Question pc won't post. passed paperclip test. replaced parts but issue persists

[ughtho]

my pc build one day wouldn't turn on at all after going into sleep mode. no fans or beeps or anything on the psu. I don't have a motherboard light to determine if power is making it to the board. i ended up doing all the tests and being led to believe the problem came from the mobo, replacing the psu didn't fix the issue. replacing the motherboard did.

the old psu passed the paperclip test. I also removed and replaced cmos. discharged power button, shorted power button, used reset switch as an alternative. everything I could find. but everything pointed to mobo as the cause. I bought a new mobo.

the new mobo and new psu worked fine for a couple of months then yesterday the computer goes into sleep and again. refuses to turn back on at all.

this psu also passes the paperclip test and the aforementioned power related checks have been ineffective

What is happening? I feel like I must be having some kind of power fluctuation that is killing my motherboards. I don't know what to do.

both times the computer died was overnight.

this happened once before with my old components, and the pc just randomly started working again after being left alone for a while.
help

 

[rgd1101]

full system spec? include make and model of the psu

 

[ughtho]

i dont know what the old psu and mobo was but it would have been ryzen 2600 compatible and 500w.

the new psu is corsair cx650m
asrock b450m hdv
ryzen 2600
pv416g300c6k 2 x 8gb
a couple of crucial 250gb ssds
a big case that is quite airy. i dont know what type it is but ive checked it thoroughly and the stand-offs are clean, ive tried shorting the power switch directly, taking the case buttons out of the equation
a gtx1070, and an rx580 as a spare/test

had it 2 years. and it has been working fine the entire time until i replaced psu and mobo after issues as described above a few months ago. was having solid performance from the pc with no cooling issues up until spontaneous death.

the room its in can get quite cold at night, and it did die both times while in sleep mode at night. none of the capacitors on either mobo were visibly bulged at all.

 

[geofelt]

The "paperclip" test is useless.
It can only confirm that a psu is dead.
It says nothing about the proper operation of the power supply.
It takes a $5000 tester to do that.

You would hope that a new power supply would run correctly out of the box.
Unfortunately, some units do fail early.

650w should be plenty.
See if you can't borrow a known good power supply of 550w or more to test with.
Your CX650 is not rated high on this psu quality list:

PSU Tier List [OLD]

Credit to @Starelementpokeand @Energycore for helping out with this long-awaited refresh of the old PSU tier list by Aniallation (who previously went several months without logging in), leaving the old list in need of help. This is simply an update of the list that was flawed in creation by the P...

linustechtips.com

If, indeed it is the psu, corsair is good about a rma.
In the mean time, buy a top quality power supply like seasonic focus with a 7/10 year warranty.

 

[ughtho]

do you work for a PSU manufacturer or something? 'Pretty limited response, geofelt. did you miss the part of my query where i said that when i replaced the first PSU, it did not resolve the issue? Replacing the motherboard resolved the issue.

Moreover, you are quoting the out of date PSU list. the up-to-date list says that my new PSU is one of the best units in tier B, mid-range systems.

exact same issue occuring in the same way with 2 different PSUs manufactured by different companies with different wattages.

<<Removed by moderator>>

 

[ex_bubblehead]

Alright, simmer down. Civility is required at all times here.

Seeing as you still haven't identified the exact (by model name/number) power supplies you are referring to geofelt's comments were spot on.

 

[ughtho]

it's a cx with a grey sticker that i bought new 3 months ago. rps0029. the up to date psu list says those are gold tier B. the old one was 500w and someone else because i bought differently. couldn't tell you who. it ended up in a landfill.

tell me, should i only buy tier A? even in a relatively budget system like above? seems pretty wasteful to follow that suggestion.

 

[rgd1101]

using a surge protector or ups?
try another outlet?
unplug the ssd to see if it boot to bios?

 

[geofelt]

PSU quality lists are opinions and are generally valid.
No doubt opinions vary.
I use that admittedly out of date list for simplicity.
It is easy to read and many are asking about cheap or old units.
If you want a solid opinion on a particular unit, find a review by a reviewer with proper experience and testing equipment.

