Optimizing Airflow in Meshify C

[Anon7mous]

My system:
Motherboard: Asus TUF x570 plus with Wi Fi 
CPU: Ryzen 3700X

CPU Cooler: Dark Rock Pro 4
GPU: 2070 Super Zotac AMD! Extreme

Fans: 2x intake Noctua 140mm redux 1500 RPM PWM fans, 1x exhaust at the back 120mm Fractal Design, 1x exhaust at the top 120mm Fractal Design (the ones that come with the case)

I did NOT overclock my GPU or CPU

 

My temps:

Ambient temp: 21-22 C

CPU idle: ca. 45 C

CPU under load: 67-70 C (Furmark CPU Burner)

GPU under load: 80 C (Furmark)

While gaming the temps are a few degrees lower than under full load

 

I believe my temps are a bit too high considering what case / fans and CPU cooler I have. I believe I messed up when applying the thermal paste and I didn't adjust any fan speeds. I would like somebody advise me how to optimize my airflow to reduce the temps.
 

1. I bought Thermal Grizzly Kryonaut today and I will apply it on the weekend. Should I use the spread method advised by team grizzly or some other?

2. One of the exhaust fans is connected to AIO port on the motherboard, is this a problem or should I connect both exhaust fans to one connector and connect both fans to one PWM port on the mobo?

3. Probably the most important question: How should I adjust the fan speeds in BIOS? Is there an ideal fan curve I should be aiming for? SpeedFan doesn't show case fans, only GPU and CPU so I guess I have to use BIOS

4. Should I change any settings for the CPU or GPU?

 

Thanks!

[Meganter]

That looks more than fine to me  

[Bacon soup]

6 minutes ago, Anon7mous said:

3. Probably the most important question: How should I adjust the fan speeds in BIOS? Is there an ideal fan curve I should be aiming for? SpeedFan doesn't show case fans, only GPU and CPU so I guess I have to use BIOS

 

I think those idle and load temps are normal. My ambient temperature is 26C and my CPU at idle is 41C using a scythe fuma 2 cooler. I have four 140mm intake fans and one 120mm exhaust fan on a cooler master NR400 case.

This is what I set for case intake fans:

30 degrees C, winter & night idle temp, fan speed 0%.

40C, summer day time idle temp, case fan speed 1%

50C, gaming temp, fan speed to 50%.

I set fan speed to 100% at 69C.

CPU fan and exhaust fan are linked on a splitter cable and I left it to default PWM settings.

[Anon7mous]

34 minutes ago, Plouffe said:

That looks more than fine to me  

They are not critical but some people are achieving same temps with stock cooler and only 2 case fans and that's signaling me that something is not optimized. I think my fans curves are not ideal. I never adjusted them before that's why I'm a bit unsure what to do.

[DoctorNick]

Kryonaut. I would just use a dot. Tried spreading it, but didn't get good temps compared to just pea method. Just make sure it spread nicely all over the die. 

If you want more airflow in the case you could remove front dust-filter, but then again.. More dust.

[Anon7mous]

31 minutes ago, Bacon soup said:

I think those idle and load temps are normal. My ambient temperature is 26C and my CPU at idle is 41C using a scythe fuma 2 cooler. I have four 140mm intake fans and one 120mm exhaust fan on a cooler master NR400 case.

This is what I set for case intake fans:

30 degrees C, winter & night idle temp, fan speed 0%.

40C, summer day time idle temp, case fan speed 1%

50C, gaming temp, fan speed to 50%.

I set fan speed to 100% at 69C.

CPU fan and exhaust fan are linked on a splitter cable and I left it to default PWM settings.

I can give it a try and tinker the curves a bit. Btw do you have any idea why can't I see my fans in SpeedFan but only in BIOS?

[Anon7mous]

1 minute ago, DoctorNick said:

Kryonaut. I would just use a dot. Tried spreading it, but didn't get good temps compared to just pea method. Just make sure it spread nicely all over the die. 

If you want more airflow in the case you could remove front dust-filter, but then again.. More dust.

Yeah I'll spread a thin layer of it. 

I'd prefer keeping the dust filter. 2x 140mm 1500RPM fans should draw in enough air but I think they are just running on low RPM because ithey're set on automatic. The fans are so quiet I wouldn't mind bumping the speed a little bit up to further reduce the temps. I'm just not 100% sure how to balance the RPMs between 2x 140mm intake fans and 2x exhaust fans.

[Bacon soup]

6 minutes ago, Anon7mous said:

I can give it a try and tinker the curves a bit. Btw do you have any idea why can't I see my fans in SpeedFan but only in BIOS?

maybe speedfan developers havnt added your bios. I forgot to mention my temps are from ryzen 2600

[WoodenMarker]

4 hours ago, Anon7mous said:

They are not critical but some people are achieving same temps with stock cooler and only 2 case fans

I find that a bit hard to believe. Are they also using Furmark? Any examples you can link? 

