PC No Longer Posts

[SD-[Inc]]

Was working perfectly fine. Wouldn't wake one morning. Powered on and off many times and no Post

It's an ASUS TUF Gaming x570-plus motherboard. The DRAM LED turns yellow and stays yellow. No other lights came on at all. I tried swapping memory into different slots and such and same thing. I have 2 sticks and tried different configurations of 1 or 2 with no luck. I doubt both sticks could have failed at same time. So probably MB died. Any other ideas before I tear the whole thing do to replace MB?

[Nunz]

Have you cleared CMOS? Any dark/burnt spots on the board, specifically near the DIMM slots?

I'm assuming no POST LED, eh?

[bill dennison]

swap the PSU and video card if you can

[SD-[Inc]]

Clearing CMOS didn't change things. Can't See any visible damage on top of board. I'll have to yank the whole thing out to look underneath. I tried pulling video card and using onboard video still nothing. Swapping PSU could be worth it. Need to see if I can find one.

[Babel-17]

IIRC I not too long ago had the same problem of the PC not posting. Yadda yadda yadda, I disconnected an external USB HD bay that had a hard drive in it, and then the PC booted fine.
Same motherboard, though the WiFi version.

P.S. A week ago I noticed horrible read speeds on an internal SSD. I could write to it full speed, but reading from it was terrible. I have a Phantek case with the SSD mounted in little doors, and I guess a cable came slightly loose. Pressed it in, problem solved.

P.P.S. I have a hub for external drives, and yadda yadda yadda, I found that the ones powered from it stopped performing poorly once I hooked them directly up to USB ports. I now only use that hub for the keyboard, mouse, and an external drive that gets its power from the wall.

[SD-[Inc]]

Another hint is that it’s not the first time the PC would not wake. But power off/on fixed it. Not this time.

[Babel-17]

Try turning off the PSU, unplugging it, and then holding the power on button of the PC. Very doubtful that would do anything at this point, but that might help empty any capacitors that are holding some juice. Do that, then try booting again. You can also try holding down the power button. I had an old Dell XPS 420 that that trick was sometimes necessary with. It was a backup PC, and I refused to let it die for a long time.

[Trunks0]

It's probably the PSU. Just be happy it didn't take the whole rig with it

[digitalwanderer]

Have you tried removing the CR2032 battery from the motherboard for about 20 minutes and putting in a fresh one? I've had all sorts of weird posting grief fixed by that on some motherboards after about 3-5 years of owning 'em.

If not that then the PSU would be my first choice also, but you'd be surprised what a fresh battery can do.

[SD-[Inc]]

Have you tried removing the CR2032 battery from the motherboard for about 20 minutes and putting in a fresh one? I've had all sorts of weird posting grief fixed by that on some motherboards after about 3-5 years of owning 'em.

If not that then the PSU would be my first choice also, but you'd be surprised what a fresh battery can do.

Thanks. That’s a cheap and easy try. Even makes sense given the symptoms. Fingers crossed.

[SD-[Inc]]

Darn, new battery made no difference. Next try will be PSU.

[digitalwanderer]

Darn, new battery made no difference. Next try will be PSU.

Sorry to hear, I recently had that one come up on my own rig about 2 weeks ago. Best of luck!

[Shadowless]

this sounds like every psu that has failed on me.

[SD-[Inc]]

It sound like the easiest way to debug this thing is to pull it out of the case and set up a test bench. What do you guys do for a power switch when running on a test bench? I have never run a MB like this before. Is there any other tips? I bought an nice anti-static pad. What else might I need?

[Nagorak]

One thing you may try is putting your memory in, making sure it's seated firmly, and then jiggle it a little bit back in forth in the slot (not with extreme force, but enough that it shakes back in forth in the slot a bit).

For some reason that has helped me get boards back to booting again when showing strange DRAM errors on boot. I guess the sticks might not have been making good contact and that technique manages to get them seated more firmly? I'm not sure. I actually just had that problem with one of my boards earlier today.

The other thing is unplug everything from the board including all SATA, GPU and other PCI-E cards, and see if that helps get it to boot. It should boot past memory error lights even with no GPU in. Some boards actually boot entirely sans-GPU although obviously you get no video signal. In any case, that can troubleshoot whether something else is causing a fault.

It sound like the easiest way to debug this thing is to pull it out of the case and set up a test bench. What do you guys do for a power switch when running on a test bench? I have never run a MB like this before. Is there any other tips? I bought an nice anti-static pad. What else might I need?

