How can I make sure my PSU is faulty? is there a stress app for PSUs?

[Vayra86]

IMO if you really want to know if a PSU drops out of ATX spec is to buy a volt meter and probe while stressing the PC. Software readings can be off, and this is the only way to really determine PSU failure before it actually does fail.

I say this because I dont see a problem with temperatures, and even if it is NOT the PSU this method will allow you to exclude the PSU and zoom in on other possible causes. Additionally having such a tool on hand doesnt break the bank and will prob prove useful to have in the future.

 

[Dippyskoodlez]

The best overall load ive managed to find via my kill-a-watt was a loop of 3dmark extreme combined test.

IMO if you really want to know if a PSU drops out of ATX spec is to buy a volt meter and probe while stressing the PC. Software readings can be off, and this is the only way to really determine PSU failure before it actually does fail.

Dmm wont see ripple, only voltage spikes, and if its a multi rail psu, you wont be able to see all of them. Good for a single niche out of spec scenario only

 

[Master]

Just finished running memtest. two passes and no errors thanks to dear God.
@newtekie1 : I swapped ram sticks as well, performance seemed to have improved! thanks
by the way is my +5v,+3v and 12+v stats OK? I see +12v has alot of fluctuations 11.6 to 11.9! and mostly staying around 11.7 11.8!

 

[FordGT90Concept]

Did the computer restart itself or just go off?

Have you disabled "automatic restart" on system failure?

There could have been a BSOD with more details but Windows' default behavior is to restart so you can't see it. A BSOD can log a Kernel-Power 41 event after the automatic restart basically telling you something broke but not giving any details as to what or why. Instructions how to disable the automatic restart are at the bottom of that page.

Hell, even an interruption in power to the PSU can cause a Kernel-Power 41 event log.


Windows basically has a "proper shutdown" bit on your HDD/SSD which is set to false when you start Windows. When you shut down Windows properly, it sets the bit to true. If the computer shuts down for any reason that the bit doesn't get set to true, when Windows next boots, it creates the Kernel-Power 41 event log indicating Windows itself had no control over the shut down (hardware fault, power loss, or as simple as holding the power button in for 5 seconds).

 

[Master]

Did the computer restart itself or just go off?

Have you disabled "automatic restart" on system failure?

There could have been a BSOD with more details but Windows' default behavior is to restart so you can't see it. A BSOD can log a Kernel-Power 41 event after the automatic restart basically telling you something broke but not giving any details as to what or why. Instructions how to disable the automatic restart are at the bottom of that page.

Hell, even an interruption in power to the PSU can cause a Kernel-Power 41 event log.


Windows basically has a "proper shutdown" bit on your HDD/SSD which is set to false when you start Windows. When you shut down Windows properly, it sets the bit to true. If the computer shuts down for any reason that the bit doesn't get set to true, when Windows next boots, it creates the Kernel-Power 41 event log indicating Windows itself had no control over the shut down (hardware fault, power loss, or as simple as holding the power button in for 5 seconds).

Thanks alot.
No computer restarts itself, usually it takes several seconds for it to restart . if I restart the system it doesnt act like that, I can almost instantly see the Asus logo when the system wants to boot up.
I just disabled it, it wasn't before.
Usually when a BSOD happens, the moment windows is restarted, a windows usually pops up saying a crash occurred, and you know the rest. but I never faced such window when I had these kinds of restarting!

 

[FordGT90Concept]

Next time it does it then, you should see a full BSOD and there should be a line on that with full caps and underscores that say what actual fault is occurring.

 

[Vayra86]

ATX spec 2.2
Voltage Rail Tolerance Minimum Voltage Maximum Voltage
+3.3VDC ± 5% +3.135 VDC +3.465 VDC
+5VDC ± 5% +4.750 VDC +5.250 VDC
+5VSB ± 5% +4.750 VDC +5.250 VDC
-5VDC (if used) ± 10% -4.500 VDC -5.500 VDC
+12VDC ± 5% +11.400 VDC +12.600 VDC
-12VDC ± 10% -10.800 VDC - 13.200 VDC

A bit of riople doesn't kill a system btw, so whether or not you can see this is irrelevant unless you're overclocking. Many cheaper PSUs have ripple and they run fine. However, dropping out of spec even briefly on +12V is instant death. In addition, ripple in my experience doesn't magically fall out of the sky, its inherent to the unit and its design, while +12 V can degrade over time.