[Anon7mous]

52 minutes ago, WoodenMarker said:

I find that a bit hard to believe. Are they also using Furmark? Any examples you can link? 

I don't know if they are using Furmark. There's always a chance they're not telling the truth but here you go:

1. 

2.

 

 

[OldHunter]

44 minutes ago, WoodenMarker said:

I find that a bit hard to believe. Are they also using Furmark? Any examples you can link? 

Is Furmark that much more intense than a blender & heaven loop?

I have a different case but a 3700X with the stock cooler I can get it as low as 64 degrees on a blender & heaven loop, thats with crazy noise tho, 68 degrees with moderate noise, and 72 degrees running quiet.  I agree with the OP i would expect better from a dark rock pro 4.

 

OP, I would crank all the fans to 100% in the BIOS, setup your benchmark, let it get up to temp, and start playing, partially restrict fans one at a time, switch orientations, etc to see what the best case scenario is, from there turn down the fan speed to something tolerable noise wise and see where the temps land.

[Anon7mous]

2 minutes ago, OldHunter said:

Is Furmark that much more intense than a blender & heaven loop?

I have a different case but a 3700X with the stock cooler I can get it as low as 64 degrees on a blender & heaven loop, thats with crazy noise tho, 68 degrees with moderate noise, and 72 degrees running quiet.  I agree with the OP i would expect better from a dark rock pro 4.

 

OP, I would crank all the fans to 100% in the BIOS, setup your benchmark, let it get up to temp, and start playing, partially restrict fans one at a time, switch orientations, etc to see what the best case scenario is, from there turn down the fan speed to something tolerable noise wise and see where the temps land.

When I use furmark I get 64C on my CPU. During gaming it goes over 70. I realized it had to do with my GPU because during CPU stress test the GPU is idle. If both are running, they both generate heat. I just need to get the heat out of the case.

 

SpeedFan doesn't recognize my mobo. Can you by any chance recommend any good software to adjust the fan curves or am I forced to do it in BIOS?

[OldHunter]

10 minutes ago, Anon7mous said:

When I use furmark I get 64C on my CPU. During gaming it goes over 70. I realized it had to do with my GPU because during CPU stress test the GPU is idle. If both are running, they both generate heat. I just need to get the heat out of the case.

 

SpeedFan doesn't recognize my mobo. Can you by any chance recommend any good software to adjust the fan curves or am I forced to do it in BIOS?

I do my fan control in the BIOS.  I have an ASUS x470 prime, I'd imagine the fan control would be the same as your X570 tuf.  I ended up setting all fans to the silent profile, the CPU boosts to almost the same regardless of 68 or 72 degrees and the near silent running is worth way more to me than 25mhz.  When getting the fans dialed in, fan orientation made the biggest difference in temperatures.  I run negative pressure and flow from back to front and that made most of the difference, flow balance between a couple of the fans also impacted things more than I would have thought.  That said with a different case and a different cooler about all I can really comment on is that I think your CPU temps should be better with the dark rock pro.

 

I should have mentioned in the post above, GPU is a gigabyte 2070 super, it sees between 66-70 throughout the blender & heaven benchmarks

[Anon7mous]

1 hour ago, OldHunter said:

I do my fan control in the BIOS.  I have an ASUS x470 prime, I'd imagine the fan control would be the same as your X570 tuf.  I ended up setting all fans to the silent profile, the CPU boosts to almost the same regardless of 68 or 72 degrees and the near silent running is worth way more to me than 25mhz.  When getting the fans dialed in, fan orientation made the biggest difference in temperatures.  I run negative pressure and flow from back to front and that made most of the difference, flow balance between a couple of the fans also impacted things more than I would have thought.  That said with a different case and a different cooler about all I can really comment on is that I think your CPU temps should be better with the dark rock pro.

 

I should have mentioned in the post above, GPU is a gigabyte 2070 super, it sees between 66-70 throughout the blender & heaven benchmarks

I think I messed up with the thermal paste. Today I bought Grizzly kryonaut thermal paste and I will reapply the paste on the weekend and play with the fans a bit. I read here that somebody tried out different fan setups for this case and ended up having exactly what I have. Gamers nexus came to a similar conclusion. It was positive pressure, 2x iintake at the front, 1x exhaust at the back and 1x exhaust at the top.

[WoodenMarker]

On 11/7/2019 at 9:15 AM, OldHunter said:

Is Furmark that much more intense than a blender & heaven loop?

Yes, probably. I haven't used FurMark in a while since it's really heavy and can kill some graphics cards.

You could just try Blender and Heaven on your system to compare. 

 

If you're going to compare temps, at least use the same stress test. Here's an example of how the temps can differ: https://www.passmark.com/reports/cpu-heat-generation-benchmarks.php