A power supply tester can identify some power supply defects. But it can never report a supply as good. A power supply tester costs almost as much as a multimeter. And does not report useful facts. No numbers is a first obvious reason why a PSU tester is useless. How to obtain assistance from the few who actually know this stuff? Numbers are required. A PSU tester cannot even report a PSU as good. Puts nothing on a "definitively ..." list. Cannot elicit a useful reply. And creates no accomplishment. PSU testers are not useful. Tester cannot report any supply as good. Useful testing requires a major load (ie connected to a computer) and numbers. Tester has neither. Only a multimeter can report a supply and all other power 'system' components as good - without doubt.
I gfo with this, my grandpa recently needed to use the pressure software for his range rover, the serial port was bad so he got a new one and it fried the board... £5 well spent.
Now you're weeing me off. A power supply tester tests each individual rail and connection of the power supply to check for any faults with a certain rail or connection. Sure, a multimeter can obtain numbers of the voltage running through each rail and connection but that will hardly tell you anything about a computer's PSU. Please, stop calling any suggestions to at least try and solve the problem "useless" while you continue speaking of numbers, the OP is looking of ways to sort the problem. Try to be useful, you've suggested that at least TRYING something won't help, I'd get off a help thread...
Please learn how power supplies work. A defective supply with no load can measure OK. To locate a defective supply means it must be connected to a major load. The only load a layman has is a computer. If only connected to a tester, then failures get ignored. Multimeter numbers report many other 'defects' that a supply tester cannot see. Again, the supply tester can only identify a few type of power supply defects. Will not find defects that cause intermittent computer crashes and other problems. OP’s problem is not being solved because other suggestions are useless. May make the problem exponentially more complex. He had two choices. Keep replacing good parts and software until something works. Or identify the problem before fixing anything. He did the former. And is now more confused. Of course. Those diagnostic methods are tepid. A very next post can result in accomplishment if using a meter (not a useless power supply tester) to have numbers. Accomplishment means good or bad - without any doubt. A power supply tester can never do that. Reality does not change because you do not like it. A meter can test the many components of a power system in one minute. Do you even know what those many parts are? Your PSU tester can only test one - and not sufficiently. A meter is the tool routinely used by better trained techs. You can learn or you can become angry. Your choice. Does not negate a reality. Not using a meter has created confusion. And no solution. Reality does not change no matter how angry you become. That tester remains a nearly useless tool that can report defective power supplies as good. Can create confusion. Can make problems exponentially more complex. Why is that a good thing?
He does have a point, you can only really test a PSU properly if it is put under varying levels of stress, for instance the voltages may be within ATX or 80+ spec when it is at 50% load, but the more stress put on it, the less stable the voltage it may be able to produce. It's not easy to get the equipment that can do this without spending thousands.
And a good PSU tester will do this? Hook up each connection like the 24 pin and each will be pt under stress to check the voltages being given out? Still numbers? Right? This then allows you to check for fluctuating or unstable voltage under different levels of load.
Just for the record I've read most of this thread because one of my own machines has the same problem as the OPs and what westom is suggesting is not wrong. He may not be suggesting it in exactly the correct wording and may keep going on about what has been done wrong as opposed to what can be done right, but I believe his general line of thinking is right. I haven't yet gotten to the bottom of what is causing the power failure on my machine. I have a new, higher rated PSU waiting to go into it although I do have my doubts as to whether that's the problem or not because the old PSU tested fine (and was put under heavy load on a PSU test bench and monitored on all rails, not just in a PC with a multimeter). I've also ran the memory (12GB Corsair Dominator GT) through a battery of tests and it checks out fine. I'm not trying to hijack this but it's relevant to my interests because I have a machine exhibiting exactly the same fault and I, who would call myself an experienced system builder and electronic engineer, have not yet gotten to the bottom of it.
Its nice to see that someone else is having issues along with me, im almost at the stage now where im just gonna upgrade my Mobo, ram and processor out of pure frustration.
Might be time for an upgrade anyway? It seems like a good time as nothing better comes out in the Intel party 'till the end of the year.
This is the problem, a problem like this can be caused by many different parts so it is best to test them all first.
Thats the plan, Just an update, when the computer died over the weekend, it wouldnt power on until i turned the PSU off and on, but the Mobo still had its power light when she wouldnt boot. Update: Took my RAM chips out and it says my RAM is 4-4-4-12 @ 1066mhz but the BIOS has it as 3-3-3-15 @ 800mhz and i cant get it back to the 4's.. could this be one of the issues?
I'd think so, sounds like the board is at fault. When I first got mine, I did have to RMA it due to a similar problem.
It can probably do both quite happily, but if you want to be sure then use windows memory diagnostic to test your memory.
Did you try doing what I said? It fixed my laptop lastnight. Left it running, doesnt randomly shutdown anymore.
Completely stripped down the PC last night, reapplied the colling paste and tested it and 1 stick of ram. still dies. So its narrowed down to Mobo,ram,processor and PSU.... joyful
So, you didn't try a new hd with a fresh install of another os as I said? That would have been my first port of call. As I was saying. my laptop had the same issue, and it's nearly 7 years old now and runs really well. Yes, try everything, but you seem to have bypassed a solution which is has a good chance of suceeding. my symptoms wouldnt power on unless I removed cmos battery and then replaced it again. Would die at any time. All you need is a 40gb sata 9one you never used on that machine)/ ide and a copy of linux...
