HELP: Heat issues on AIO cooled i7-4790k

[Hoover1979]

Hi. This is my first thread here, but in recent weeks I have been getting overheating issues on my i7-4790k.

For cooling, I have a Corsair H115i 280mm radiator dual-fan CPU cooler.
My Motherboard is an ASUS Z97-k.
I also have 32Gb of Corsair Vengeance Pro 2400 DDR3 RAM.
Power Supply is a Corsair HX850i.
My GPU is a Gigabyte Xtreme Gaming GTX1080 Waterforce rev 2, connected to an ASUS VG278QR via Displayport.
Windows 10 v1903 is installed on my boot drive which is a 256Gb Samsung 860 pro SSD.
The sound card is a Sound Blaster Zx connected to Logitech Z906 speakers as 5.1
The chassis is a Fractal Design Define XL-R2 Black pearl full-tower case.
The optical drive is a Pioneer Blu-ray burner
I also have 3 HDD's installed. A WD 1.5Tb, a WD 4Tb and a WD gold enterprise 10Tb.

I have recently started to get BSOD crashes (0xc00000124 WHEA Uncorrectable Error) when performing CPU intensive tasks like rendering filters in Photoshop CC2019 to use in my UltraHD Texture Pack I have been making for the original Doom, and also when using Format Factory or Handbrake to convert audio or video files. Sometimes I get a BSOD during bootup and sometimes it happens randomly when playing a game, watching a movie or even idling at the desktop.

At first, I didn't know it was heating related, as Corsair Link was set to adjust fans due to the temperature of the H115i cooler itself and not the CPU, but a few days ago I configured it to adjust the fans according to CPU core #1 and instantly the RGB display went red and the fans took off like a jet. I then rebooted and on startup ran HWmonitor and was aghast to see the temps hitting 95 degrees while booting up. I went into MSConfig and disabled most non-Microsoft services, most startup programs, shut down, went into BIOS, disabled XMP tuning, disabled turbo mode entirely and ran the PC again, this time with the side panel removed and it gets to 55 degrees during boot up, idles at between 48-52 degrees, and goes into the 60 mark when I launce Firefox, especially if I go onto Facebook on firefox. I tried to drop the core voltage offset a tad in Intel XTU but it instantly triggered a BSOD so I won't be increasing or decreasing voltage from hereon in.

I used to be able to render filters in Photoshop for hours at a time and convert large videos (particularly converting my Blu-Rays to media files for my SHIELD tablet), without issue, but now within 10 seconds the PC freezes and I get a BSOD and it takes a large amount of time to collect the dump and the MEMORY.DMP is usually between 1-3Gb in size.

I can't move the PC to remount the cooler as I have a bad back (from a car crash in 1996) and my PC is too heavy for me to lift off the ground without putting myself in the hospital but I am wondering is there anything more I can do or is this hardware failure and do I need a new Motherboard, CPU and RAM.

 

[trents]

Sounds to me like your AIO pump isn't pumping coolant to the radiator such that the radiator remains cool but the processor temp skyrockets. If the pump rpm is where it should be then there could be a blockage in the hoses or radiator do to an air bubble or algae buildup. How old is the AIO?

I would also check your CPU core voltage in bios to make sure it isn't excessive. If something went wrong with a voltage regulation component on the motherboard such that it significantly increased the CPU core voltage this would explain the high temps.

 

[Hoover1979]

I powered off my rig and gave it a bit of a shake and drummed my fingers up and down the water pipes for the cooler and then went to BIOS and in BIOS the temp was reading 32 degrees, but the instant I went into Windows and immediately fired up HWmonitor it was still between 45-60 degrees. What would the voltage need to be? I am no longer game to adjust voltages due to every attempt I made so far caused BSOD (0xc00000124 WHEA Uncorrectable Error) crashes.

 

[trents]

I think stock would be somewhere 1.2-1.25 volts.

Idle temps don't mean much unless they are abnormally high.

How old is the AIO cooler? I'm still thinking the cooler is the problem.

 

[Hoover1979]

The cooler was purchased in 2018 after my Noctua NH-L12 conked out.

 

[EarthDog]

Try running a stress test once in windows with a temperature app up and see what it does.

In the meantime, I'd make sure the bios is at default and perhaps update it to the newest version as well.

X0124 is typically a lack of vcore issue, at least when overclocking.

 

[EarthDog]

Try running a stress test once in windows with a temperature app up and see what it does.

In the meantime, I'd make sure the bios is at default and perhaps update it to the newest version as well.

X0124 is typically a lack of vcore issue, at least when overclocking.

 

[maxfly]

Make sure your cpu block isnt loose or installed incorrectly. Is your cpu running at stock speed now? Just clarifying in case i missed it being mentioned earlier.

 

[trents]

The cooler was purchased in 2018 after my Noctua NH-L12 conked out.

Your Noctua NH-12 conked out? That's an air cooler. What's there to go wrong with it? I note that is a compact cooler made for SFF systems. Is this a small form factor system or ITX build? I note also from your original post that your temps dropped considerably after you removed the side panel, though you made some voltage/frequency adjustments at the same time so hard to tell which was most responsible for the temp drop.

Has this system been overclocked?

 

[Hoover1979]

The system is not overclocked, it was on the factory overclock but I disabled XMP tuning and disabled Turbo mode. I never pushed it past the factory turbo 4.4Ghz mode. i did a test with the side panel off and on last night and the temp difference was only a couple of degrees, so disabling XMP tuning and turbo mode is what got the temps down, but they are still to high for comfort as when I play a game the CPU gets into the 70-80 degree mark, and I am not game to use my games that have CPU hogging DRM like Denuvo, while I am getting these temps. Last time I tried to play a Denuvo protected game (Star Wars Battlefront 2015) I had a BSOD crash (0xc0000124 WHEA Uncorrectable Error) while adjusting the controller sensitivity in the menu.

