Got a gigantic temp spike on both my CPU and Motherboard during a game. Can someone look at this real fast?

[Triburos]

Hey folks.

So I just had a gigantic jump in temperatures as I was playing a game. Happened about 10 or so minutes into the game. I've just been getting over what I thought was a fixed overheat problem, but it seems not.

Just for some normal context; I usually play on about 50 to 60 in terms of CPU temperature, and around the 60 range for the board. Temperatures for folk seem to vary wildly on weither or not the range is acceptable or not. Some say this is fine, some say its a bit too hot.

Personally, that range is fine for me.

What ISN'T fine is when I start jumping as high as 80c on the CPU and even what I think was 90c on the board.

This came out of fucking nowhere, and I immediately shut down the game got up and looked inside my case, and the fans are still spinning. After I shut the game down, the temperatures dropped back down to normal So I'm confused what might have caused this.

Tis worth noting; I use 2 fans; one case fan and one heatsink fan. This obviously isn't counting the fans within the graphics card and power supply.

Both are pretty stock. I'm aware this isn't ideal, however I'm not the kinda guy who goes for advanced builds and premium perfect airflow. I just want the PC to not struggle and work. But if it's time for a replacement, I'm all for it.


Anywho, here's a Gyazo of Speccy (just to show some of my specs) and on the right is temperatures from HWMonitor.

You can see that even though I'm at perfect idle temperatures at the time of writing (imo anyway), the Max values on HWMonitor show just how high I could have potentially hit had I not shut down my game when I had that spike.

Notice that I'm reading absurd MB temperatures (an idle of -88c?)

Restarting HWMonitor has returned my MB minimum temperature values back to normal. No clue why, but Here's the image

I'm personally suspecting an issue with VRMs on the board. I find it hard to believe both the CPU and MB were rising so drastically at the same time and both are failing at the same time. But I need more input from you folk.

Anyone else feel the same? Thing is; I'm not sure if a faulty / dusty / poorly mounted heatsink can also cause the motherboard temperature to rise along with the temperature of the CPU. Does anyone know if that can happen?

 

[Rooster__]

Your cpu didn't overheat that much, 80oC is not that high for a cpu- unless you have benchmarked it on full load at a lower temperature? I would try that with aida64 and some temperature monitor software. Same with the motherboard. The -88C might be a random error, notice it is a minimum, and board temp is 43C. All seems very odd, but run the benchmarks for an hour or so and report the temperature at the end of that period of time. It does seem like you are having some problems with detection on the motherboard, however temperatures for some motherboard components are built to withstand much higher temperatures than that of a cpu- up to 150C.

 

[Triburos]

Your cpu didn't overheat that much, 80oC is not that high for a cpu- unless you have benchmarked it on full load at a lower temperature? I would try that with aida64 and some temperature monitor software. Same with the motherboard. The -88C might be a random error, notice it is a minimum, and board temp is 43C. All seems very odd, but run the benchmarks for an hour or so and report the temperature at the end of that period of time. It does seem like you are having some problems with detection on the motherboard, however temperatures for some motherboard components are built to withstand much higher temperatures than that of a cpu- up to 150C.

Heya, thanks for the input.

As I said, temperatures seem to vary from folk to folk. Some say 80c is perfectly fine, some say it's too hot. Some say even 60c under load is too hot for a CPU.

I personally say that a CPU should never go above 70C under load. Not sayin' I'm right, it's just that after looking at some tables of recommended thermal limits for various CPU manufacturers, the over-all pattern seems to be that CPUs should try to stay at 70c or below whilst under load.

And while 80c might not be too bad to you, it's concerning to me since I've never had a spike that high. I've always been running at 60c at most according to Speccy, HWMonitor, and so on. So there is a problem going on if I'm all of a sudden getting gigantic boosts in temperatures not just to the CPU, but to the motherboard as well.

I've also been told that motherboards tend to be cooler than the CPU, so it's strange that mine is typically warmer by about 10-15c on average.

Anywho, the reason it only got to 80C is because I quickly shut the game off. It coulda gotten higher if the max value on HWMonitor is anything to go by. Note that HWM says it could have gone as high as 100C+. That's not normal. Of course, HWMonitor also said it could reach -88c, which is obviously impossible. So perhaps that was just a temp sensor malfunction.

What's most worrying is that BOTH the motherboard and cpu temperatures started rising at the same time. But from what I understand, CPU and MB have different temperature sensors and drivers. Meaning one was potentially causing the other one to heat up. Which is why I'm thinking the VRMs are putting out too much power. That would heat the board and since VRMs are directly under the CPU, it'd also heat the CPU.

 

[nuts32605]

The numbers that HWmonitor samples come from a very simple chip on the motherboard. Any little transient jumps in the signals it reads will cause really weird numbers to come out. That's why you're seeing all kinds of just plain wrong data in HWmonitor.

For example, FANIN2 doesn't appear to have a fan plugged into it, and yet has a max RPM of 2647? And I'm sure you don't have a gas turbine or a turbocharger in your system so 675,000 RPM from a fan is completely insane.

