[SOLVED] Random crashes with 3 different RAM kits

Oct 20, 2020
9
0
10
My PC started crashing a few months ago. It would just freeze and power off and then turn back on. Sometimes it would give me the bsod. I attributed it to faulty memory, so I ran memtest, and one of the modules didn't complete the test. So I RMA'd it and in the meantime I was given a back up kingston memory. The back up memory worked fine and then 2 weeks later it started crashing as well. But by this time, my replacement memory was available which was brand new. Fast forward 40 days later PC started crashing again, this time no bsod, just freezes. So the odds of all these memory sticks suddenly failing was way too low so I thought maybe something was wrong with the motherboard. So I got a spare board from a nearby store and tried it out. Still no improvement. I've tried disconnecting all my hard drives except the boot drive, disconnected the LEDs, and makes no difference. Any ideas on what could be the potential issue here?
Sometimes these crashes happen during the bios screen, and sometimes during usage, and sometimes it doesn't even get to the bios screen.

Specs
OS: Windows 10 64bit 2004
Processor: i5 6500
Motherboard: gigabyte B150M-D3H (F22 bios version)/Biostar B250
GPU: MSi GTX1060 Gaming X
RAM: team group vulcan 3000MHz 2x4GB/kingston 1x8GB/ team group dark 3000MHz 2x4GB
PSU: FSP Hexa+ 550W
Storage: Kingston UV300 240GB(boot drive), Seagate barracuda 2TB, Western digital blue 2x1TB
Case: Asus TUF GT501
 
Solution
As a matter of elimination, based on all that you have done, the CPU could be the culprit.

The other thought is, just by simple bad luck, the new PSU is not up to par.

Next time you boot, go into BIOS and change the POST reporting to be as verbose as possible and to allow time for you to read the screens in the interim.

Knowing what POST is doing and reporting may prove helpful.

Probably worth the effort to pull, clean, reseat, and put new thermal paste on the CPU.

Check CPU, socket, and surrounding motherboard area for signs of damage and overheating.

Noted that you cleared CMOS - install a new battery - also worth a try.
Oct 20, 2020
9
0
10
Look in Reliability History and Event Viewer for error codes, warnings, and even informational events that correspond with the crashes, etc..

PSU:

https://shopee.com.my/FSP-hexa-550W-power-supply-i.4190965.1283539852

How old is that FSP PSU?

Has the PSU been heavily used for gaming, graphics work, maybe even bit-mining?
It's slightly over 4 years old. And it hasn't been heavily used for mining. But the pc has been used almost every day since purchase. And specially this year due to quarantine it's turned on for like 15-16h a day
 

Ralston18

Titan
Moderator
In my mind, all that makes the PSU the primary suspect.

Especially if the PSU has survived 4+ years.....

Like most products these days, PSUs have a designed in EOL (End of Life). Long use at high power demands will just bring that EOL all the sooner.,

Do you have access to another PSU that you can install and test?
 
Oct 20, 2020
9
0
10
In my mind, all that makes the PSU the primary suspect.

Especially if the PSU has survived 4+ years.....

Like most products these days, PSUs have a designed in EOL (End of Life). Long use at high power demands will just bring that EOL all the sooner.,

Do you have access to another PSU that you can install and test?
I can test it using a friend's PSU. But if the ram is already fried, it probably wouldn't make a difference unless I can get my hands on some new ram as well.
 

Ralston18

Titan
Moderator
What error codes, if any?

Would not be sure about the RAM being "fried" per se. Maybe the modules were simply not getting the correct voltages.

Test with your friend's PSU but start with the old RAM.

Key is to determine if the test PSU results in system stability - no crashes, etc..
 
Oct 20, 2020
9
0
10
What error codes, if any?

Would not be sure about the RAM being "fried" per se. Maybe the modules were simply not getting the correct voltages.

Test with your friend's PSU but start with the old RAM.

Key is to determine if the test PSU results in system stability - no crashes, etc..
"WHEA_UNCORRECTABLE_ERROR"

"UNEXPECTED_KERNEL_MODE_TRAP"
These showed up a long time ago, during the first few bsods
 

Ralston18

Titan
Moderator
Oct 20, 2020
9
0
10
Multiple occurrences of each error - correct?

Any number codes?

First error suggests hardware:

https://support.microsoft.com/en-us...le-error-7c49d78a-2792-96cf-2268-abbe9d9eb29f

https://answers.microsoft.com/en-us...bleerror/8f95262a-8c31-4ff7-ac9c-ab22d3eb4c3c

If there are varying hardware errors the common factor is the PSU.

Take a look at both links. Do you see any other patterns or indications of specific problems?
Yeah, multiple times.
I think there were number codes once or twice, but I didn't take screenshots of them so I don't have them, sorry.

I did go through the given links long time ago, but that did not seem to resolve the issue. However, the time between failures increased. For example if I leave the computer turned off for a while, then when I turn it on, it stays on for like a few minutes. But once it shuts down the first time, it doesn't stay on much longer, with subsequent boots being shorter to the point where it just turns on and off sometimes
 

Ralston18

Titan
Moderator
"Time between failures increased"?

Did you perhaps intend to use "decreased"?

Or do you mean that you must allow more time between failures before a reboot succeeds?

Works while cool and, once warmed up, failures start.

The first failure occurs and you reboot. However, the system is already warm and after the reboot simply takes less time to hit whatever threshold temperature is causing the crashes.

I am still leaning towards the PSU being the culprit. But need to find a way to prove that either way. No vested interest....

Do you have a multi-meter and know how to use it? Or a family member or friend who does?

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

In any case, just reading the link may prove helpful to you based on the overall problem history.
 