The NH-L12 was a remnant of my smaller case and went back into my build when I got my full-tower case. It eventually stopped working and when I was booting up into windows (Windows 7 at the time) the temps skyrocketed and the system BSOD'd within seconds of booting up, there was what looked like heat damage on the cooling pipes. My CPU has taken a beating over the years as not only am I a gamer but I am also a modder and have been working on UltraHD texture replacements for the original 1993 Doom and when I render filters through the Filter Forge plugin for Photoshop CC, it pushed the CPU to 100% usage sometimes for several hours at a time and this has been repeated frequently from 2016 onwards until recently when the system BSOD's when I try to render a filter. I decided to get the most expensive AIO cooler I could at that time (sometime in 2018, I can't remember the exact date) from my local MSY store, hoping it could keep my CPU cool due to the load I usually put it under and everything was coming up roses until a few weeks ago.

I already set the bios to default a couple of weeks ago and it was a headache as it stopped listing my SSD as a boot drive and I had to set it manually as it wasn't in the list of bootable drives and the PC was acting like it had no OS until I fixed that. I never went into looking at voltages in BIOS and I am loathed to toy with voltages as when I tried it on intel XTU (which helped keep the heat down on the Acer Predator Helios 300 laptop) even a single increment drop of the core voltage instantly triggered a BSOD (0xc00000124 WHEA Uncorrectable Error).

In a weeks time I am taking my PC into my repair shop and if it turns out the CPU is faulty, I will likely try to get the funds together for a 9th gen i7, a new Mobo and 32Gb of DDR4, as I really need to get back to making my textures as I have amassed a large following on ModDB, and don't want to let people down.

 

[maxfly]

Its not a big deal adjusting your cpu and memory voltages. The worst that can happen is your rig will reboot to the bios defaults. Thats not going to happen because once we know your bios settings we will get them set up correctly. so we can move on to the next troubleshooting step.
The only way to test your system for stability is to first find out what speed your memory is running at. Then find out what voltage its running at. Once we know if it needs to be adjusted we will walk you thru the steps to correct any issues.
If the memory checks out we will move on to checking what speed your cpu is running at and what voltage. Again if they need to be adjusted we will walk you thru changing them.
Dont sweat it these are literally the easiest troubleshooting steps.
Do you know where in your bios to find the cpu and memory settings?

 

[Hoover1979]

The memory was running at 2400Mhz when I had XMP tuning on, now it got set to 1333mhz by BIOS when I switched from XMP to Auto, CPU Voltage is reading as just over 1.2 (I forgot the exact number as I have short term memory problems from an acquired hypoxic brain injury from October 2000, but I just checked it in BIOS couple of mins ago, and it was 1.2??

 

[trents]

Hoover, I hope if and when you get the new 9th generation stuff you don't find that it was the AIO cooler that was failing. But then again, it's probably time to upgrade as your current rig is getting long in the tooth, especially given your uses. But be aware that you might have to wind up replacing the cooler as well. With your back issues, who would do the new build?

 

[Hoover1979]

I used to do all of it myself but these days my back got so bad there is a shop that's a fair bit from my house but the techs do a good job and charge a fair price with cheap labor charges. It's best to pay a little extra in labor and not end up in a hospital these days than to do the work myself and save a few bucks. They can also test the cooler and if it needs replacing I will get it added to the list, but first I would need to get the funds together and that's easier said than done. Hopefully a 9th gen i7 can run 3.2Ghz DDR4 at full speed.

 

[EarthDog]

Hopefully a 9th gen i7 can run 3.2Ghz DDR4 at full speed.

It will have zero problems running that speed.

 

[Hoover1979]

My rig is going into the shop today. Hopefully, they can help me get this fixed with the minimum of expense.

 

[Hoover1979]

The CPU was faulty so I now have a Ryzen 9 3900X CPU, a Gigabyte Aorus X570 Pro Wifi MoBo, 32Gb of Crucial Ballistix 3200 DDR4 Ram (dual-channel) and I a Deepcool Castle 280RGB AIO Cooler. I also got a 500Gb Samsung 970 EVO Plus M.2 SSD drive and am now running the OS on that. Once I have everything reinstalled, set up, and reactivated I will see how much better performance my 12 core (24 thread) CPU gives when rendering filters for my texture work. For some reason Filter Forge 8 won't add itself to the plugins for Photoshop CC2020 so I have to use Photoshop CC2019 until Adobe fixes this.

 

[DaveB]

My rig is going into the shop today. Hopefully, they can help me get this fixed with the minimum of expense.

The CPU was faulty so I now have a Ryzen 9 3900X CPU, a Gigabyte Aorus X570 Pro Wifi MoBo, 32Gb of Crucial Ballistix 3200 DDR4 Ram (dual-channel) and I a Deepcool Castle 280RGB AIO Cooler. I also got a 500Gb Samsung 970 EVO Plus M.2 SSD drive and am now running the OS on that. Once I have everything reinstalled, set up, and reactivated I will see how much better performance my 12 core (24 thread) CPU gives when rendering filters for my texture work. For some reason Filter Forge 8 won't add itself to the plugins for Photoshop CC2020 so I have to use Photoshop CC2019 until Adobe fixes this.

Minimum of expense!!!!!!!!!!!!!

If it was just the CPU, a used i7-4790K on ebay goes for under $200.

Nice new setup, just a bit pricey.