Computer Boot Loop, only a week old

[Roundbaby]

I built this PC a week ago, and I woke up a few mornings ago to a PC that would cold restart every few minutes after I booted into Windows. I tried looking into this issue in the BIOS, reverting settings to factory presets to no avail. After long, the cold restarting became more frequent, barely getting past the ASUS POST splash screen before restarting again. I can still access BIOS fine, but all other parts of the computer are completely unreachable now.

I originally believed it to be a PSU issue, maybe the voltage regulator was faulty and causing hard reboots prematurely, but upon replacing the PSU there is still no end in sight.

I saw another post on the forums about clearing CMOS, and I successfully reverted the BIOS to factory settings but the boot loop continues.

Could it be a shorting issue? Maybe bad mobo? Please help!

Specs:
Motherboard: ASUS Prime X570-P Ryzen 3
RAM: 32gb 3200MHz GSkill (4x8gb)
CPU: Ryzen 9 3900X w/ Noctua NH-D15 cooler
GPU: RTX 3080 FE
Boot drive: Corsiar M600 1TB NVME PCIe Gen 4

 

[XAIXER]

What's your PSU wattage?

[Roundbaby]

PSU: EVGA G5 1000W 80+ Gold

The one I had originally was 850W, and I also thought it could be a power target issue so I upped it to 1000W, no avail.

[truescot]

try different memory stick combinations of 2, also check thermal paste of cpu

[Helpful Tech Witch]

Just now, truescot said:

try different memory stick combinations of 2, also check thermal paste of cpu

thermalpaste wont boot loop, unless he's The Verge.

[truescot]

Just now, HelpfulTechWizard said:

thermalpaste wont boot loop, unless he's The Verge.

no, but if there isnt enough pressure between the cooler and the cpu it could and refreshing the paste can make you revisit that, i speak from experience on that stupidity  

[Helpful Tech Witch]

1 minute ago, truescot said:

no, but if there isnt enough pressure between the cooler and the cpu it could and refreshing the paste can make you revisit that, i speak from experience on that stupidity  

? If theirs not enough pressure it would just over heat, he should get to windows tho still.

[truescot]

15 minutes ago, HelpfulTechWizard said:

? If theirs not enough pressure it would just over heat, he should get to windows tho still.

maybe, then again maybe not,  I had this exact issue when i had a corsair h80i water cooler strapped on with the wrong mount (i jerry rigged it until i could get the right thing), for quite a long time it worked fine with marginally high temperatures, then one day it just spazzed out and got stuck in a boot loop, i was convinced the pump had died on the cooler or something took it off, put it back on a few times then realised it wasnt sitting very hard against the CPU, got a hold of a different mount from another pc and bobs your uncle it started working perfectly again. *actually thinking about it, there is a video on LTT with Linus trying to hold down a cpu cooler because he didnt have the right brackets for a xeon i think and he had the same problem, would either boot loop or not even post

Edited October 20, 2020 by truescot

[Roundbaby]

Ok, the problem has progressed just now. Now, I cannot even get to the bios splash, or at least there is no signal reaching the monitor. Also, the lights on the GPU are no longer turning on either, which was normally immediate after hitting the power button. I was really hoping I wouldn't have to check the CPU because the NH-D15 is a beast of a cooler and is impossible to reach around to plug the fan cables in, but it appears I have no other option.

It should be noted that I have the PC on it's side now, just so I can more easily pick and prod at things without having to move it to the kitchen table every time I wanted to re-arrange something.

P.S., I tried removing some RAM to see if it would get any better (A2 & B2 slots), but still no good.

I'll update on the health of the CPU seating

[truescot]

3 minutes ago, Roundbaby said:

Ok, the problem has progressed just now. Now, I cannot even get to the bios splash, or at least there is no signal reaching the monitor. Also, the lights on the GPU are no longer turning on either, which was normally immediate after hitting the power button. I was really hoping I wouldn't have to check the CPU because the NH-D15 is a beast of a cooler and is impossible to reach around to plug the fan cables in, but it appears I have no other option.

It should be noted that I have the PC on it's side now, just so I can more easily pick and prod at things without having to move it to the kitchen table every time I wanted to re-arrange something.

P.S., I tried removing some RAM to see if it would get any better (A2 & B2 slots), but still no good.