As to your current problem, something is wrong.
If you are having power fluctuations you could buy a quality UPS.

It is unlikely that two power supplies could each have caused damage to the motherboard.
One, yes, two is unlikely.

What might have changed since all was well?
Have you moved the pc, possibly causing a short somewhere?
As a simple functional test.
Run memtest86.
It takes windows out of the picture.
It boots from a usb stick and does not use windows.
You can download the free edition here:
https://www.memtest86.com/download.htm

If you can run a full pass with NO errors, your ram should be ok.

 

[ughtho]

yes i use a surge protector. tried different outlets. different cables. with monitor connected. without. card in. card out. no boot. no lights. no fans.

i've unplugged everything unneccessary for the most rudimentary boot. took the board off the stand-offs and put it own on cardboard, swapped around the RAM,, etc.

the issue is, and i really dont reject the advice that it's the PSU. i'm sure i need to replace either the PSU or the motherboard. that's a given. one of them is definitely fried and based on last time, evidence points to one or both needing replacing again.

HOWEVER i feel its likely because this issue has persisted through 2 PSUs and 2 motherboards i think it actually stems from something else. I don't want to buy another one of either of these items without getting a sense on how i can troubleshoot any other potential causes to stop a faulty component from frying another PSU or motherboard. this is what is frustrating me because the only advice i get sends me in an expensive loop for a cheap computer.

what other factors can stop a mobo like this? can SSDs? RAM ? CPU? a HDMI cable? usb devices? cold temperature? anything? where can i start?

 

[rgd1101]

the room its in can get quite cold at night

exact how cold?

 

[ughtho]

If you can run a full pass with NO errors, your ram should be ok.

RAM was fine when I did memtestx86 after replacing the PSU and mobo 3 months ago.

 

[ughtho]

exact how cold?

Not as cold as some places no doubt.
Hard to say exactly. Never freezing. Between 1-10 Celsius at night. It's cold right now so this is why I'm wondering it might be a factor.

 

[geofelt]

it's a cx with a grey sticker that i bought new 3 months ago. rps0029. the up to date psu list says those are gold tier B. the old one was 500w and someone else because i bought differently. couldn't tell you who. it ended up in a landfill.

tell me, should i only buy tier A? even in a relatively budget system like above? seems pretty wasteful to follow that suggestion.

Most of us have budgets.
You buy what you can.
But, if there ever was a component that should be bought with quality first in mind, it is a power supply.
If you buy a sufficiently strong650/750w power supply with a 7 to 10 year warranty, it is not going to go obsolete quickly.
You will be better insulated from wall power issues and a stronger unit will allow a future graphics card upgrade.

Anything else inside is going to be old an two to 5 years.
Perhaps a good case might last you 10 years.
A good monitor can also be a long term investment.
Many times, I have bought cheap and regretted it for a long time.
More often when I bought quality, the price stung for a little while, but I never regretted it.

 

[Mr.Spock]

cold can have 2 consequences, condensation on components can cause a short - one reason they suggest new components be brought to room temperature (if shipped in a truck/etc), and 2nd it can cause contacts to expand and/or shrink (requiring you to reseat memory or GPUs.

Not as cold as some places no doubt.
Hard to say exactly. Never freezing. Between 1-10 Celsius at night. It's cold right now so this is why I'm wondering it might be a factor.

 

[ughtho]

Replacing the motherboard solved the problem, not the PSU. The PSU works fine in a different system and the same system.

If you stumble upon this page, look at what advice I received. It's an easy answer for someone to just say a power-related issue is the PSU's fault and tell you to go spend another hundred, I provided more than enough extra detail to explain why the issue is more complicated than that, and I still got told to just get a new PSU. which didn't solve my problem.

 

[Mtop]

I myself would be RMAing the old PSU or even getting another, as I dont want to be replacing the motherboard a third time.
Not to mention a new surge arrestor.