For power switch you can just use a screwdriver to jump the power switch pins on the board. It's easiest with a Phillip's head since it's not as wide and you can kind of just slide the diamond shape in between the two pins.

[SD-[Inc]]

One thing you may try is putting your memory in, making sure it's seated firmly, and then jiggle it a little bit back in forth in the slot (not with extreme force, but enough that it shakes back in forth in the slot a bit).

For some reason that has helped me get boards back to booting again when showing strange DRAM errors on boot. I guess the sticks might not have been making good contact and that technique manages to get them seated more firmly? I'm not sure. I actually just had that problem with one of my boards earlier today.

The other thing is unplug everything from the board including all SATA, GPU and other PCI-E cards, and see if that helps get it to boot. It should boot past memory error lights even with no GPU in. Some boards actually boot entirely sans-GPU although obviously you get no video signal. In any case, that can troubleshoot whether something else is causing a fault.



For power switch you can just use a screwdriver to jump the power switch pins on the board. It's easiest with a Phillip's head since it's not as wide and you can kind of just slide the diamond shape in between the two pins.

Thanks. That’s all the first stuff I tried. I did find on Amazon a cheap button setup you can just plug into motherboard headers that is perfect for this job. I am bummed my motherboard doesn’t support BIOS flashback feature so I can’t reload bios in case it is corrupt. That’s another thing I wanted to try. Now I need to start swapping PSU, mb, memory, and/or cpu until I get it working again.

[koralis]

It sound like the easiest way to debug this thing is to pull it out of the case and set up a test bench. What do you guys do for a power switch when running on a test bench? I have never run a MB like this before. Is there any other tips? I bought an nice anti-static pad. What else might I need?

Note that I take no responsibility for someone frying their PSu by doing something wrong.


To test the power supply independently from anything else, you can push a wire into the motherboard connector, shorting the green wire (power on) and an adjacent black wire. This is the PSU pin that the motherboard shorts in software via the BIOS when you hit the momentary power button on your computer. If it works, it should start supplying 5 and 12v.

If you had a fan or something wired in you could see it. If you have a voltmeter you could test the lines directly (yellow/black = 12v, red/black = 5v)

IF you short the green pin and nothing at all happens then your power supply is dead as a doornail. If it works, then that's not saying that the PSU is 100% (maybe a coil died, or it tops out at a certain amount of current) but it's an indicator that your motherboard or one of the things in it could be the problem.

A short in the system (loose screw in a place it shouldn't be?) will cause the PSU to shut off immediately.

[bill dennison]

psu tester

https://www.amazon.com/Computer-PC-T.../dp/B076CLNPPK

[SD-[Inc]]

psu tester

https://www.amazon.com/Computer-PC-T.../dp/B076CLNPPK

Awesome! Thanks Bill. I didn’t know such a thing existed. I’ll give it a try.

[Sasquach]

I'm also going to go with PSU dieded.

I experienced something similar.

It worked one day, died sometime in the middle of sleep mode and wouldnt wake. no post or nothing. MOBO lights were still on and pulsating though.

I left it alone for a day (turned off PSU), and it started right up again the next day.

Soon after same thing happened but nothing would make it wake anymore and when it finally did it didnt stay awake for very long. I put in a new PSU and havent had any issues since. Lucky for me none of the times it died was when i was using the PC, so it either died in sleep mode, or it died just idling.

I've had that PSU since building a Phenom II system way back when, so it most likely outlived its useful service life.

[SD-[Inc]]

I'm also going to go with PSU dieded.

I experienced something similar.

It worked one day, died sometime in the middle of sleep mode and wouldnt wake. no post or nothing. MOBO lights were still on and pulsating though.

I left it alone for a day (turned off PSU), and it started right up again the next day.

Soon after same thing happened but nothing would make it wake anymore and when it finally did it didnt stay awake for very long. I put in a new PSU and havent had any issues since. Lucky for me none of the times it died was when i was using the PC, so it either died in sleep mode, or it died just idling.

I've had that PSU since building a Phenom II system way back when, so it most likely outlived its useful service life.

I hope so. I don’t have a backup PSU to try. I did buy the tester Bill linked to see if that also points to the PSU. The thing is supposed to be on warranty. If the tester says it’s bad I am going to see if I can get a replacement. But the headache may not be worth the $120 a new one costs. Let’s see what happens when the tester arrives.

