Desktop receiving power but not turning on

[Nelson37]

Yep, those are the correct pins.

Hard to tell from the picture but the large cap, next to the small one, just above and to the left of the VGA connector, looks a little different. Bulged or swelled at all?

Have you tried cycling the switch on the psu itself? Or connecting power with the board outside of the case?

When was the heatsink and fan removed?

Also, you did not say if the beeps you mentioned happened AFTER the siren wail. If not, this ain't lookin' good at all.

[StoneColdWhat]

Took a look there and it looked alright. I managed to get a better picture.



Cycling the switch is just flipping the thing on and off right on the PSU? If so I did that already. Also tried connecting power with board on wooden surface and that didn't do anything.

Removed the heatsink and fan after the thing stopped working. After the siren there was no reaction since. I managed to find a board locally here, I really hope it isn't my CPU.

Did Redwudz wanted me to do something about his pic or was he just pointing something out?

[Nelson37]

Cap looks OK, was just the light in the previous picture.

Yes, you have "cycled the switch" correctly. Hasn't been mentioned, but there is a "110/220" slider on the PSU which should be checked. Probably OK, but worth a look.

There are such things as a "cpu tester", you might try to find a local shop that has one. They'll charge you for the service, maybe get a discount if you agree to buy the board there,
but you already have one.

High odds it's the board and not the chip, but no certainty. Be exceptionally careful during the install process. If you have read that the CPU will get hot enough to put a blister
on your finger in less than 5 seconds, I can personally attest that this is true. Just had to see for myself.

Also, surge protector or battery back-up has not been mentioned. HIGHLY recommended, APC is the only brand I recommend, though Minuteman and Panamax used to be good, as well. Just
far more expensive. PC's are VERY SENSITIVE to both over- and under- voltage conditions. A good surge unit will likely extend the life of your PC.

[turk690]

@nelson37 This "siren" I only heard momentarily as I figured something was wrong so I turned the power off pretty much immediately. The siren and spinning of fans came on when I flipped the switch on the PSU and not from the power button or shorting the 2 pins, which did not work.
Just to confirm, you guys wanted me to short the shown in diagram E pin 6 and 8 correct? Or were you guys talking about the ones in Redwudz's pic? In any event I tried to put a jumper on pin 6 and 8 and that didn't work after removing it momentarily in getting the power on. I also tried the screw driver approach and that didn't work either. The sequence of event was: Was working > Stopped working > Fans and PSU came on after flipping the switch and siren came > Power led on mobo, cpu fan still functional but unable to turn on power by any means.

We first determine the likelyhood of a good PSU by taking it out of the case, disconnecting everything from it. Then, after laying it down so that the vents & fan are not obstructed, apply AC mains. Measure the output between any black wire (GND) and the purple wire (+5VSB); you should be able to get +5Vdc,±10%. Then take a paper clip, bend it into a "U" shape, and short out the green wire pin (PS_ON) with any black wire pin (GND), but preferably the one closest beside the green wire pin, and hold it there shorting the two pins out (push the paper clip tips into the connector to hold them in place). Observe the fan (and any lights); if the PSU is likely OK, the fan should start spinning. You may take the meter again and measure the voltages: black and red is +5Vdc, black and yellow is +12Vdc, black and orange is +3.3Vdc, black and blue is -12Vdc. Since there is no load, the voltages may be a little off over or under these rated ones; black/red may be as high as as +5.6Vdc, or black/yellow as low as +3.1Vdc. Something is bad if a voltage is way out, like just +10V or so for the +12Vdc line. High current is especially demanded of the +12Vdc line by the CPU, GPU, hard drives, etc so it may fail first.
The mainboard has a latching circuitry powered from the purple +5VSB line, that converts the momentary contact of the SPST switch in your computer's front panel when you press it into the logic high to low transition, then low state for the entire duration the computer is powered on for the green wire. The test above is a quick way of mimicking what the mainboard does and isolate some powering-on problems.
Some computers' power switches do get damaged (permanently open, permanently closed, intermittent and delicate (in this latter case a mere cross look at it is enough to close its contacts, with you gasping why the computer just shut down)).
The siren means there is something really wrong, like instant overheating of some component; wrong type, wrongly-installed, defective, or absent system RAM; or a huge fat short somewhere on the mainboard (CPU and/or supporting chips). Nimble pros at this point quickly test all of the voltages to see which one it may be, after of course in the initial standalone test above, the PSU did give the correct readings.
What I would now do is remove all the connectors and RAM, and CMOS battery (this coin battery should never measure below +3.0Vdc; even at +2.9V, I throw it out & get a fresh one). I'll tip the board face down and shake it and see what loose bits come out. I take a (grounded) air blower to the board to further ensure fluff and lint (some of which may be metallic) are blasted away. Then I spray isopropyl alcohol on the solder side, let it drip dry. After it dries, some caked-out whitish areas may form (which are dried salts and oils); I take an old toothbrush and gently scrub these out. I blower one last time.
I can see by the board pictures that you have (wisely) retained the CPU in its socket. That is a good step, because the pins on those sockets are incredibly and amazingly fragile; some over-enthusiastic people have been known to mistakenly brush a screw-driver or so against them, rendering their sophisticated and expensive x79 mainboard downright useless. Even if one pin (depending on CPU/socket there are 775, 1155, 1156, or 2011 of them) does not make reliable contact with its corresponding pad on the CPU, you get the siren.
Lastly, the grey sludge on the CPU means that is the original heatsink compound. It could have dried and makes inefficient thermal coupling between CPU and heatsink, leading to possible overheating. I'd gently clean it off the two surfaces with isopropyl alcohol and apply fresh one. Whether you will use inexpensive white silicon heatsink compound or the $10 little tubes of some gooey silvery-looking stuff that couldn't possible be more than pea-size or two in volume in them is up to you.

