Jump to content
Corsair Community

600C Inverted case GPU temps high!?


AladarZC
 Share

Recommended Posts

System:

600C Inverted Case

Asus Prime X370 Pro

Ryzen1700 w/ Corsair AiO Cooler

Corsair 16GB 3000mhz DDR4

Corsair 750M Bronze PS

RX580 XFX 8GB GPU

 

 

When I play games (Mostly Overwatch) my GPU temps rise so much, and when I hit around 80 degree mark my GPU crashes. (Rendering Device has been lost). I feel as if my cooling in the case isnt good enough? Reviews I've seen before getting the case say that temps are not a real issue with it, so I went ahead and got it. Now my GPU screams everytime I play.

 

Should I be looking at getting new fans? New fan setup? I've tried a multitude of different AMD drivers as well. I know some people will say 80 degrees isnt bad, and it can handle it but when i see other peoples numbers or streamers or whoever, they are much lower temps than mine. What gives?

 

What should I be looking into? I tried getting MSI afterburner but I am clueless on using it. :(

 

Any help would be appreciated. Thanks

Link to comment
Share on other sites

I'm assuming that you haven't overclocked your GPU?

Can you verify that the fans on the card are spinning when it heats up? Afterburner should tell you that, as would any number of other monitoring software applications (including Link).

What is your ambient/room temperature?

Can you validate any of the temperatures, besides CPU, that your motherboard is reporting?

And what variable are your case fans using to ramp up? Since you have liquid cooling, the default variable (CPU temp) isn't relevant for controlling your case fans. You'll want to control it by one of your motherboard temperatures that provides some indication of the case interior temperature. And since that won't get as hot as a CPU would get, you'd also want the fans to ramp up at lower temperatures.

Link to comment
Share on other sites

That's kind of the tricky part. The difference between a decent airflow case and a lackluster performer might be a 5C on the ambient temp. So absent some other restriction, that means you would be at 75C instead of 80C. When at the limit, that might be a little helpful, but I suspect you think it should be lower that that anyway. That likely means there is another issue.

 

Ideally, you would use temp probes to take reading from around the outside of the GPU and compare those to the rest of the case, etc. Most people won't have the gear for that. However, you do a fairly primitive assessment to see what direction to look next. If the case is not getting GPU waste heat out of that top area, it is going to be pretty hot to the touch on the case panels and interior. If it's not overly warm, that would suggest the problem is with the card and you can try the things suggested above. If the case is screaming hot, than we need to figure out why. I was a bit curious as to how the top mounted GPU would do in close proximity to the PSU above when under long gaming loads. It was a caution sign for me, but you are the first person I can remember since launch to specifically address this. There may be unique reasons why your's is hot, but I am not sure the design is fully responsible or this would be a weekly post.

 

Are you able to get temp readings on a 750M PSU in Link? Does it feel overly hot? It is possible this plays a part in all of it as well.

Edited by c-attack
Link to comment
Share on other sites

Alright so I downloaded HWmonitor to see if my case fans are doing anything, and these are what I get:

 

Fans:

CPU - 1000-1100

FANIN1 - 800-900

FANIN2 - 800-900

FANIN4 - 4200-4400

 

all Fans in PWM are 0%

 

I opened the case while playing, and i noticed the case fans dont ramp up. And in HWmonitor they still say 800ish. I am also wondering what is running at 4000RPM as i dont see anything screaming like that in the case. Maybe its reading the pump?

 

I am in a room thats about 65 degrees since its pretty chilly outside.

 

I havent been able to find out my PSU fan rpms, but i put my hand in the back of the case where the PSU is, and i feel like its not ramping up either. I BARELY feel anything. Even under game load, the breeze from the back doesnt kick up.

 

The gpu is very hit when i puit my hand around it, or even touch the bottom side. While typing this, nothing else running except HWmonitor GPU is sitting at 50 degrees 900rpm fans. Its factory OC'd i believe its the XFX GTS 8GB fancy dancy one. the mV on MSI afterburner remain at 0, is that normal?

Link to comment
Share on other sites

Question: when you opened the case while playing, did the GPU stay any cooler?

 

Your profile indicates an AIO cooler ... where is that plugged in? Typically it's the CPU_FAN (per instructions) but it doesn't have to be. That should be set to 100% or Manual. The pump will report RPMs that are about 1/2 the pump RPM.

From there, where are the rest of the fans plugged in? You can stop the fans with your fingers to see which one is which on HWMonitor/Link. That motherboard has at least 1 on-board temperature sensor in addition to the CPU and PCH sensor. What does this temp sensor read when you are gaming, especially with the case closed?

Link to comment
Share on other sites

OK. Some suggestions:

 

- The Radiator fans should be connected to the pump, not the CPU fan header. This will allow the radiator fans to be controlled by the pump temperature, which is the relevant thing that they cool. By putting them on the motherboard, you can't do this ... plus, they will ramp up and down frequently based on CPU speed, which is something less than pleasant. The default fan curves in Link are based on controlling fan speed by the AIO coolant temperature, not CPU temperature.

- Typically, the AIO should be on the CPU_FAN header, this gives the CPU Fan warning if the pump isn't working. You would need to make sure that it's set to 100%/Manual mode (in the ASUS BIOS, it's DISABLE Fan Control). BUT ... since CPU_OPT mirrors the CPU_FAN header and you are a little short on fan headers, just leave it there. Instead, put your intake fans on CPU_FAN and CPU_OPT; this allows them to be controlled together. In the BIOS, set this CPU_FAN to be controlled by your motherboard temperature (you should be able to do this) ... according to the Asus site, your board has a temperature sensor between PCIEX16_2 and PCIEX16_3. With the inverted case, this is at the top and probably the temperature that you want to watch most. Start with having the fans max out at, say, 45C and start slowly ramping up at about 30-35C.