 

[EarthDog]

All psus have ripple, not some.

Dropping out of spec isnt instant death, its instant power off or never powers on.

Ripple, like 12v voltage, can degrade over time.

 

[SnakeDoctor]

The power supply testers you buy from ebay(got one) are only going to tell you the Power supply switches on and show if the volts are 12,3.3,5v volt and that's all .Does not actually test the psu on load or anything
If the voltages are low yes the psu is faulty but normally unlikely
I more just use one to see if the power supply is dead or not , Check if any damages caps in psu

To test the power supply properly you will need the equipment like that of TPU uses . Check the equipment$$$ they use on testing power supply with load tests

Op - The voltages do look a bit low on the 12volt rail - 11.62 seems but low, but seems pretty stable
The voltage should not be greater/lower that 3% tolerance

 

[EarthDog]

The power supply testers you buy from ebay(got one) are only going to tell you the Power supply switches on and show if the volts are 12,3.3,5v volt and that's all .Does not actually test the psu on load or anything
If the voltages are low yes the psu is faulty but normally unlikely
I more just use one to see if the power supply is dead or not , Check if any damages caps in psu

To test the power supply properly you will need the equipment like that of TPU uses . Check the equipment$$$ they use on testing power supply with load tests

Op - The voltages do look a bit low on the 12volt rail - 11.62 seems but low, but seems pretty stable
The voltage should not be greater/lower that 3% tolerance

5% is the atx specification (see vayara's post two above yours), not 3%. Lower is better of course for stability, but so long as its within spec, it should still work. Outsode of that things get wonky and shut down or just dont work.

 

[Master]

Did the computer restart itself or just go off?

Have you disabled "automatic restart" on system failure?

There could have been a BSOD with more details but Windows' default behavior is to restart so you can't see it. A BSOD can log a Kernel-Power 41 event after the automatic restart basically telling you something broke but not giving any details as to what or why. Instructions how to disable the automatic restart are at the bottom of that page.

Hell, even an interruption in power to the PSU can cause a Kernel-Power 41 event log.


Windows basically has a "proper shutdown" bit on your HDD/SSD which is set to false when you start Windows. When you shut down Windows properly, it sets the bit to true. If the computer shuts down for any reason that the bit doesn't get set to true, when Windows next boots, it creates the Kernel-Power 41 event log indicating Windows itself had no control over the shut down (hardware fault, power loss, or as simple as holding the power button in for 5 seconds).

Hello every one, OK, last night I suddenly faced another restart. I had unchecked the automatic restart on system failure, so the system doesnt restart in case it was windows related. Yet the system was restarted, upon getting backed, I checked windows event viewer and again saw the infamous Kernel-Power41.
Since last night I' have tried to run some tests again to see if I can finally know what could have been causing it. I went to the bios and set the profile to performance, this profile seems to overclock my CPU (two cores 4600, one 4500 and the last one 4400 Mhz). I did that to see if its the temperature that caused the restart.
I started windows and ran OCCT PSU benchmark for two hours. It went well without any issues until it restarted just 2 minutes before the two-hour test is finished. I checked the temps , the highest was 99C in the very beginning but after an hour and a half which I checked again, it was around 82 ~85C . The +3V and +5V seemed fine. the 12V was fluctuating between 11.81 11.71 and rarely 11.61~11.62.
I tried it again, and once again around the same time , system was restarted.
For the third time I thought maybe its because of me having different RAMs (different speeds but from the same brand, Geil, two modules are Geil 1333 2G and the other two are Geil Evo 1600 8G), and the overclocking may have caused them to malfunction or something like that. So I removed the 2 Gig modules and ran the test again with 16 Gig of ram total. (I need to say that, I had already ran memtest and all modules were fine.)
Any way, to my surprise it restarted at around the same time as the previous tests .
Having acted like this, what could have been the cause ?
power seems to be working fine I guess right?

 

[EarthDog]

Bad ram? Try the other set alone amd see what happens..

 

[FordGT90Concept]

You don't have this system on a UPS do you? Perhaps the computer is fine but there's a slump in incoming power causing the computer to go off/restart.

You can verify this is the case by going into the BIOS and finding the option where it sets behavior for power loss. If it's set to "turn on" or "restart," change it to "stay off." When it happens again and the computer stays off, that confirms it's line power causing the issue.