Plugging PSU into wall or multisocket

[
nope ;/ sadly. i don't think it's a power problem.

Which items do you know - without doubt - are good? If the list is empty, you have done lots of work without any accomplishments.

Your solution starts with collecting numbers - not changing anything. For example, what are the voltage in the purple, green and gray wires (from power supply to motherboard) both before and when the power switch is pressed. You currently have little idea how much additional facts are in those numbers. Post them so that others with significant experience can report those numbers.

Also critical are voltage on any one red, orange, and yellow wires when the system is fully loaded. Prime95 does provide a maximum load - is not sufficient. A maximum load is when you are playing complex graphics (ie a movie), while downloading from the internet, while playing sound loudly, while searching the hard drive, while reading a CD rom, while powering a USB device, while ... Multitasking is the heavy load - not Prime95. Now measure any one red, orange, and yellow wire.

Those numbers from six wires will result in answers that are definitive. Then you have items on your list of accomplishments.

Do not swap any parts or change any voltages. Leave everything as is. First collect the important numbers. If your computer is from the more responsible manufacturers, then execute the comprehensive hardware diagnostics that execute without Windows. More information that combined with those voltage numbers may even identify the problem in the very next post. Again, you don't know how much information is in those numbers - yet.

Your power supply must output the same voltage when AC mains are at 260 volts or at 180 volts. All those voltages are ideal to every computer. All those voltages cause no change to the DC voltages. Anyone who did not know that would make silly recommendations such as plug it directly into the wall. If your power cord was causing low voltage, then it is also a fire hazard. Ignore the myths. Get numbers. Post them. Have a useful reply immediately.
 
westom

thanks for your help but to be perfectly honest the chances of these stability issues being down to a brand new Corsair 850W Modular PSU are extremely low. Plugging directly into wall hasn't made any difference to stability as far as I can tell. Surely if a PSU successfully boots a PC, runs games and windows generally etc then it's fine? I doubt any crashes during heavy CPU usage (I do 3D rendering, probably the most intensive task a CPU can take) will be down to that. More likely to be temps or an unstable OC.

I've also tried a couple of power cords, so that's not the issue.

Cheers,
 
westom thanks for your help but to be perfectly honest the chances of these stability issues being down to a brand new Corsair 850W Modular PSU are extremely low.
TBH, if it's brand new, you'll probably have more problems with it now than at any point in the future, assuming the bathtub curve and all... I had something similar with an Antec PSU a year ago.
Having said that, I couldn't think of what else it could be currently...:(
 
Surely if a PSU successfully boots a PC, runs games and windows generally etc then it's fine? I doubt any crashes during heavy CPU usage (I do 3D rendering, probably the most intensive task a CPU can take) will be down to that.
Normal is for a defective power supply to boot and run a computer for months. Then many months later, strange events occur. Does not matter how many watts it has. Furthermore, a power 'system' is more than just a power supply. Those numbers report on all supply 'system' components - not just the power supply.

A long list of reasons - all in the supply 'system' - could explain your symptoms. I will not even begin to list them. Did you even know about those other 'system components? Without those numbers, I cannot provide any help and you still do not know about the 'system' integrity. Voltages from those six wires will either identify a suspect or exonerate the entire 'system'. Only then do you have some items on a list of 'definitively good'. Only then are you making progress.

Your struggle to zero in on the failure is determined by the length of that list. If you do not understand why, then get the meter to make those measurements. Again, if you keep making assumptions, then one who vast knowledge cannot provide anything useful.

Your only other alternative is to keep replacing parts until something works. That's it. What you have done has even caused you confusion and erroneous speculation (ie if the computer boots, then a power supply is good). May have even exponentially complicated your problem. If you don't understand why, well, then you have little idea of the power and usefulness reported by only six wires.

Again - because it should have caught your attention the first time: "More information that combined with those voltage numbers may even identify the problem in the very next post. Again, you don't know how much information is in those numbers.". You assumed those numbers are reporting stability rather than learn what is really being measured. Don't assume. Not yet learned is the information contained in those numbers. Did you also know about the power supply controller? If not, then why are you assuming? You provided zero reasons to believe that Corsair is good. It is neither good nor bad; it is still undefined.

Long before using Prime95, you should have known it would not report anything definitive at the system level. Long before you plugged directly into the wall, obvious, it would not change anything. Normal is for a defective power supply to boot a computer without failure for months. Why? This post is already too long and those reasons too numerous. But then one who really does know this stuff has suggested how to get useful answers almost immediately. Each reply will only be as useful as the numbers you provide.
 
That's pretty sarcastic westom for a request for information, in the same body of text you could have outlined what unusual voltages on the supply lines imply.

But, in the interests of using your vast body of knowledge, here's mine.
All voltages test out ok when not connected to a computer. All of them, done with multimeter by borrowing an electrical engineer. The 8 pin atx power does not run a system, however the 4 pin does. Both report 12/0V as you'd expect. How do I determine further which is the issue?

It also doesn't boot a system at all these days, but I think this is a contact issue with the atx pins rather than anything else.
 
All voltages test out ok when not connected to a computer. All of them, done with multimeter by borrowing an electrical engineer.
Defective power supplies can test just fine when out of the computer. Essential to better testing is to not remove or disconnect anything in a computer. Power supply must be under heaviest load to see the defects.

Also not posted are three digit numbers. Saying those numbers are OK is equivalent to knowing the sky is falling. Without actual numbers, the important information is lost.

Why are power supply testers a useless waste of money? At least four reasons. IIt tests the supply disconnected from a computer. No loading. Rest of the power 'system' also is not connected. And no numbers.

How to obtain a useful reply was provided. Your replies will only be as useful as the information you provide. No three digit numbers, as requested, means no information. A useful reply is not possible.
 
Last edited:
Surely it would have to be a massive surge to do that?
View numbers on that fuse. For example, a fuse rated for 250 volts cannot interrupt a higher voltage. The blown fuse continues to conduct electricity until that greater than 250 volts is interrupted elsewhere.

What happens if you use a 32 volt automotive fuse on a 230 volt plug? The fuse may continue conducting electricity after excessive current blows the fuse. A fuse could never stop a surge. Surge voltages exist where something tries to stop that current - typically thousands of volts. Even a tiny surge would blow right through a fuse and keep going.

That was reason one - voltage. Reason two - time. 300 consecutive surges could pass through a fuse before the fuse even considers blowing. Fuses take tens of milliseconds (or longer) to open. Surges are done in microseconds. A second reason why fuses cannot provide surge protection.

Meanwhile, that is irrelevant to the OPs problem. Facts remain - anything that might reduce AC voltage either is a fire risk (large voltage drop) or does nothing harmful (tiny voltage drop). A power strip, extension lead, or power cord is completely irrelevant to the OPs problem.
 
Cool. The reason I asked is that my i7 bundle is still unstable, even after upping voltage to 1.3 (and now 1.35) from OCUK's 1.275, which they tested and claim was stable at their end. .

lol quite a lot of people seem to have unstable clocks that ocuk have claimed stable.
 
Back
Top Bottom