- Flip the fans for the radiator to have them set to exhaust mode. You need to be actively pulling warm air DOWN and out the bottom and you have physics working against you. If you look at the diagrams on Corsair's site for this case, that's what pretty clearly suggested for a bottom-mount radiator. What may be happening is that there's no good way for the hot air from your video card to get exhausted so it "collects" up at the top. This prevents your GPU from cooling effectively; it can only be as cool as the ambient air around the GPU. Plus, if your fans aren't spinning up as you increase load, you just aren't moving enough air around to effectively cool the GPU.

- Put your exhaust fan on CHA_FAN1 or CHA_FAN2 and set it like the CPU_FAN. Don't have it ramp up quite as fast, though ... perhaps 100% at 55C. You want to try to maintain positive pressure in the case - more air coming in than going out. With the radiator exhausting air, it'll be easy to have negative pressure unless you are pretty conservative about the rear fan exhaust.

 

Note that the temps for the fan curves are just starting recommendations. Tweak that based on the results as well as the desired noise profile (faster fans are noisier).

Edited by DevBiker
Link to comment
Share on other sites

Actually, I just looked at the video for that case and noticed that you have a fan controller on the case. Alternatively to my babbling above, you can use that for your intake fans and then manually increase fan speed when you game. IF you do that, put your pump on CPU_FAN and set it to 100%. :-)
Link to comment
Share on other sites

I am wondering if your AIO is a Corsair H60. That would account for the 4000 rpm reading. If not, please specify. It doesn't matter if it's not Corsair.

 

You likely do not want anything on CPU OPT on an Asus board. The OPT header is a mirror of CPU fan and cannot be controlled individually. It is useful for two fan air towers or two radiator fans. Perhaps the AMD series is different, but probably not.

 

If you have a case exterior fan controller, that would perhaps be the easiest for you to manage. However, your current fan speeds are not minimal and when it comes to GPU VRM or diode temp, only the GPU's owns fans can really make a difference. If you want to see 65C on the GPU, you would need to lower the case ambient temperature by 15C with your case fans. That's not going to happen at any speed and I am not sure you have 15C to drop.

 

The two concerns are:

 

1) Too much heat in that top corner. If so, your GPU and PSU should be really hot. Are the exterior case panels hot or very warm to the touch around there? (The GPU will always be hot. Don't touch.) The PSU may or may not be helpful with moving air out of that area and it likely changes its fan speed on a low/high toggle based on PSU temperature. I wouild only be concerned if your PSU exhaust feels extremely hot. It likely would be in the 40-50C range when gaming. If it is hotter than that, the exhaust will be very uncomfortable to put your hand near. The PSU and GPU fans are going in opposite directions. Most of the GPU heat should go out the back of the card or radiate off the bottom of the inverted back plate, thus the underside. If local temp and case flow really is the issue, it would build up above the card and influence the PSU as well.

 

2) I think its time to get GPU model specific. I don't know your card, but ultimately its nature and fan speed has more influence than all of this other stuff combined. You may need to find some comparative data specifically for the RX580.

Edited by c-attack
usual typos
Link to comment
Share on other sites

OK ... so with that in mind, allow me to revise my previous.

Assuming that you aren't using the fan controller built into the case ...

Keep the pump and the fans right where they are. That's the best place for them. However, do reverse the cooler to exhaust mode, not intake. I think that's exacerbating the problem; as c-attack pointed out, you have a "heat collection zone" at the upper right of your case (GPU and PSU, especially the GPU) and you need to get that hot air moving and exhausted out of the case. Dumping the heat from your radiator back into the case with so little exhaust and so little intake is turning the inside of your case into a sauna. It won't matter if your fan is push or pull (although pull does make it easier to clean off the radiator fins from dust) just as long as it's exhaust and out the bottom ... preferably, I think, towards the back.

You do want to set the fan curves for the intake and exhaust fans on that motherboard temperature sensor though. That's going to be close to that "heat collection zone" and it should let you know when it starts heating up. Now, which sensor corresponds to the one on the Asus diagram ... I can't say but you don't have many to choose from ... 1 is under the CPU, 1 is on the chipset and the other one is that motherboard sensor between the chipset and PCIEX16_3. Part of the problem is, I think, that your fan curves are based on CPU temperature (which is the default). That works for an air-cooled system but it's not really relevant for a liquid-cooled system. Now, since you're going to be setting them based on an ambient air temperature, you'll want to adjust the high temp of your curve down accordingly (probably between 40 and 50C, at most). If you have AI Suite running, you can do it in there. I, personally, found AI Suite to be a PITA but YMMV. If you don't have AI Suite running, you'll need to do it in the BIOS.

BTW ... the 4000 RPM fan header is the pump; there's nothing else it can be. And there's really nothing special about that AIO_PUMP header; it's just another PWM fan header that's defaulted to full speed. But having the other fans spinning at only around 800-900 RPM when you are gaming and creating a bunch of heat in the system from the GPU isn't helping and I do think that's directly related to your GPU overheating.

Link to comment
Share on other sites

 Share

×
×
  • Create New...