PC stuck in reboot loop after cleaning

[wheresmycar]

I take it you've tried popping out the CMOS battery/run everything on default at the BIOS level?

Remove any third party device during testing... just keep the basics hooked up (KB, mouse and display)

For something like this, i'd disassemble the entire build again! You mentioned you used cloth/tissues... check each and every connector (mobo incl CPU socket/RAM slots, PSU, cables, etc etc) for any debris or other cleaning product remnants. Inspect the board carefully with good lighting for broken-off/loose components, swollen capacitors, burnt areas, etc. If everything looks in order, assemble the system on a non-conductive surface (outside of the case) and see if you have any luck.

If the problem persists, its either the board or memory sticks which have gone kaput (most likely the board). A 5 min reboot setback in my experience is usually the mobo or some sensors on the mobo playing up.

 

[depakjan]

Yes I took out the CMOS battery for 30 mins and yea I am currently in barebones setup. I checked the MB for any swollen capacitors I cant see any but then I am not that good in the this area. Below is how it is

ignore the missing screws and the CPU fan hanging around, I was thinking maybe I tightened the cooler screws too much, so I was testing it this way as well.

 

[Bobaganoosh]

No offense, but I can see why you were trying to clean it. I'm starting to wonder if you have a bit of a piece of something mildly conductive stuck in the dust that's intermittently shorting a component, causing your PSU to reset from triggering a protection circuit. Just curious, do you hear a "click" when it restarts? You might want to get a can of duster and try to get as much of the dust off your components as possible (don't do this with the system running). You can also use something like a soft bristled toothbrush and isopropyl alcohol and very carefully wipe away grime. The air is useful for blowing out PCIe slots, memory slots, etc. that you can't get to otherwise.

Other theories at this point would be that you have some other damage in your motherboard or CPU causing the instability or that it's heat-related as things can flex and move as they warm up and cool down. Plugging things in and unplugging can flex the board and if there's a crack or something, it'll only short when it heats up or flexes. You can also check for loose cable contacts. If you wiggle the cables a bit while it's running, does it reset? (don't yank on anything, just gentle wiggles).

 

[maxfly]

I would call it a day and rma that mb at this point. You're down to that or a bad CPU. A CPU dying is far more unlikely than a bad mb.

Personally, if I swap PSUs, run memtest86+ and test the CPU out with no change. I've done my due diligence. My time is worth far more than any component in a rig to spend all the time you have in your troubleshooting quest. You've got a malfunctioning component and your most likely is the MB.

 

[depakjan]

Yes, I have come to the same conclusion, thanks all. I will post here once the MB comes back from RMA.

 

[Mussels]

Sorry, here is the exact RAM, XPG DDR4 D60G 3200 MHz CL16

So its getting more crazy , the new PSU Deepcool PM850D came in today, so plugged in without screwing in the PSU, initially it worked good, I watched 2 videos , browsed some time then I started a game, once the game loaded it rebooted , god damn it!! , once the PC logged back in, I will simply checking the temps, they were all in 40c to 45c ranges, now I ran Cinebench R23 , I could run 3 to 4 multicore benches, temps reached only 68 to 70C , then to my surprise it again rebooted after maybe 30 mins, highest it has been on since this issue started, after this the reboots became frequent, now its back to rebooting every 5 mins
what craziness, so I hope its not the PSU!! , I cant understand why it could work well for sometime with the new PSU.

Got no clue now I am not gonna buy anything new, I will send both the CPU and Mobo for RMA, wait for months or whatever it takes for those to be back.

EDIT: One strange thing I am noticing is with my Old PSU, reboots are very frequent almost immediately after I enter my PIN it reboots but with the New PSU I can sometimes stay logged on for even 1 hour and the restart are not frequent though , I was even able to run one unigine benchmark as well , does this explain anything ?? Maybe my new PSU is also faulty ? or we can safely assume PSU is not at fault here

SPECTRIX D60G DDR4 RGB Memory Module (xpg.com)
3200 C16 at 1.35v, 3600 C17 at 1.4v

It doesnt say if that's two XMP profiles, or two seperate kits - or how many ranks they have. Not your fault, just poor information on their end. Frustrating, and now i understand why you didnt share more information - they don't give you any!