The same thing is happening in the Temperatures section, If something was -126 degrees or 250 degrees, you would've noticed. The package temperatures from Vishera CPUs I've never really seen them as very accurate. I have a FX-9370 and HWmonitor says it's min is ~10 degrees and it bounces around constantly. I doubt it's 57 degrees in your house.

My advice is to ignore the Package temp and watch the value, not the min or max too much, of the 3 temp sensors above.

 

[Triburos]

The numbers that HWmonitor samples come from a very simple chip on the motherboard. Any little transient jumps in the signals it reads will cause really weird numbers to come out. That's why you're seeing all kinds of just plain wrong data in HWmonitor.

For example, FANIN2 doesn't appear to have a fan plugged into it, and yet has a max RPM of 2647? And I'm sure you don't have a gas turbine or a turbocharger in your system so 675,000 RPM from a fan is completely insane.

The same thing is happening in the Temperatures section, If something was -126 degrees or 250 degrees, you would've noticed. The package temperatures from Vishera CPUs I've never really seen them as very accurate. I have a FX-9370 and HWmonitor says it's min is ~10 degrees and it bounces around constantly. I doubt it's 57 degrees in your house.

My advice is to ignore the Package temp and watch the value, not the min or max too much, of the 3 temp sensors above.

Lmao, holy shit I only just noticed the RPM of one of the fans. Man that's funny.

But hey, thanks for the advice!

But I'm still kinda concerned. I defo felt a shit ton of heat from my PC case, much more than usual when this happened. Even if the temps might not be incredibly accurate, they did tell me that the temperatures were atleast going haywire. Weither the values were exact or not, I dunno.

It's just worrying. When you've always seen around 60C for CPU temps and around mid to mid high 60C for your motherboard, then you all of a sudden see 80C and 90C pop up for no reason PLUS you have physical heat evidence proving that its not just faulty program readings, that causes me to worry.

I know that overheating nowadays doesn't cause much damage thanks to failsafes, but it's the failsafes that bug me. It's no fun to have your PC shut down when you're trying to do stuff because a certain temperature got too high y'know?

But you are without a doubt correct. I've always known that these temperatures can be a bit funky. Trying to say that it's 57C when my room temperature is around 70F is obviously off-putting. -88c is taking the piss.

It's just that I have physical evidence that the heat was actually rising and that this wasn't just a 100% malfunction of the sensors. I felt an enormous amount of heat from my case after all. Much more than normal for gaming.

 

[nuts32605]

I took a look at your specs again, and recognized your motherboard. Ye olde M5A78L-M LX Plus. Made back in the Phenom II/Core i5-550 days. Couldn't get that thing to run right with a Phenom || X4 960T until I added heatsinks to the VRMs. ASUS says 120W, but it's more comfortable at 95W. It's only a 3+1 VRM setup.

You need to get some kind of forced induction down there. Fresh, cool, outside air into the CPU area. If you're still using the stock downdraft CPU heatsink that's going to blow all the hot air from the CPU down to the motherboard around it, including the VRMs. I finally got mine happy with a 92mm tower-style CPU heatsink and little individual heatsinks for the MOSFETs in the VRMs.

 

[Triburos]

I took a look at your specs again, and recognized your motherboard. Ye olde M5A78L-M LX Plus. Made back in the Phenom II/Core i5-550 days. Couldn't get that thing to run right with a Phenom || X4 960T until I added heatsinks to the VRMs. ASUS says 120W, but it's more comfortable at 95W. It's only a 3+1 VRM setup.

You need to get some kind of forced induction down there. Fresh, cool, outside air into the CPU area. If you're still using the stock downdraft CPU heatsink that's going to blow all the hot air from the CPU down to the motherboard around it, including the VRMs. I finally got mine happy with a 92mm tower-style CPU heatsink and little individual heatsinks for the MOSFETs in the VRMs.

Aye. It's defo a budget board I ended up getting with some bare bones package I got off of Tiger Direct at some point. I knew it'd end up causing issue at some point after I did some looking into it. But I had already built the rig so I had to stick with what I got 😛.

And indeed, the fan and sink I'm using is stock. I felt that as long as it works, it's fine. I'm not into anything major like water-cooling, y'see. So I typically just grab whatever can work, and then pray that it works for a healthy amount of time.

I think I getcha so far. But just some confirmation;

1. Would yah recommend leaving the side panel off for the time being? I'm not too fond of that due to disrupting the air-flow, but then again my air flow isn't that amazing either. Read below for more info on that. But when you said getting some clean air into the case, that's my best guess as to how I should do that.

2. Seems like I defo want to replace the heatsink and fan regardless. I was gonna do this anyway, so why not? If so, any clue what you'd recommend for this particular mobo? As much as I'd love to replace the entire board, it's not in the cards right now. So I'm just looking for a bandaid as it were. I have a mid-sized tower.

Funny story about the air flow in my case actually. I was having a hell of a hard time fitting in my PSU whilst I was building this PC. Not the case's fault, just the PSU having its wires come out of a side rather than the middle made fitting it in how I wanted almost impossible.

