Question PC only switches on by flipping PSU switch and/or power strip switch. Can anybody help me out?

[THJohn]

Hello all!

I'm having a problem with my pc and i'm hoping that someone out there could help me solve the issue or point me in the right direction. The problem itself is simple: my PC won't power on normally by using the power button.*
It just started to happen that i went to my PC, pressed the power button and nothing happened. Eventually i got the PC running again by flipping the PSU switch off and on, and it just went on - WITHOUT pressing the power button on the case!
Like how does that even work?

So past few weeks i (may) need to:

Flip the switch on the PSU, Flip the switch on the power strip, unplug the power cable from the PC and re plug it for my PC to 'jump on'. (Still without pressing the power button on the case, and in no particular order)
Now it gets even more strange- have a read at this:

Some days only flipping the PSU switch will be enough to make the PC go on.
Some days i need to do a combination of all three actions to make the PC go on. (no particular order)
Some days it may take more than 15 minutes of trying the above to eventually get the PC on.
Very rarely the PC WILL power up by using the case button.*
Some days the PC WON'T power up using the screwdriver method.

It really has me confused. Once the PC is on, it's on as if nothing is wrong- it does not shutdown or loses power.

These are the system specs:

CPU - i5 4590
RAM - Hyper X FURY 2x 4GB
MOBO - Gigabyte GA-Z97P-D3
Video Card - MSI R9 390 8GB
Storage - Samsung Evo SSD 250 GB and a 1TB WD normal harddisk
PSU - Seasonic G-Series 750W (80 Plus Gold) semi-modulair
OS - Windows 10 64bit

All parts are +/- 6 years old and by the looks of it are in good condition.

Here are some things i tried but to no avail to remedy the issue:

Used another power strip, Used another power cable, Used another outlet, Switched off Windows FastBoot, Holding the power button for more than a minute. CMOS battery seems fine. I also reseated al the cabling inside of my case. Screwdriver trick sometimes work, sometimes does not.

So yeah how do i narrow down if it's like the PSU or the motherboard? I keep thinking something is shorting somewhere which is giving my PSU a 'not-okay' signal to power on? In the next few days i will disassemble my PC and give it a good clean.

If you need extra info hit me up!

Thanks,

John

 

[Ralston18]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition.

One immediate risk is loss of data. Power related problems, especially those where power is suddenly lost, can cause data corruption.

Windows needs to go through the full and proper power down process to prepare for the next boot. Continuing power issues just create more problems. Snowball effect.

At the first opportunity when the PC is on and working as if "nothing is wrong" be sure to back up all important data. At least 2 x and to a target source not within your computer.

Then look in Reliability History for error codes, warnings, and informational events that correlate with the power problems.

Numerous and varying errors are a sign of a faltering PSU. There can be and could be other issues. Faulty case power switch for example.

Immediate priority is, again, to ensure that all important data is backed up, proven recoverable and readable.

 

[THJohn]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition.

One immediate risk is loss of data. Power related problems, especially those where power is suddenly lost, can cause data corruption.

Windows needs to go through the full and proper power down process to prepare for the next boot. Continuing power issues just create more problems. Snowball effect.

At the first opportunity when the PC is on and working as if "nothing is wrong" be sure to back up all important data. At least 2 x and to a target source not within your computer.

Then look in Reliability History for error codes, warnings, and informational events that correlate with the power problems.

Numerous and varying errors are a sign of a faltering PSU. There can be and could be other issues. Faulty case power switch for example.

Immediate priority is, again, to ensure that all important data is backed up, proven recoverable and readable.

Thanks for your quick reply!

I updated my original post with my system specs i hope this is sufficient. I don't keep any important stuff on my PC so a backup is not needed but thanks for the heads up anyway!
I checked Reliability History but i can't find any type of event relating to power issues.

 

[specterk]

There's a lot of possible issues to cover here, and you've unfortunately introduced a lot of variables with your troubleshooting process. Based solely on having troubleshooted this issue a few hundred times, my gut says suspect a loose case switch, loose front panel header, or loose mobo power cable.

Other important info for outside troubleshooters to have would be: do your PSU and/or CPU fans spin up at all when your system fails to start?

Before getting really deep into more troubleshooting, though, I'd try borrowing a different case and hooking up just its power switch to your motherboard (with everything else hooked up exactly like it is now), and see if that solves the problem. Also, you should definitely stop messing with the PSU power cable, power strip, and PSU rocker switch; leave those hooked up and turned on like they normally are, and only test your motherboard pins OR your case switch when attempting to start the system. Test one system, rule it out or swap it out, move on to the next; don't test multiple systems at once.

Assuming that by the "screwdriver trick" you are referring to shorting the front-panel-header's power switch pins on your motherboard to start the system, my next immediate concern is that you damaged the motherboard the first time you did that (either by accidentally shorting something else, or by shorting those two pins for more than an instant and crippling that circuit) - but if that isn't the case, and if you're doing it safely, a correctly-performed short-start like that will only fail if the motherboard experiences a fault or doesn't get power.

