• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

i7-4790K Heating issues. I'm an idiot so could really use some help

Evolutiondag

Distinguished
Jun 14, 2016
16
0
18,510
Hi guys, i made a similar post in the motherboard section because i was concerned about a sensor temperature on Corsair Link about my motherboard, but i decided to do a much more detailed one here in case that trouble is caused by this.

Essentially i built this computer around august last year, i had no trouble with it for quite a while, but i think an issue has popped up over the last month or so, i'm not 100% sure how long because i only actually felt the heat yesterday from one of my vents. So originally my CPU has ran idle temps around the low 20s, now though i am seeing it constantly idling between 40 - 50 mostly, even when i turnt on my PC earlier within the first couple of minutes of loading desktop i saw the sensor say the CPU cores were around 80c. I am really confused with what the problem could be, running a game like Witcher 3 for E.G is putting the CPU around high 70s/low 80s. Beforehand it was in the 50s i think on that game.

A tech friend of mine believes it to be that my H100i GTX Water cooling has failed somehow, but it still worries me if it is because i'd have to send for an RMA and hope that's the issue. He helped me build the computer because i'm not very knowledgeable. Anyway the thing that has confused me is, i asked for his help yesterday because of the temperatures, he came round and took the water cooler off and the radiator, sprayed them out for dust and reapplied thermal paste. After booting up my temperatures seemed normal again, they was idling around low/mid 20s and i played some games and sat around mid 50s or so. The reason i'm really confused is, i turnt the computer off went to bed and woke up today to the same problem again, 40 - 50 idle temps and 80+ in games. Does anyone have any idea what might be going? I don't know what to do really, please talk to me like i am an idiot if possible because honestly i am not great with this stuff.

This is a Gyazo screenshot of Corsair Link right now. https://gyazo.com/17e155010d447ccc4cc7436705572e6d
For some reason that Temp#4 on the motherboard heats up way more than the other 3 which might as well be static. The Temp#4 has already reached upwards of 70c which worries me.

My specs are:

HyperX Beast 1600MHz DDR3 4 x 4GB for 16GB
Intel Core i7 4790K
Asus MAXIMUS VII RANGER Z97 Motherboard
EVGA GeForce GTX 980 Ti Superclocked
Corsair RM850 PSU
Corsair Hydro H100i GTX Liquid CPU Cooler
Corsair Graphite 780T Full Tower ATX PC Case

If there's anything else anyone wants listed, just let me know. As far as i can tell it would be these parts that are of importance.
 
Normally this CPU gets really hot but considering you're on water cooling it's quite noticeable that it's acting like this. Now I do know that it's about 2 things on the top of my head. The sticker on the plate is still on. Yeah that happens.2. It's not mounted in properly.
 


Is there anyway surefire way for me to tell. I've felt the hose and feel like i can feel water moving but i can't say 100%. Of course it's hard to listen to with fans
 


Funnily enough there is a little bit of sticker on two parts, one near bottom of the board and another on the CPU plate, but what i'm confused about with that is, they would have been there the whole time since i had it, so why would it act up now as well and be fine yesterday after taking it out and putting it back in?
 


One idea would be to use a regular old heatsink cooler and see if that helps. If it does then you will know it has something to do with your watercooler.
 


You'll hit 90c for sure with intels stock cooler, its total junk.

On a hot day 70c from a 100i isn't unheard of. After a few years any AIO cooler drys out, I replaced a number of them already in friends computers. Its easier to just go with air unless you are serious about overclocking.

If the pump went bad you would be see 90c+ and if the fan controller died either they would be full speed always (which has happened to me twice) or they don't spin at all. I'm thinking if its not heat related (summer time) or bad fan controller. than your 100i is runner out of fluids and starting to heat up.

If you have taken it apart, make sure to apply new thermal paste because you cannot reuse whats on it already.
 


I haven't even had it for 1 year is the sad thing. I could understand it being a hot day and all obviously ambient temperature affects it, but it's been at these temps when i've had rain and a storm with open windows. Thanks for the information about them drying out though, i'll keep that in mind. Would you happen to have any idea about what the issue may be with Temp#4 on the motherboard sensor? It's the odd one out considering it's like 20 - 30c higher on average than the other 3
 
I think there maybe some confusion. Temp #4 is on the motherboard and is not a core temp. Therefore this has nothing to do with your CPU and H100i GTX. It's hard to say what Temp #4 is measuring. It could be a VRM if it's getting that hot. Have a look in your BIOS and see what each sensor is monitoring on the motherboard (not the CPU). Hopefully there will be some labeling that will shed some light as to what that temp is for.

I will have a look at you motherboard manual and see if I can make an educated guess.
 
OK, well I've had a look at the manual, and it looks like there are three sensors on the motherboard. You have CPU package (thermal diode), MB (presumably for southbridge), and Sensor 1 which is an external sensor. Sensor 1 is for a temp sensor wired to a connector on the motherboard (see top of 1-47 of the mobo manual), and it's likely you aren't using it.

So there are a few possibilities. Temp #4 could be reading something from Sensor 1's input, or there is an unused temp sensor input on the heath monitor IC. In either case you should be able to safely ignore that sensor as it's just reading garbage. This happens when you use monitoring utilities that haven't be specifically written for your motherboard. I'm sure if you used Ai Suite III, you wouldn't find this temp listed anywhere. However don't take that as an endorsement of Ai Suite, because in my opinion it's garbage (caused BSOD's and events in Event Viewer). I only mentioned it because it is written exclusively for your motherboard.
 


Ok thank you a lot. That's one bit to be added for piece of mind. I really appreciate that a lot. Thanks so much
 
Yeah mine friend computer had a antec kuhler in a computer that was "overheating". I check the Kuhler and noticed it felt very light and no signs of fluids inside it. So I decided to cut it open and it was nearly empty lol. It completely evaporated after only 2 years. No leaks expect for where I cut it open.

My friends Corsair H50 was similar, though it had enough fluid to keep it going for maybe 6 more months. These AIO coolers only last 3 years if yo have it on all the time without refilling them.
 


His issue isn't with CPU temps, that Temp 4 reading is on the motherboard and it's a bogus input.
 


His motherboard only has three valid temp inputs. CPU package, MB and Sensor 1. The fourth is likely an unused channel on the health monitoring IC. I've seen this with a few other motherboards. The utility detects the IC, and knows how many channels are available on the device, but it doesn't know which channels are actually used. These usually behave erratically. Sometimes they'll read "normal" and then next it will go up to some unrealistic value or even down (negative). The only way to know for sure would be for the OP to use Ai Suite (albeit temporarily) Ai Suite to see if it shows a corresponding temp value.
 


Ok.. Yeah I have had a couple boards that had funny temp sensors too.
 


There is 5 fans in the case
 

Oh I I see, I thought this was a new build but yeah thanks for bring that up. Hm, right about now all I can think of using another cooler to see a comparison.