Power button on 2+ year old build suddenly does nothing

First up, here's my build.

https://pcpartpicker.com/user/nacustis/saved/rQzscf

Yesterday my PC was working fine. No complaints, no weird noises, no errors. I shut it down like normal and walked away. Today nothing happens when the power button is pressed. No clicks, no whirrs, no extra lights lighting up. The red power symbol (line and broken circle) LED on my motherboard lights up when the power supply is connected and switched to ON state. The motherboard's digital display is dark. No error code.

What I've done:

Swapped the Reset and Power button connections from the case to the motherboard to test for a button problem. Made sure all connections were firm. No change.

Checked the connections and mechanical function at the Reset and Power buttons. Nothing seems amiss.

Checked and reseated both connections from the PSU to the motherboard. No change.

Clicked the red LED power button, then held it for about 20 seconds. Tried to hit the case's power button while holding the button on the motherboard. No result.

Turned the power supply off at the external switch for a minute or more, then back on and reattempted. Reseated the power cable at the outlet and the PSU. Plugged power directly into the wall at an outlet known to work. No change from any of this.

I'm losing my mind here. Someone please help?
Reply to Nacustis
19 answers Last reply
More about power button year build suddenly
  1. Try clearing the CMOS (unplug the power cord and remove the CMOS battery for a minute). If still no go, try testing the power supply: https://www.youtube.com/watch?v=5FWXgQSokF4
    See if a fan connected to it does spin.
    Reply to alexoiu
  2. alexoiu said:
    Try clearing the CMOS (unplug the power cord and remove the CMOS battery for a minute). If still no go, try testing the power supply: https://www.youtube.com/watch?v=5FWXgQSokF4
    See if a fan connected to it does spin.


    I have attempted to test the power supply as indicated here... https://www.youtube.com/watch?v=5FWXgQSokF4

    It seems the power supply' fan does come on (albeit very gently) when powered on.

    Can you offer a guide on how to remove the CMOS battery? Is that the disc battery on my motherboard?
    Reply to Nacustis
  3. Yes. For the power supply test, it would be good to connect a fan to it and see if spinning.
    Or try borrowing a power supply and testing.
    Reply to alexoiu
  4. alexoiu said:
    Yes. For the power supply test, it would be good to connect a fan to it and see if spinning.
    Or try borrowing a power supply and testing.


    I have several fans that are part of my case, but the connectors don't look like they're compatible with the PSU's. They're small and connect directly to un-sheathed pins, and they all have different numbers of pins than any of the connectors for the PSU. At the moment they all connect to the motherboard.
    Reply to Nacustis
  5. Try borrowing a power supply, just for testing.
    Reply to alexoiu
  6. alexoiu said:
    Try borrowing a power supply, just for testing.


    I'll give it a shot. Will report back.


    Update: Hooked a different power supply to the 8- and 24-pin connectors on my mobo, no change. Removing the battery also produced no effect.
    Reply to Nacustis
  7. ok.
    Reply to alexoiu
  8. alexoiu said:
    ok.


    Update: Hooked a different power supply to the 8- and 24-pin connectors on my mobo, no change. Removing the battery also produced no effect.

    I have also attempted to manually bridge the GROUND and POWER SW pins on the motherboard that the POWER plug from the front plate was plugged into, also with no effect.
    Reply to Nacustis
  9. It might be a defective board. If under warranty, try RMA-ing it.
    Reply to alexoiu
  10. I was afraid you'd say that. I'm going to try unplugging a few other components and give it another shot. If Corsair gets back to me about the possibility of it being the PSU I'll do whatever they say, but at this point I'm not terribly hopeful.
    Reply to Nacustis
  11. Please post back the result.
    Reply to alexoiu
  12. alexoiu said:
    Please post back the result.


    I will. Hoping Corsair is helpful. Would be very nice if they can get me a new PSU, as they've manufactured both my case and the PSU. I'm thinking if they can't or won't, I'm going to buy a new PSU and try that before caving and buying a new motherboard; it's cheaper if that is what the problem is, and if it isn't I can keep it in a box until I make my next build.
    Reply to Nacustis
  13. For the power supply test it is not necessary to hook anything (the psu's own fan is a load). And if you leave an hdd/optical drive or other pheriphelas is ok (but if for whaterver reason overvolts, it may fry anything that is hooked.. so beter not stress hdd's as test device).. it is important to unhook the main 24pin ATX, the 4pin-8pin CPU, the GPU 6-8pins if present.

