do you have a cooler on the cpu or is it just the fan supplied with the cpu? Did you use thermal paste when installing the cooler? Do you have any entries under computer management critical incidents to coincide with the shutdowns? Try using ram without xmp activated to see if it clears up.
 

David0ne86

Upstanding
Mar 11, 2021
418
110
390
10
Yeah i would look to windows log to see the nature of the crash. Try to type "events" in the windows search bar (or after you pressed the start button key). Open the event viewer, windows log and then system. The crashes should be marked with a white x on a red circle called critical.
 
Jun 19, 2021
2
0
10
0
Yeah i would look to windows log to see the nature of the crash. Try to type "events" in the windows search bar (or after you pressed the start button key). Open the event viewer, windows log and then system. The crashes should be marked with a white x on a red circle called critical.
It fixed itself for 3 weeks but came back again I have no idea why. It says kernel power what does that mean?
 
Last edited:
Jul 8, 2021
3
1
15
0
eeehhhhhh

somehow this description of the error sounds very familiar to me :rolleyes::tonguewink:
do you have HWinfo/Aida or any other Systeminfo Software?

the green text is maybe not really important (just about i wasted 4 weeks of my life with corsair ram)

At the end of 2019, after 10 years of abstinence from windows computers, I ordered cheap but very good reviewed last gen parts with high OC
potential.
The "16gb/2 Corisair Vengeance LPX 3200 CL15" worked at a max of 2866mhz CL19 (to start the computer) and causes sporadic crashes.
sometimes every 3hours... sometimes after20min... 4/5/6 times in a row ^^ other people OC em 20% without a single crash!
and my vengeance ram always boots with Jedec settings (2166mhz) if xmp was activated. dont matter witch xmp profile was set +the ram
voltage spikes way to high, especially 4 this clockspeed!

lucky amazon sends me a canceled order with 16gb/2 HyperX3200 4 weeks late :love: without charging
:smilingimp: 3 days later i got 16gb @3600mhz CL18

sry for this less important story, but i already cant belive kingston hyperx runs like daredevil on crystal meth on my mainboard and the vengeance lpx
running in the computer of my roommate with Ryzen ?2600x? with a 2933mhz ram-controller @ 3400mhz (asus mainboard^^)
But Maybe it helps anytime a frustrated guy, like me in 2019. At that time nobody could help me and since other ram worked for me and the vengeace on every intel and most amd (except b350, z290 and b450) it was clear that the memory is not defective either ^^

First: I would always disable xmp! Because the auto voltage is always much to high and the range much to big. Sometimes this Voltage drops/spikes can cause safety shutdowns with the msg: kernel power ID:41 ;-)

If the voltages/offsets/limits whatever, adjusted manually you can safe a lot of hardware lifetime, increase performance/stability AND
it's easier to detect/describe malfunction's "if you know how what is set,".

If you wanna try it, set your preferd clock speed manualy in the OC menu.
https://flic.kr/p/2madq3o If you are not the unlucky one of thousand how catched minus performce cpu sample , clocks up to +3666mhz should be easily possible, if your mainboard also likes the dimms you have!!!! ;-) But OC are your problems of tomorrow :p

-->extended DRAM configuration:
a) be sure the rates in automatic mode
b) your bios sould have the same look like mine -> at the end of this page (pic 3)
gear down "on" - power down "off" - tsme "off" - dram latency enhance mode..... pfff until your system is not stable "off" ;-)

https://flic.kr/p/2maiC15 https://flic.kr/p/2maiC1v
sooo now i got some "homework" for you:ROFLMAO:

WindowsPower Plan settings:
Balanced or better Performce mode until hunting the error 🔫
https://flic.kr/p/2maf5r7
screenshots from a Sysinfotool like HWinfo would be awsome (vcore/vsoc/dimm voltage/gpu socket power | one while computer idles on desktop 1-2min after reboot & with a running game/benchmark running in windowmode/backround)? :ouimaitre:

Or ideally, recorded log files from both situations ;-)
HWinfo 7.06 heise.de direct DL - virus checked - no adware

https://flic.kr/p/2maaspE
How many Monitors connected?
How many external devices connected?
CPU Overclocked?
Graphiccard Overclocked?
Any system-boost-software installed? (razer cortex activ?)
Any energysaving tools/options/software active?
wenn exactly crashes the system? while browsing? after computer idles?
wich software running in background?
full partnumber f DRAM? like this -> Corsair CMK16GX4M2B3200C16

It says kernel power what does that mean?
to protect the hardware the system shuts down if a voltage drop happens for 0.X miliseconds at any device or the needed voltage for any action of a device dont reached fast enough!
Until feb/march 2020 i used a 630W powersupply wich was enough for runnig the system in all situations. But sometimes there a tiny little spikes in the powerusage of some
devices witch detected as thread 4 the hardware = shutdwon
or they are really to high for the system (Mainboard/Powersup)
since i got a 730W i never had this error again, caused from the system.
bad amd/win10 updates in april/may are the last known reason for that i remember (because of ghost workloads without an programm is doing anything)

but the power consumption of my gfx since the last win10 update rises.... hmmmm.... detected thru 20°C higher average Temp of my rx5700xt!
full model description of your gfxcard und driver version? :ROFLMAO:

exsample: My graphiccard is software limited to a max auf 240W allowed. But spikes up to +350W over a timeo 1,2,3ms can happen
in idle they ned just 10-15W... if irq (request to hardware from software) drops in (starting a game + tabbingin out the gameto browser where a youtube was until this point in backround =
the gfx sucks immediately with the maximum the cable can transport from the powersup. but after 2ms the radeon driver says no dude.. max 230W!!!
but the cpu, ram Mainboard soundcard, southbridge, ssd, fans doin the same, at nearby same time..
and everytime its the same like a driving car.... my skoda just neds 7L/100km. but while accelerating 0-100 km/h with a 100% pressed gaspedal the motor sucks with a average usage auf +12L/100km ;-)

sry its way to hard to building a description for things like these & translate in english for me, at the same time :homer:


If I have run over you too much with my novel and you are not afraid of help via remote connection, I can also help via Teamviewer.
-> easy remote tool with chat and paint on desktop function


greetz, Curtis


ps: hopefully there are no dramatic errors in the whole description:
copy text - change - rewrite - move text - rewrite - insert 5 things at the same time 🤣:homer:

To be honest, I'm too lazy to read it all over again
 
Last edited:
Reactions: Fix_that_Glitch

ASK THE COMMUNITY

TRENDING THREADS