I'll update on the health of the CPU seating

instead of removing it etc maybe just try to see if it has any real free play on it and tighten the cooler down a bit more, helpfultechwizard is right that it shouldnt be an issue with thermal paste, also what thermal paste did you use? i have heard of people using liquid metal like thermal grizzly Conductonaut (i use it all the time) and messing it up witha bit getting on the mboard and causing a short after movement

[Roundbaby]

I used a really fancy thermal paste my dad bought me a few years ago that's IC Diamond 7 carat compound. I was worried about tightening it down too hard, but there appears to be a little play. I am willing to tighten it down more, but how much would you say is tight enough? I don't know what the likelihood is of being strong enough to tighten to the point of cracking the IHS, but I'm very nervous when it comes to these kinds fo things 

[Roundbaby]

I've tightened it down considerably more, and there is still some play but that might just be from the flex of the heat pipes. I originally counted CPU mounting issues out because it operated fine for 7 days with absolutely no crashing, and for it to suddenly have very immediate issues seemed impossible to be a mounting problem me... regardless, hold while I spend copious amounts of time trying to plug in the cpu fans again

[truescot]

6 minutes ago, Roundbaby said:

I've tightened it down considerably more, and there is still some play but that might just be from the flex of the heat pipes. I originally counted CPU mounting issues out because it operated fine for 7 days with absolutely no crashing, and for it to suddenly have very immediate issues seemed impossible to be a mounting problem me... regardless, hold while I spend copious amounts of time trying to plug in the cpu fans again

well, i have never managed to brake a single cpu yet by overtightening and i litterally must have fitted in the high hundreds and i am a firm believer in "fuck the torque wrench manhandle that bitch", i have one time a long long time ago (talking pentium 1 days) seen someone push down on the middle of the chip with his thumb when fitting and crack the chip right down the middle, but with a cooler that wont happen, too much surface area and pressing it down into the pin holes not into the space in the middle.

[Roundbaby]

CPU heatsink is back together, still no POST or LEDs on the gpu. Perhaps I should take a look at the pins on the CPU and reseat it?

I would imagine tightening down the CPU heatsink mount would help check the seating of the CPU, but I guess not...?

[truescot]

 

7 minutes ago, Roundbaby said:

CPU heatsink is back together, still no POST or LEDs on the gpu. Perhaps I should take a look at the pins on the CPU and reseat it?

I would imagine tightening down the CPU heatsink mount would help check the seating of the CPU, but I guess not...?

I would doubt it to be honest, its unlikely it wasnt seater properly and still worked, especially with the amd which has pins instead of the pads on an intel, just one other thing i can think of off the top of my head, you mentioned that you replaced the psu, did you definately remember to put the cpu power cables back in the motherboard, that would explain it not posting but fans etc coming on, also try giving the 24 pin motherboard connector a bit of a push

[Roundbaby]

CPU power is definitely plugged in, given it's a 3900X I have all 3 4-pin connectors hooked up on both ends. Still no post.

[Roundbaby]

I think I'm going to go into seating-land, will give updates

[Roundbaby]

Here's a picture of the IHS -- The thermal paste looks exceedingly brown compared do other thermal paste, is it a feature of IC Diamond? 

P.S. the seating is fine, and the pins are pristine 

[Roundbaby]

Do GPU LEDs come on if no display is connected? I would imagine so... that's the only reason I could imagine the GPU is no longer showing lights.

[Roundbaby]

I'm back to BIOS splash! turns out it must have been an issue with the ram, causing the GPU to not be recognized by the mobo. With one stick in, we're back to the computer cold restarting after bios. I read something about CSM playing a role, could enabling it better my chances?

[Roundbaby]

Here's a breakthrough-- with no display connected, the crash frequency is greatly reduced (I can tell by seeing if the RTX logo on the GPU turns off), seemingly a crash every few minutes as opposed to every few seconds. Could it be a GPU problem? Maybe peripherals?

[Roundbaby]

I am beginning to feel this issue is software related, since there are two main areas where the reset occurs: 
1) give or take 3 seconds after BIOS splash
2) after "Preparing Startup Repair" when the screen says "diagnosing this PC"
I can sit in the BIOS as long as I want and even access my windows disk image without error. However, after clicking "install now" when installing windows in the install image, a few moments before the product key screen would appear, the restart occurs

Sometimes I get lucky and can get past "diagnosing this pc" and reach the advanced start options, and this time I decided to "Reset this PC" and completely wipe it. Maybe now I can get to windows.