[StoneColdWhat]

Well I don't have volt meter so I can't test the things you wanted me to but I have pretty much narrowed it down to the mobo at this point. I took the PSU and was able to power on another computer that is working for sure and that I was able to power it on. And just in case I was not shorting the pins properly... I put the same non working mobo in the same case that has a working power switch and still wasn't able to power it on. I had a close look at the CPU and the slot and they looked fine.

I also tried to power on just the bad mobo by itself, no cpu, ram, nothing at all and it didn't beep so I think thats the end of that. I took a picture of the back of the board and some of the stains do raise the question if thats why the board isn't working.

[Nelson37]

You won't get anything at all with no CPU on the board, that's one component you must have.

Sure sounds like it's dead, let us know how the new board works out, just to verify. I'd clean the gunk off the CPU while still mounted on the old board,
hopefully you already did that.

[WinSpecToR]

Alright guys so I replaced the mobo and now I am able to boot. But my windows has some how become corrupt, I booted into Windows RE:

Chkdsk can't run because the volume is used by another process and the only way to access it would be to force a dismount which I did, but it only makes it to about 10% and says readable file segment xxxxx is not writeable, failed to transfer logged message to the event log with status 50. Then it waits for another user input.

If select no for the dismount, it says chkdsk will run the next time the computer starts but it didn't. I was initially able to get into safe mode for a brief moment none of the safemode options successfully loads anymore, last known good config didn't work and repair your computer didn't either.

I would like to avoid a reinstall if possible.

[Nelson37]

Unless the mobo is absolutely identical to the original, and sometimes even if it is, you must do a repair install.

[jagabo]

Chkdsk can't run because the volume is used by another process

Put the drive in another computer as a secondary drive (or in an external USB enclosure) and run CHKDSK from there.

[aedipuss]

hope you did a better job with the heatsink compound this time. the cpu was only about 50% covered looked like there wasn't sufficient downforce on the hsf and was probably overheating.

[WinSpecToR]

@Nelson37 The repair install must be done from within windows 7 from what I just researched correct?

@jagabo Would that do the trick? The dismount method from Win RE didn't complete it.

@aedipuss The rest of the heat sink compound is on the heatsink. In any event, I don't have extra compound hanging around right now.

[aedipuss]

heatsink compound is not re-usable. it needs to be replaced everytime the hsf is removed from the cpu.

[Nelson37]

EEK, yes, you need new heatsink compound. If you did not, POWER OFF NOW and go get some, 5-6$ and it only takes a small amount. Find a geek and borrow some
if necessary. Just enough to cover the surface.

I always do repair installs from the installation disk. The reason to do it is because something is wrong with the boot OS, so don't boot from it.
Use the CD.

[WinSpecToR]

@aedipuss Ok will go find some.

@Nelson37. I tried to do the repair install from the disc but it said I had to do it from within in windows.

[Nelson37]

Here's a link with pretty detailed instructions. http://www.sevenforums.com/tutorials/3413-repair-install.html

[WinSpecToR]

Well there was no fixing it so I decided to reinstall. But now I'm running into a problem. I tried installing several times from my original win7 dvd as well as the win7 sp1 iso from m$ when it gets step 5 of 5 of installing the OS, it restarts like it is supposed to but it says that the bootmgr image is corrupt and stops dead there. I know both images are good. I went back and checked the size of the OS and only 9gig or so on the hd was used and it is suppose to use about 20-25 so its not done installing. New mobo is P5K-SE for those interested in seeing if there is any bios settings that needs tweaking.

Also, it is an IDE drive, I am wondering if its a jumper setting but that shouldn't be the case since it was my booting drive before the mobo swap. Mind you I do have several other drives but I have temporary disconnected those for now.