[SD-[Inc]]

The handy little tool Bill recommending is reporting that the PSU is bad. It’s only 2 years old should I try and get a replacement? How is EVGA support? Or, is it just worth buying another instead? Anyone had any experience with them before?

[bill dennison]

The handy little tool Bill recommending is reporting that the PSU is bad. It’s only 2 years old should I try and get a replacement? How is EVGA support? Or, is it just worth buying another instead? Anyone had any experience with them before?

both

buy a better one now and RMA the other as a spare

[SD-[Inc]]

both

buy a better one now and RMA the other as a spare

I did. New one on the way. Thanks again for the Tester tip

[SD-[Inc]]

New PSU arrived plugged it in and ….. No post….

Same symptoms exactly. Not sure next steps. Maybe I can find a refurbished version of the exact same MB for cheap and try that? I am possibly one step from ending my PC gaming career on this one. If someone wants to make an offer on the whole setup let me know (it comes with 2 perfectly good power supplies!). Otherwise, I may just trudge along with a few more part replacements…. I hate these type of problems. Whenever I get them it is always the LAST part I replace that fixes the problem.

[bill dennison]

New PSU arrived plugged it in and ….. No post….

Same symptoms exactly. Not sure next steps. Maybe I can find a refurbished version of the exact same MB for cheap and try that? I am possibly one step from ending my PC gaming career on this one. If someone wants to make an offer on the whole setup let me know (it comes with 2 perfectly good power supplies!). Otherwise, I may just trudge along with a few more part replacements…. I hate these type of problems. Whenever I get them it is always the LAST part I replace that fixes the problem.

well when it's fixed you stop

..
does the old PSU still test bad still ?

could be the MB & PSU one took out the other

think i would look for somewhere to take it to diagnose it cheap

[Nagorak]

It's tough when you don't have other parts on hand to troubleshoot with. I'd guess the problem is likely the board. It's not impossible it could be the CPU but much less likely.

Assuming the board is still under warranty, you could just order a cheap B450 to keep you going while you RMA the board. Doesn't really need to be great to last for a few weeks. I wouldn't pay for an expensive board if you can get that one replaced.

[Lazy8s]

Is the MB installed in the case? If so, try taking it out. Possibly something shorting out….

[SD-[Inc]]

It's tough when you don't have other parts on hand to troubleshoot with. I'd guess the problem is likely the board. It's not impossible it could be the CPU but much less likely.

Assuming the board is still under warranty, you could just order a cheap B450 to keep you going while you RMA the board. Doesn't really need to be great to last for a few weeks. I wouldn't pay for an expensive board if you can get that one replaced.

This one is reasonable and if works I’d probably just keep it with the RMA as backup…

https://www.asus.com/us/motherboards...ing-b550-plus/

I want to play around just a bit more with the PSU Tester to find out if it was really operator error when I got the bad test result on my old PSU or not before shelling out another $150. That’s still a bit of a mystery. The possibilities are:

1. I really didn’t have a bad PSU, just operator error
2. My second PSU is bad also
3. Board and PSU are fried

I really hope the CPU isn’t the issue. Last time I went down the rabbit hole like this it was the CPU and that was the last thing I swapped. But the symptoms were totally different back then (and nastier). That time everything posted just fine. The problem was it would always blue screen randomly during Windows installation. That was a nightmare that turned out to be a defective CPU. In that case though, the processor was brand new. Never heard of a CPU wearing out after 2 years of usage. But, I have heard of capacitors failing over time on MB. So that still feels like the likely candidate.

[koralis]

This one is reasonable and if works I’d probably just keep it with the RMA as backup…

https://www.asus.com/us/motherboards...ing-b550-plus/

I want to play around just a bit more with the PSU Tester to find out if it was really operator error when I got the bad test result on my old PSU or not before shelling out another $150. That’s still a bit of a mystery. The possibilities are:

1. I really didn’t have a bad PSU, just operator error
2. My second PSU is bad also
3. Board and PSU are fried

Just to be complete:
4. The PSU tester is bad.

But, I have heard of capacitors failing over time on MB. So that still feels like the likely candidate.

You should be testing the power supply without anything being plugged into it. The tester only is testing the main MB plug also. Depending on your supply there may be multiple coils that supply diffferent things... for example, the aux power plugs for the MB might be hosed even if the main ones are ok.

A voltmeter could tell you. What exactly did the PSU tester say was wrong?

[SD-[Inc]]

Just to be complete:
4. The PSU tester is bad.