Oct 20, 2020
9
0
10
"Time between failures increased"?

Did you perhaps intend to use "decreased"?

Or do you mean that you must allow more time between failures before a reboot succeeds?

Works while cool and, once warmed up, failures start.

The first failure occurs and you reboot. However, the system is already warm and after the reboot simply takes less time to hit whatever threshold temperature is causing the crashes.

I am still leaning towards the PSU being the culprit. But need to find a way to prove that either way. No vested interest....

Do you have a multi-meter and know how to use it? Or a family member or friend who does?

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

In any case, just reading the link may prove helpful to you based on the overall problem history.
Nope. I did not mean decrease, meant increase.
"Time between failures increased"?

Did you perhaps intend to use "decreased"?

Or do you mean that you must allow more time between failures before a reboot succeeds?

Works while cool and, once warmed up, failures start.

The first failure occurs and you reboot. However, the system is already warm and after the reboot simply takes less time to hit whatever threshold temperature is causing the crashes.

I am still leaning towards the PSU being the culprit. But need to find a way to prove that either way. No vested interest....

Do you have a multi-meter and know how to use it? Or a family member or friend who does?

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

In any case, just reading the link may prove helpful to you based on the overall problem history.
I checked my ram on a different PC. It worked fine.
So I bought a new PSU and tested it. It ran okay for a while. And then again the same problem.
 
Oct 20, 2020
9
0
10
New PSU: make, model, wattage?

How long was "awhile"? Same errors as before?

Try running "sfc /scannow" via the Command Prompt.

Reference:

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161

Then maybe "dism".
I removed my GPU and tried, hasn't crashed a single time so far. I can't believe I missed that. But the issues don't even correspond towards a GPU problem. I'm planning on running some tests on the GPU on a friend's bench, and I'm gonna try the 750Ti I have lying around on mine to see if any issue pops up. I'll update after the tests.
Thanks a lot for the support so far :)
 

Ralston18

Titan
Moderator
Very good. You are welcome.

With what appear to be intermittent and varying problems try to remain as methodical as possible with respect to the swapping/testing etc.

Change only one thing at a time and also put an interim pause on Windows OS or other software updates.

Keep an eye on the error logs and also system performance via Task Manager and Resource Monitor.

Hopefully there will be a "Eureka" moment sooner or later.....
 
Oct 20, 2020
9
0
10
Very good. You are welcome.

With what appear to be intermittent and varying problems try to remain as methodical as possible with respect to the swapping/testing etc.

Change only one thing at a time and also put an interim pause on Windows OS or other software updates.

Keep an eye on the error logs and also system performance via Task Manager and Resource Monitor.

Hopefully there will be a "Eureka" moment sooner or later.....
I plugged in my GPU again and using afterburner, set power limit to 80%, it crashed after a few hours. I then tested it at 60% power limit, and it's been up and running for days without crashing. I'm not sure why this is happening though, as now I'm leaving quite a lot of performance on the table.
 

Ralston18

Titan
Moderator
That is progress.

Take a closer look at your system's power demands.

If that 550 Watt PSU cannot keep up then you may be able to improve performance with a new, higher wattage PSU.

What games do you play and what configurations (overclocking) do you use? Double check recommended/required hardware game specs with regards to your system.

Again use either Resource Monitor or Task Manager to observe system performance.

Work your back up from 60% towards 80% and then 100% as incrementally as you can. With and without afterburner etc..

Watch what resources are being used , to what extent (%), and what is using the resource. When do the crashes start with respect to any given resource?

It may well be that the only way to regain the required performance is a new higher wattage PSU.

Overall, continuing to delve in a bit deeper will help you learn more and that in turn will help you regain performance.

How much gain? TBD....
 
Oct 20, 2020
9
0
10
That is progress.

Take a closer look at your system's power demands.

If that 550 Watt PSU cannot keep up then you may be able to improve performance with a new, higher wattage PSU.

What games do you play and what configurations (overclocking) do you use? Double check recommended/required hardware game specs with regards to your system.

Again use either Resource Monitor or Task Manager to observe system performance.

Work your back up from 60% towards 80% and then 100% as incrementally as you can. With and without afterburner etc..

Watch what resources are being used , to what extent (%), and what is using the resource. When do the crashes start with respect to any given resource?

It may well be that the only way to regain the required performance is a new higher wattage PSU.

Overall, continuing to delve in a bit deeper will help you learn more and that in turn will help you regain performance.

How much gain? TBD....
Well, I'm back. After no issues for quite some time. Now it's even worse with PC not booting or barely booting. Sometimes it turns off before getting to the bios screen, sometimes after, sometimes it boots into windows and then turns off. All of this happened one night after a perfectly working PC. I'm completely stumped. I don't even know what to test at this point. Removing the GPU and using onboard graphics didn't help either - same result. Clearing CMOS didn't help. Nothing seems to be working.
Edit- I just breadboarded the system. Same issue. I'm inclined to believe it's the cpu causing the problem as I have tested the RAM, PSU and used a different mobo. The only component I haven't tested is the cpu, as it's hard to come by a 6th gen cpu these days.
 
Last edited:

Ralston18

Titan
Moderator
As a matter of elimination, based on all that you have done, the CPU could be the culprit.

The other thought is, just by simple bad luck, the new PSU is not up to par.

Next time you boot, go into BIOS and change the POST reporting to be as verbose as possible and to allow time for you to read the screens in the interim.

Knowing what POST is doing and reporting may prove helpful.

Probably worth the effort to pull, clean, reseat, and put new thermal paste on the CPU.

Check CPU, socket, and surrounding motherboard area for signs of damage and overheating.

Noted that you cleared CMOS - install a new battery - also worth a try.
 
Solution