So, IF you're absolutely, positively, 110% certain you're doing the short-start correctly every single time, AND IF it sometimes works (meaning your system boots and runs completely normally) but sometimes doesn't, then that effectively rules out your case power switch/cable. (If your case were to blame, the motherboard would start up correctly every time when bypassing it). If your case isn't the problem, and if you are in fact doing the short-start correctly every time, your apples-to-apples power-on success rate should also be the same as it was when starting via your case power switch (in which case, just go back to using your case, as it's much safer). However, if there's been a marked difference in power-on success between the switch/short methods, I'd stop here and test with another case. More than one intermittent fault occurring simultaneously is extremely unlikely, unless your method is one of them. =P

If testing with another case isn't an option, try using a jumper cap instead of a screwdriver to short-start it; it'll make sure that you achieve reliable contact between pins, and reduce the risk of shorting out anything else nearby. Just be sure that you remove it QUICKLY. Shorting those pins for more than a second or so has the potential to do some real damage. If you also have a dongle/pin-extender attached to the front panel header, try removing it to access the pins directly for all tests.

Beyond that, I'd suspect a loose motherboard power or cpu power cable/socket, or even the front panel header itself. Like USB headers, it's not uncommon for them to wear out over time - particularly the motherboard power socket, which sees a lot of abuse. You can also check the soldering points on the underside of the motherboard directly underneath the 3 aforementioned areas, and if they're good, give the baseplates a gentle wriggle to make sure they feel secure and don't move or flex at all. Inspect the contacts inside your case power switch cable's cap, as well as the contacts and pins inside the CPU and motherboard power sockets and power cable caps; make sure they aren't severely misaligned or damaged, and that the top edges of each contact/pin are level with the rest. Sometimes one will back out of its intended position and intermittently fail to make a solid connection.

It is theoretically possible that your power supply could also be at fault, but intermittent power supply issues are extremely difficult to diagnose without a multimeter (and there are plenty of possible internal mechanical issues that can't be identified even with one) - but from what you've said so far, a switch, cable/socket/contact or motherboard issue is more likely. If you have a meter and are comfortable going down that road, it's still worth doing so to rule out an obvious failure. If you get a good meter read on an instance when your system doesn't start, then you know your PSU is either fine or failing in a way that you can't measure, in which case the only sure way to rule it out is to test with a known-working one. If you don't have a meter, try a power-torture test when you manage to get your system running. A PSU that's bad or close to dying is extremely unlikely to run normally and should fail very quickly - just be ready to pull the plug from the wall in case it goes bad, monitor your real-time voltages closely throughout, and don't attempt with any XFX hardware anywhere in your system (seriously - not a joke).

With all that said, to be honest, there are unfortunately just WAY too many variables at play here to jump to any obvious conclusion. Failure to start can be CPU-related (bent pin, damaged CPU socket, loose/faulty CPU power cable or power socket, bad CPU fan, loose/faulty CPU fan cable or header, bad heatsink mount or TIM triggering a thermal fault, or just a bad CPU), RAM-related (less common these days, but faulty modules especially in slot 0 or 1 can still sometimes cause POST failures), or any of the other issues mentioned above.

The most likely scenario is still that your case power switch/cable/circuit are at fault, and you just aren't correctly shorting the power switch pins by hand when testing.

 

[Karadjgne]

Unplug everything from the back of the pc that's not absolutely necessary. Pretty much that means everything except the monitor and network cable.

Then try powering up as 'normal' usually is.
If that works, then start plugging things back in one at a time, until it doesn't work. That last thing is usually the culprit.

Unless you turn off the rocker, or pull the plug, psus and motherboards are never actually 'off' they are in a Standby state, and pulling power, especially on the 5v rail to the various USB devices governed wake timers.

Inside the psu there's a power ok circuit that once you push the on button, does the various protection checks for shorts etc, then starts the power cycle booting up. If the 5v rail pulls enough juice to hang that up, or there's any other issue in that circuitry, you get nothing. Power not ok. That's when you'll need to rocker off the psu, to allow that circuit to reset. From the sounds of it, the power ok is failing somewhere, like a relay hanging up, and not even allowing the reset at times. Pushing the power button, shorting the pins, none of that helps.

That G is as old as your pc, they discontinued them Years ago (2017?) and had a 5 year warranty. So you might still be covered, maybe.

The only way to determine without doubt if it actually is the psu, is by temporary replacement, and that includes All cables attached to it.

 

[THJohn]

Thanks for you extensive reply's Karadjgne and Specterk! I'll go ahead and perform the suggested tips to try and troubleshoot! I will post again if if made any progress!

 

[THJohn]

To everyone who replied to my original post; thanks for your advice and taking the time to help me out. Bless. After trying many things suggested by you guys and since i didn't had any spare parts i decided to let a computer store have a look at it.

In this case it was a faulty PSU!

 

[laughingstock]

To everyone who replied to my original post; thanks for your advice and taking the time to help me out. Bless. After trying many things suggested by you guys and since i didn't had any spare parts i decided to let a computer store have a look at it.

In this case it was a faulty PSU!

I actually have the same problem. I have been turning on my PC by switching the PSU switch on and the case power button ON at the same damn time Often times my PC would turn on between 11 PM until 1 AM. I will be buying a new PSU for this as we are really having the same problem.