New Build - Random overheating

[KimSim]

I have built over 20 systems over the years, and the one I just built has me puzzled. First the specs.

NZXT H700 Case

ASUS ROG Crosshair VII Hero X470 Motherboard

AMD 2700X CPU

CORSAIR HYDRO Series H100i PRO AIO

64GB Corsair Vengence LPX Memory (DDR4 3200Mhz)

ASUS GeForce RTX 2070 8G EVO Turbo Edition

EVGA 1000 GQ, 80+ Gold 1000W Power Supply

HP EX950 M.2 1TB PCIe 3.1 x4 NVMe 3D TLC NAND Boot drive

Seagate Barracuda ZA2000CM10002 2 TB Solid State Drive Data Drive

I had a 140mm Noctua NF-A14 fan exhausting at the back,  (2) 120mm Noctua NF-F12 PWM fans (intake) on the radiator in front, plus a 120mm Noctua NF-A12 fan (intake) under the radiator in front as an fresh air intake.

 

Ok, I haven't overclocked anything, all I have done is changed to memory profile to take advantage of the 3200Mhz memory. I have updated the BIOS so it is the latest. Otherwise the BIOS is stock. I didn't install any of the tweaking software, gaming software, etc. I depend on BIOS for my fan control.

 

The thing seems to run fine for days, but then will shut down for no apparent reason.  I had to physically unplug the power supply and plug it back in to restart it.

So I replaced the power supply - still happened.

I replaced the motherboard - at first it seemed to fix the problem except now it seems to run fine for many hours or days nice and cool (35-40 at idle, 60-70 under load) but then overheats for no apparent reason, sometimes while sitting idle, and shuts down.

I cleaned and reapplied the thermal paste, still does it.

I added 3 more 120mm Noctua NF-A12 fans to the top of the case as more fresh air intake.  Still overheats. (standard fan settings in BIOS)

I added 2 more 120mm Noctua NF-F12 fans to the radiator (push pull) and jacked up the minimum speed to 50% in bios on the cpu fans  

 

Today I was doing a simple spreadsheet (not paying attention to temps) and first excel shuts down out of the blue, then after restarting Excel I got a blue screen of death and the computer shut down.  That was when I noticed the smell of hot electronics.  After a few minutes I restarted it and the temps dropped as the fans cooled things down.

The only thing I haven't replaced are the drives, the cpu, gpu, and the memory. I don't think the drives are the culprit. The only thing I can do next is replace the cpu and see if the problems persist.  I have a AMD 3600 laying around so I may try that.  Any other suggestions?

 

 

[Jurrunio]

...overheats? 70C doesnt count as overheating, in fact I dont think the board will shutdown until it hits 100C, more on the VRM.

 

Run hwinfo64 in sensor mode, and keep all the reported temperatures monitored.

 

3 minutes ago, KimSim said:

That was when I noticed the smell of hot electronics. 

New hardware can sometimes smell chemical, not a big concern.

[Windows7ge]

Did you try a different CPU cooler? It sounds like the pump is failing.

[KimSim]

9 hours ago, Windows7ge said:

Did you try a different CPU cooler? It sounds like the pump is failing.

Yeah, I was thinking about that last night.  I think I will throw on the stock Spire cooler to see if the problems persist. It could be that I got a bad Corsair H100i. I have used them for years with no problem, but because this problem has persisted across all the other components, I think it may be the cooler itself.

[KimSim]

10 hours ago, Jurrunio said:

...overheats? 70C doesnt count as overheating, in fact I dont think the board will shutdown until it hits 100C, more on the VRM.

 

Run hwinfo64 in sensor mode, and keep all the reported temperatures monitored.

 

New hardware can sometimes smell chemical, not a big concern.

No, 70s are where they are under load when I test it.  I am very happy with that. But randomly it overheats to the point of shutdown even when idle. Almost like the AIO pump stops running, so I think I may try the stock cooler for a few days to see if the problem persists.  It could be that I have a bad AIO that has intermittent issues.

[KimSim]

OK, yesterday I pulled the Corsair H100i Pro AIO and installed a AMD stock cooler from a 3700X which I had laying around.  It runs slightly hotter at idle but so far I haven't had any shutdown from overheating. Highest temp reached is 70C, even during benchmarking. I guess I got a bad AIO out of the box. It either has a pump that randomly stops or it has something floating in it that restricts the flow from time to time, causing the temperature spike and subsequent shutdown.  Now I just hope I haven't damaged the CPU by the multiple overtemps.

[KimSim]

After three days running fine with the stock AMD cooler, I installed a Cooler Master Hyper 212 Black Edition. Now my temps are matching what they were with the AIO, even better!  When running cinnebench, I only get a max temp of 66C. That Corsair H100i Pro AIO must have been bad right from the factory. Too bad I have had it too long to get my money back.