This lets them part swap out under the same model name, which is frustrating for situations like this. I guess CPU-Z is going to be our saviour there.

Sorry, here is the exact RAM, XPG DDR4 D60G 3200 MHz CL16

So its getting more crazy , the new PSU Deepcool PM850D came in today, so plugged in without screwing in the PSU, initially it worked good, I watched 2 videos , browsed some time then I started a game, once the game loaded it rebooted , god damn it!! , once the PC logged back in, I will simply checking the temps, they were all in 40c to 45c ranges, now I ran Cinebench R23 , I could run 3 to 4 multicore benches, temps reached only 68 to 70C , then to my surprise it again rebooted after maybe 30 mins, highest it has been on since this issue started, after this the reboots became frequent, now its back to rebooting every 5 mins
what craziness, so I hope its not the PSU!! , I cant understand why it could work well for sometime with the new PSU.

Got no clue now I am not gonna buy anything new, I will send both the CPU and Mobo for RMA, wait for months or whatever it takes for those to be back.

EDIT: One strange thing I am noticing is with my Old PSU, reboots are very frequent almost immediately after I enter my PIN it reboots but with the New PSU I can sometimes stay logged on for even 1 hour and the restart are not frequent though , I was even able to run one unigine benchmark as well , does this explain anything ?? Maybe my new PSU is also faulty ? or we can safely assume PSU is not at fault here

It's *a* voltage drooping. The better PSU sags less, so the relevant hardware after it has less trouble. This is why some people have more problems than others - a better PSU can make weak motherboards perform better.

It's a weakish VRM setup on that board, a 4+2 setup (4 CPU, two SoC) - not bottom of the barrel but on the weak side.
Video here has the CPU side of the VRMs tested and quite happy about them for stock speeds and slight overclocks, but they don't go into RAM support in this video.

For once we have a fantastic chart showing how it's set up internally

The high side FET's:
MOSFET — Power, Single, N-Channel, SO-8 FL 30 V, 46 A (onsemi.com)

11.2A at 80C, and you have 8 of them for roughly 90 amps when they're quite toasty - yours shouldnt be that hot (check in HWinfo with cinebench running?) but that's still enough to get most of the lineup running at stock speeds, even when they're 'overheating'
Good for a budget board, overall

Definitely set the ram to 2933 and SoC to 1.15v and see if it helps. My 3700x is really weird with 3200MT/s on some boards, your CPU should not be but it's a starting point to work up from in case the BIOS has some buggy defaults.

 

[depakjan]

Thanks for your information.

It's *a* voltage drooping. The better PSU sags less, so the relevant hardware after it has less trouble. This is why some people have more problems than others - a better PSU can make weak motherboards perform better.

Funnily enough my Old PSU is a Corsair RM850 (supposed to be Tier A) and the new one in a Deepccol PM850d( tier B speculative) , i assumed maybe MB is spiking the voltage and the Corsair PSU was restarting to avoid any component and the Deepcool is slow to restart

Definitely set the ram to 2933 and SoC to 1.15v and see if it helps. My 3700x is really weird with 3200MT/s on some boards, your CPU should not be but it's a starting point to work up from in case the BIOS has some buggy defaults.

When i removed and kept only 1 RAM, it already defaulted to2933 mhz and it still restarted.

Ii recently changed from RTX 3060ti to RXT 3080ti , so upgraded my PSU from 650w to 850W, not sure if this MB couldn't handle this load

 

[depakjan]

Hi All, My Motherboard cam back from RMA and so far so good. they said one SMD was weak and they had to replace that one. previously my pc wont boot with the 4th stick of RAM, I thought it was due to incompatibility with the RAM module but now it boots with all 4, seems something was wrong even before but other than that one RAM everything was working fine.