    Take a paperclip and short the green wire to one of the blacks
    If the psu fan's turns on, is on the right track.

    Now if you have a multimeter check all the voltages: (yellow and black 12v, red and black 5v, orange and black 3.3v)

    And to test it even furter.. try to start another system with it. There are cases when outside tests are good on voltages and start but the mb does not receive the proper pwr_ok signal and shuts off.

    Better do this first before buying a psu.
    Reply to Azzyasi
  14. Azzyasi said:
    For the power supply test it is not necessary to hook anything (the psu's own fan is a load). And if you leave an hdd/optical drive or other pheriphelas is ok (but if for whaterver reason overvolts, it may fry anything that is hooked.. so beter not stress hdd's as test device).. it is important to unhook the main 24pin ATX, the 4pin-8pin CPU, the GPU 6-8pins if present.

    Take a paperclip and short the green wire to one of the blacks
    If the psu fan's turns on, is on the right track.

    Now if you have a multimeter check all the voltages: (yellow and black 12v, red and black 5v, orange and black 3.3v)

    And to test it even furter.. try to start another system with it. There are cases when outside tests are good on voltages and start but the mb does not receive the proper pwr_ok signal and shuts off.

    Better do this first before buying a psu.


    I've tried jumping the green and black wires on the 24-pin connector while it is powered and nothing is connected to it. The PSU's fan comes on, albeit very gently and quietly. I can't say for sure that isn't its normal operating condition, though.

    I'm afraid that while I have a multimeter, I'm not skilled in its use. I'll have to look up a tutorial tonight.
    Reply to Nacustis
  15. Well it's not hard at all to measure with a multimeter.

    Be sure to select Volts DC (VDC or the simbol ---, and not the one with ~)
    If it has a large rotary dial in center, go to 20V indicated on the VDC scale (that means measure vdc up to 20v.. it is good to measure on right scale.. measuring on higher scale will give low resolution, measuring on lower scale will not measure at all)

    And hook the tester leads one to a black wire, and one on a yellow wire. (go with the tester leads inside a molex conector for 12v and 5v for easy measurement, and the 3.3 is optional but you find a few in main mb connector.
    It's only 12v, - no risk involved.

    Same for 5v - (red wire and black)
    and 3.3V (orange wire and black)

    This way you can almost rule out the psu (sure.. it does not verify the psu under load, and also does not check ripple and other things.. but if the voltages are about right

    Don't worry!.. as long as the multimeter is not set to measure amps, the multimeter acts as very tiny resistance.. so in reality it does not matter where the leads go.. you can't short anything (sure with perseverance anything is possible.. but under normal conditions.. it is VERY safe for the integrity of measured psu, and the multimeter itself, and for yourself)
    Reply to Azzyasi
  16. Azzyasi said:
    Well it's not hard at all to measure with a multimeter.

    Be sure to select Volts DC (VDC or the simbol ---, and not the one with ~)
    If it has a large rotary dial in center, go to 20V indicated on the VDC scale (that means measure vdc up to 20v.. it is good to measure on right scale.. measuring on higher scale will give low resolution, measuring on lower scale will not measure at all)

    And hook the tester leads one to a black wire, and one on a yellow wire. (go with the tester leads inside a molex conector for 12v and 5v for easy measurement, and the 3.3 is optional but you find a few in main mb connector.
    It's only 12v, - no risk involved.

    Same for 5v - (red wire and black)
    and 3.3V (orange wire and black)

    This way you can almost rule out the psu (sure.. it does not verify the psu under load, and also does not check ripple and other things.. but if the voltages are about right

    Don't worry!.. as long as the multimeter is not set to measure amps, the multimeter acts as very tiny resistance.. so in reality it does not matter where the leads go.. you can't short anything (sure with perseverance anything is possible.. but under normal conditions.. it is VERY safe for the integrity of measured psu, and the multimeter itself, and for yourself)


    Thank you. I will try this when I get home from work and will report back.
    Reply to Nacustis
  17. Azzyasi said:
    Well it's not hard at all to measure with a multimeter.