You should be testing the power supply without anything being plugged into it. The tester only is testing the main MB plug also. Depending on your supply there may be multiple coils that supply diffferent things... for example, the aux power plugs for the MB might be hosed even if the main ones are ok.

A voltmeter could tell you. What exactly did the PSU tester say was wrong?

Right, it could be a bad tester. I've read more and according to the tester if a value blinks or the tester beeps the PSU is bad. The PG value was blinking and the tester was beeping. All voltages showed good. PG is Power Good Delay. I had no idea what that meant until I looked it up.

Power Good Delay is the amount of time it takes a power supply to start up completely and begin delivering the proper voltages to the connected devices. According to the Desktop Platform Form Factors Power Supply Guide [PDF], Power Good Delay (called PWR_OK delay in that document) should be 100–500 ms.

The value that was blinking was 130MS. This sounds good according to the definition above? I am going to try it on a known working PSU and see what I get. I could be chasing red herrings here


PS. The tester tests the main ATX power and the CPU power too. That's all I am using right now. I unplugged everything else.

[SD-[Inc]]

Well glad I kept digging. Pulled out my brand new PSU and put it on the tester. (Something I should have done before I put it in!). The tester reported the same bad values for PG. But the new PSU reported a blinking 0MS. That definitely is outside recommended specs. My son was nice enough to allow me to disassemble his working PC and his PSU tested a solid 300MS PG with no blinking at all. So the only real evidence I have now is the new PSU is also bad. I’ll try returning and trying another.

Sounds like PS Tester is good, new PSU is also bad. Definitely not buying a lottery ticket soon…

[koralis]

Right, it could be a bad tester. I've read more and according to the tester if a value blinks or the tester beeps the PSU is bad. The PG value was blinking and the tester was beeping. All voltages showed good. PG is Power Good Delay. I had no idea what that meant until I looked it up.


The value that was blinking was 130MS. This sounds good according to the definition above? I am going to try it on a known working PSU and see what I get. I could be chasing red herrings here

Definitely within the range of normal based on what you wrote before. It could be that the tester expects a range of 200-400 ms or something and flags other values. That doesn't mean it's bad though.

The tester reported the same bad values for PG. But the new PSU reported a blinking 0MS.

Instantly good power doesn't seem like it should be a problem to me. Sounds ideal, in fact. I'm voting on lousy parameters in the tester.


Ah... here it is.

https://en.wikipedia.org/wiki/Power_good_signal

The ATX specification requires that the power-good signal ("PWR_OK") go high no sooner than 100 ms after the power rails have stabilized, and remain high for 16 ms after loss of AC power, and fall (to less than 0.4 V) at least 1 ms before the power rails fall out of specification (to 95% of their nominal value).

So this isn't really about how fast the power stabilizes at the correct level. It's basically saying that the supply is "cheating" and just setting the PWR_OK signal high immediately, not after verifying that the voltage has stabilized + 100ms as the spec says.

This isn't to say that the supply isn't functional, it means that they're not adhering to the startuip specs and the tester is telling you that. Really, it's PROBABLY not a problem, but at 0ms it means that your motherboard could be undervolted initially untl the supply gets up to speed, or bounce with a startup voltage spike that gets used. Over time that could kill your electronics hence the point of the PWR_OK signal latch in the first place. What brand is this? I want to make sure I avoid it the next time I replace a supply.


The 130ms seems perfectly fine though. Again, seems like they're requiring a time that is longer than the ATX spec asks for.

Cheaper and/or lower quality power supplies do not follow the ATX specification of a separate monitoring circuit; they instead wire the power good output to one of the 5 V lines. This means the processor will never reset given bad power unless the 5 V line drops low enough to turn off the trigger, which could be too low for proper operation.

Maybe they had some experience with supplies doing the same cheat, just implementing a fixed delay to fool the testers so they pushed out the values? (though if the supply is stable by the 130 ms, it doesn't really matter WHY they set the line high)


SO given that the numbers of your original supply seem to be in spec and the tester is being bitchy, it's looking more like a motherboard problem. have to tried powering up the motherboard outside of the case? Do that with a minimal configuration. Watch fans, etc.

Uhmm... just occurred to me, do you have a water-cooled system per chance? If the pump is stuck it could draw a lot of power and cause the power-supply to shut itself down similar to a short.

[SD-[Inc]]

Definitely within the range of normal based on what you wrote before. It could be that the tester expects a range of 200-400 ms or something and flags other values. That doesn't mean it's bad though.