 

[forman313]

Yes I took out the CMOS battery for 30 mins and yea I am currently in barebones setup. I checked the MB for any swollen capacitors I cant see any but then I am not that good in the this area. Below is how it is

ignore the missing screws and the CPU fan hanging around, I was thinking maybe I tightened the cooler screws too much, so I was testing it this way as well.
View attachment 313621

Thats not cleaning. You have shoved the dust and dirt around and instead of having a thin layer of harmless dust, you now have large deposits here and there. A large portion of it is probably wedged inn under slots and components. Add inn moisture and you get problems. Next time get a toothbrush or similar and start cleaning, as in removing the dirt

Happy to see it all worked out.

 

[faye]

Hi All, My Motherboard cam back from RMA and so far so good. they said one SMD was weak and they had to replace that one. previously my pc wont boot with the 4th stick of RAM, I thought it was due to incompatibility with the RAM module but now it boots with all 4, seems something was wrong even before but other than that one RAM everything was working fine.

Glad it's working for you. My only advice at this point would be don't buy MSI for your next build - they often skimp on power delivery and do a bunch of other shady stuff as a company. Yes, they tend to have the cheapest options in a given category, but it's very much a case of you get what you pay for.

This is a generalization of course, but I've had a couple over the years and they've been more problematic than most. Others here seem to have had similar experiences.

 

[wheresmycar]

Glad it's working for you. My only advice at this point would be don't buy MSI for your next build - they often skimp on power delivery and do a bunch of other shady stuff as a company. Yes, they tend to have the cheapest options in a given category, but it's very much a case of you get what you pay for.

This is a generalization of course, but I've had a couple over the years and they've been more problematic than most. Others here seem to have had similar experiences.

For me the Mortar, Bazooka, Gaming Plus and Tomahawk class of MSI boards have always been excellent. I would still buy MSI although when buying I tend to check reviews thoroughly or when not available user feedback. I guess with each generation the build quality/features vary and im open to all. The worst of the boards ive used is a budget mATX ASROCK one (horrible experience a few years ago) and oddly enough the best board with zero issues to report on was a ASROCK Taichi (flawless experience).

Anyway i'm a casual user and run mostly everything on default. I guess extreme users who require more powerful and feature rich boards will vary in their opinions (and rightly so).

 

[ir_cow]

Everyone says don't buy from one company or another. Every brand has some crappy ones though.

 

[boomstik360]

Everyone says don't buy from one company or another. Every brand has some crappy ones though.

I agree. I've been building PC's since 2005 and used MSI pretty regularly (not exclusively at all) for GPU's and Mobo's without a single issue. Always liked the ease of use of their website. I have only had to RMA one part ever, and that was a Gigabyte LGA 1156 motherboard a LONG time ago lol and it was a used board I bought on ebay that still had warranty left. Luckily I've never had anything die on me or had something come DOA. Weird

 

[Mussels]

Everyone says don't buy from one company or another. Every brand has some crappy ones though.

Check out brands and find out their product stack, then aim for lower end parts in their mid-range series

Asus have
TUF (bottom tier)
Strix (Meaningless, usually budget products)
ROG Strix/ROG/Strix (Used to be high end, now midrange. See 'ROG maximus' for an example of how they made their own naming and broke it)
Crosshair/Maximus - Their ultimate AMD and intel series.


You'll get a far better product with the cheapest Strix than the best TUF, is the point being made here.
The worst crosshair/maximus will make the others look bad, etc.


This random image i found googling asus motherboards has two midrange, one top of the line and two budget.
You could probably make some assumptions about their VRM's and peripherals like LAN and Wifi from the series names alone.

 

[Waldorf]

@depakjan
did you carry win over, or fresh install after changing hw (not counting gpu).
did you have similar issues before swapping gpu?

most psus have more trouble with gpu spikes, than load.

@faye
in the last 2y, not a single 2-300$ Gb board used was worth its money, the Ultra not even if free (would not even run gskill kit 3600C16/1.35, and clocks/volt fluctuations on auto)
while non of the lowest grade MSIs made any trouble, and no, talking about swapping 5950 on those.
and once you exclude the lowest tier, no issues on pwm/vrms, no review seen that supports it, unless ocing 5950x.

 

[uco73]

I have same problem, but after CLEAR CMOS everything's back to normal.