Long story short, my case has a few pre-built holes at the top of it, and I was having issues having the PSU to have the fan point down (this case uses a bottom mounted PSU, with a mesh material at the bottom to blow hot air from the unit out of the case I believe), but due to issues I had to mount the PSU to blow it's air up. Aka; past the motherboard, GPU, heatsink, and fan.

Normally I wouldn't have done this since all that air would be bounced back downward, but again; there's holes at the top. So the PSU fan actually blows both it's own air, and any air that the heat-sink fan blows out out of the top of the case.

I have NO idea if this is ideal. It could be a complete asinine method of air flow for all I know. Hell, this could be HOW you were ment to build it all along and I just assume I messed up. But I figured it was worth noting because that's actually how I can tell if heat build-up is getting ridiculous without relying on programs. Plus it was extremely hard to fit in the PSU due to wire arrangement if I didn't do that.

The PSU blows the heated air from the heatsink and fan up through the holes and I can just hover my hand over the top of my case to feel. When this issue happened, I swear I could probably cook an egg on the top of my case. Normally when my CPU and MB temperatures are normal and idle, it's very luke warm. During gaming, it gets somewhat hot.

But this time, I felt extreme heat hense why I'm a tad worried that this wasn't just bad readings on a program's part y'know?

 

[nuts32605]

That PSU fan is an intake fan, not an exhaust. So it's taking the air at the bottom of the case and removing it out the back. So there's even less air left inside for cooling the CPU. I'd leave the cover off, computer fans are not "suckers", they can't really oppose an air restriction. Taking that side panel off and letting cool air in is going to do a lot more then any air stream effect is.

I know you're concerned about the heat output, but if you're not getting the hot air out of the case and cool air in, then it's not circulating. If it's not circulating, then it's baking itself with it's own exhaust. Heatsinks can only manage so much heat, if you reach the top of your cooling "budget", then you need to either add a cooler with more surface area or get more/colder stuff for the heat to transfer to (i.e. more cold, dense air).

As far as modifications go, I'll look up the parts I used on mine and get back to you.

 

[Triburos]

That PSU fan is an intake fan, not an exhaust. So it's taking the air at the bottom of the case and removing it out the back. So there's even less air left inside for cooling the CPU. I'd leave the cover off, computer fans are not "suckers", they can't really oppose an air restriction. Taking that side panel off and letting cool air in is going to do a lot more then any air stream effect is.

I know you're concerned about the heat output, but if you're not getting the hot air out of the case and cool air in, then it's not circulating. If it's not circulating, then it's baking itself with it's own exhaust. Heatsinks can only manage so much heat, if you reach the top of your cooling "budget", then you need to either add a cooler with more surface area or get more/colder stuff for the heat to transfer to (i.e. more cold, dense air).

As far as modifications go, I'll look up the parts I used on mine and get back to you.

Well shit, I'm honestly surprised it took this long for me to realize that the PSU fan isn't an exaust. Sheesh.

But now I'm confused. If it's not the PSU fan, then how am I feeling a clear blast of air from the top of the PC? The heat-sink doesn't blow up and out of the case, so I've always just assumed that the PSU fan what was causing the air flow. I'm guessing it's just the heatsink fan's air bouncing off the inside of the side panel and back out the top. Guess I'll find out when I remove that panel.

Well that's a pretty big fuck up on my part there. But again; it was either that or cause potential damage to parts trying to force wires around where they don't belong. Ah well.

I'm defo going to have to sort that out at some point. It's like I said before; fitting that PSU in was a nightmare sadly.

Well thanks very much for taking the time to do this dude, means alot :]. I'll take off the panel for the time being.

Lemmie know if you're in need of any of my other specs. And remember that I have a mid-tower, so I probably can't take a heatsink that's super large.

 

[nuts32605]

Are you sure the fan on the back of the case is blowing out? You shouldn't have that much back-pressure in the case. Do a quick test with a piece of paper and make sure there's no blockages or dust monsters.

 

[Triburos]

Are you sure the fan on the back of the case is blowing out? You shouldn't have that much back-pressure in the case. Do a quick test with a piece of paper and make sure there's no blockages or dust monsters.

Oh no no, the back fan on the case is working fine, and it is blowing out.

Im pretty sure the gust of air coming from the top of the case is the heatsink's fan blowing against the inside of the panel and then being redirected upward.

Anywho, I actually decided to shut down my PC for awhile, and I took a can of compressed air to it just incase. there wasn't alot of dust, just a small layer of it on a few things. So I think I can rule out dust.

I'll try seeing how it does tomorrow or later on tonight. Heat issues with me seem to be an on and off thing, as annoying as that is.

Also, Shortly before I powered down I noticed my temperatures were rising again quite fast, got to like 70 for both the CPU and MB just on idle for some reason. i'm assuming it was due to a program update or something. Noticed certain updates - especially window updates - will push the processor up a few degrees.

Felt like it was worth mentioning. Even though speccy and HWMonitor arent the most accurate, they can atleast tell me when something is clearly wrong. 70c idle is just wrong.