Instantly good power doesn't seem like it should be a problem to me. Sounds ideal, in fact. I'm voting on lousy parameters in the tester.


Ah... here it is.

https://en.wikipedia.org/wiki/Power_good_signal



So this isn't really about how fast the power stabilizes at the correct level. It's basically saying that the supply is "cheating" and just setting the PWR_OK signal high immediately, not after verifying that the voltage has stabilized + 100ms as the spec says.

This isn't to say that the supply isn't functional, it means that they're not adhering to the startuip specs and the tester is telling you that. Really, it's PROBABLY not a problem, but at 0ms it means that your motherboard could be undervolted initially untl the supply gets up to speed, or bounce with a startup voltage spike that gets used. Over time that could kill your electronics hence the point of the PWR_OK signal latch in the first place. What brand is this? I want to make sure I avoid it the next time I replace a supply.


The 130ms seems perfectly fine though. Again, seems like they're requiring a time that is longer than the ATX spec asks for.




Maybe they had some experience with supplies doing the same cheat, just implementing a fixed delay to fool the testers so they pushed out the values? (though if the supply is stable by the 130 ms, it doesn't really matter WHY they set the line high)


SO given that the numbers of your original supply seem to be in spec and the tester is being bitchy, it's looking more like a motherboard problem. have to tried powering up the motherboard outside of the case? Do that with a minimal configuration. Watch fans, etc.

Uhmm... just occurred to me, do you have a water-cooled system per chance? If the pump is stuck it could draw a lot of power and cause the power-supply to shut itself down similar to a short.

No. I question the PG setting too. Although, I did see something about too fast can cause MB to crash loop. I returned the PSU for another and if same results, back to trying a new MB. Although, given the only known working power supply tested perfectly fine is the only evidence I have so far. But, it was an older PSU so maybe the newer ones cheat? I dunno. This is just going to be a a long trial and error process…

[chigger_ns]

I have saw in a few pc repair videos that the 3000 series cpu tend to fail more than others. I would buy a cheep cpu off craigs list and try it. or get a shop to test the system. Depends which way you want to go with it.

[Nagorak]

I have saw in a few pc repair videos that the 3000 series cpu tend to fail more than others. I would buy a cheep cpu off craigs list and try it. or get a shop to test the system. Depends which way you want to go with it.

I didn't really have trouble with 3000 series myself, but Ryzen 1000 did seem dodgy. One R9 1700 I bought at release just stopped posting after a week. Then the second one developed instability after a couple years, when it was placed into a new system (not impossible that there could have been something wrong with the motherboard that damaged the chip). Then my 5900X had uncorrectable WHEAs that was only fixed with a replacement.

The only relevance to the topic at hand is to second that AMD CPUs (or maybe newer CPUs in general?) seem a little bit more suspect in terms of reliability. So, I wouldn't necessarily assume the old axiom that "CPUs never die."

[Sasquach]

Well glad I kept digging. Pulled out my brand new PSU and put it on the tester. (Something I should have done before I put it in!). The tester reported the same bad values for PG. But the new PSU reported a blinking 0MS. That definitely is outside recommended specs. My son was nice enough to allow me to disassemble his working PC and his PSU tested a solid 300MS PG with no blinking at all. So the only real evidence I have now is the new PSU is also bad. I’ll try returning and trying another.

Sounds like PS Tester is good, new PSU is also bad. Definitely not buying a lottery ticket soon…

well thats just bad luck.

Would he be nice enough to go a bit further and let you install his PSU into your dead PC and see if it wakes up? obviously you wouldnt use it, but if you can turn it on (if it turns on) and let it load and then turn it off then you can confirm it was a PSU issue.

[SD-[Inc]]

well thats just bad luck.

Would he be nice enough to go a bit further and let you install his PSU into your dead PC and see if it wakes up? obviously you wouldnt use it, but if you can turn it on (if it turns on) and let it load and then turn it off then you can confirm it was a PSU issue.

I thought about that. Buts it’s a prebuilt PC and the PSU doesn’t even have the connection I need. It’s a little too old.

[Hardwood]

I have the same motherboard and very curious if you've solved this SD.

[SD-[Inc]]

I have the same motherboard and very curious if you've solved this SD.

Not yet. I returned the PSU for a replacement and UPS lost it in shipment. So now I need to go through a claims process with them. It might just be a total loss. The whole mess has demotivated me. I have to wait for another week before the package is declared lost. Just sitting on things for now.