    Be sure to select Volts DC (VDC or the simbol ---, and not the one with ~)
    If it has a large rotary dial in center, go to 20V indicated on the VDC scale (that means measure vdc up to 20v.. it is good to measure on right scale.. measuring on higher scale will give low resolution, measuring on lower scale will not measure at all)

    And hook the tester leads one to a black wire, and one on a yellow wire. (go with the tester leads inside a molex conector for 12v and 5v for easy measurement, and the 3.3 is optional but you find a few in main mb connector.
    It's only 12v, - no risk involved.

    Same for 5v - (red wire and black)
    and 3.3V (orange wire and black)

    This way you can almost rule out the psu (sure.. it does not verify the psu under load, and also does not check ripple and other things.. but if the voltages are about right

    Don't worry!.. as long as the multimeter is not set to measure amps, the multimeter acts as very tiny resistance.. so in reality it does not matter where the leads go.. you can't short anything (sure with perseverance anything is possible.. but under normal conditions.. it is VERY safe for the integrity of measured psu, and the multimeter itself, and for yourself)


    Okay, so.

    I'm not skilled with the multimeter, but here's what I got. I set the multimeter to 20V as instructed and touched the ends of the probes to the inside of the sockets on the 24-pin connector.

    Yellow wire tested at about 19 volts the first time I tried it.

    Red got about .01. Then I couldn't get any.

    Orange got 0.

    Came back to yellow and it gave me less each time I took a reading. Last try it gave me .04 volts.

    Tried green for kicks and giggles. Got 4.4 consistently.

    I admit I'm not skilled with this tool. It's only my second time using it. This might be my error. But I tried multiple black wires and different ones of red, orange, and yellow. All gave consistent results.

    I just came back and tried yellow again and I'm consistently getting it flickering between 0 and 0.01.

    Am I bad at this, or is my power supply dead?
    Reply to Nacustis
  18. If you say values are consistent (the same every time).. the psu is at fault.

    As for which wires to measure.. try keeping the probes firmly in contact with the connector. It does not matter much if you touch the lead or not.. it's not harmful if you touch the 12v lead (just keep in mind measurement-wise and safety-wise.. try not to touch both contact point with bare hands at once)

    Try to pick ground from the same connector you are measuring.. so black from molex, and yellow from same molex.

    But anyway.. if you saw a consistent stable 19V reading anywhere on any connector.. well that's psu's fault and pray it did not fry anything else.
    By the looks of it, your MB might not made it as you say it's not reacting with another psu.

    This video might help you: https://www.youtube.com/watch?v=ac7YMUcMjbw

    If the psu is under warranty and it did malfunction and fried other components.. tell them that.. idk but it may cover damages resulted from their bad product. (such as motherboard).
    Reply to Azzyasi
  19. Azzyasi said:
    If you say values are consistent (the same every time).. the psu is at fault.

    As for which wires to measure.. try keeping the probes firmly in contact with the connector. It does not matter much if you touch the lead or not.. it's not harmful if you touch the 12v lead (just keep in mind measurement-wise and safety-wise.. try not to touch both contact point with bare hands at once)

    Try to pick ground from the same connector you are measuring.. so black from molex, and yellow from same molex.

    But anyway.. if you saw a consistent stable 19V reading anywhere on any connector.. well that's psu's fault and pray it did not fry anything else.
    By the looks of it, your MB might not made it as you say it's not reacting with another psu.

    This video might help you: https://www.youtube.com/watch?v=ac7YMUcMjbw

    If the psu is under warranty and it did malfunction and fried other components.. tell them that.. idk but it may cover damages resulted from their bad product. (such as motherboard).


    I may have been wrong about the 19v. I remember thinking "That's just 1 less than it should be." Might very well have been 11.

    Corsair accepted my RMA. The other PSU I used to test was lower wattage than the one that is ostensibly fried, so I'm hoping that is the reason it was not working.

    Plan now is to wait for the new PSU and pray my mobo is actually fine.

    Thanks very much for your assistance
    Reply to Nacustis
Ask a new question Answer