4000D Airflow Fan Orientation Testing; 1 v 2 CPU Fans. Extra case fans. Temps & Noise comparisons.

[Ralfi]

I got curious, was in ‘experimental’ mode & had fans to spare, so I grabbed one of the shelved fans that came with my Corsair 4000D Airflow Case & clipped it onto the pull side of my CPU cooler with the intent of seeing if it made any difference to my temps when used in conjunction with the push fan I’d added months ago.

 

(These will be in addition to 3 x 120mm front intake fans, 1 x 120mm rear exhaust & 1 x 120mm top/rear exhaust. All Arctic Bionix P120 ARGB).

 

Granted, this is a complete patchwork quilt of a CPU cooler as it now has a DeepCool heatsink, an Arctic push fan (secured with cable ties) & a Corsair pull fan, so it’s not ideal, but it’s all I had, so here we go

 

 

Spoiler

 

 

Spoiler

 

 

 

Spoiler

 

 

I picked up a nifty little digital thermometer this week, so was able to record system temps above ambient, & as the ambient temp changed, so did my calcs. Screenshots accompanying results will show the actual temps (in Celsius!)
 

Stress testing for both CPU (5600X, Auto-OC to 4.6GHz) & GPU (6700XT, OC to 2724MHz, under-volted to 1100mV) was undertaken using FurMark, at the same time.

 

Noise was measured at seating position (~1.2m from PC), using IOS App “dB Meter”.

 

I think I’ve covered everything. If not, let me know - I enjoy doing stuff like this now & then, so would appreciate any feedback

 

Onto the results️…
 

PC Thermals & Noise - 1 v 2 CPU Fans

 

1 CPU Fan

 

IDLE CPU Temp over Ambient: +14°

IDLE GPU Temp over Ambient: +12.9°

IDLE GPU (Hot Spot) Temp over Ambient: +13.9°

IDLE Noise: 27dB…

 

Spoiler

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +46.7°

LOAD GPU Temp over Ambient: +54.4°

LOAD GPU (Hot Spot) Temp over Ambient: +65.4°

LOAD Noise: 36dB…

 

Spoiler

 

 

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +46.5°

LOAD GPU Temp over Ambient: +53.9°

LOAD GPU (Hot Spot) Temp over Ambient: +65.9°

LOAD Noise: 36dB…

 

Spoiler

 

 

2 CPU Fans

 

IDLE CPU Temp over Ambient: +11°

IDLE GPU Temp over Ambient: +11.5°

IDLE GPU (Hot Spot) Temp over Ambient: +13.5°

IDLE Noise: 27dB…

 

Spoiler

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +47.8°

LOAD GPU Temp over Ambient: +55.9°

LOAD GPU (Hot Spot) Temp over Ambient: +67.9°

LOAD Noise: 36dB…

 

Spoiler

 

 

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +47.7°

LOAD GPU Temp over Ambient: +55.7°

LOAD GPU (Hot Spot) Temp over Ambient: +65.7°

LOAD Noise: 36dB…

 

Spoiler

 

So to conclude, at least for my setup, it’s not worth adding a pull fan to the CPU cooler. The noise may be the same, but the performance isn’t improved. Temps above ambient are actually higher with two fans surprisingly & this was despite the 2-fan tests being done before the 1-fan tests, when the ambient temps were slightly lower.

 

This does reflect many similar tests on YouTube - they suggest it’s not worth it either, at least on low-medium CPU coolers.

 

I realise it’s common knowledge for many here, but hopefully it helps those with similar setups, who were wondering what the deal was with 1 v 2 CPU fans.

[Somerandomtechyboi]

Maybe a push pull fan config would be worth it if you had a thiccer cooler or you set your fans to a really low rpm, but on a thin tower like that ofc theres basically no gain

 

At one point i was running dual fans on my crappy chinese tower but that doesnt seem to help at all, i still doubt it would help much considering i run a server fan that goes mid 3k rpms, even when the heatpipes were blazing hot while running 4g on my x5660 even shoving a fan pointed at the heatpipes from the side didnt help at all, better off using the spare 92mm fan included with the tower to cool the nb and running water through the heatpipes w a ghetto waterloop

[freeagent]

Those little pipes can only do so much, you need some more girth

[thrasher_565]

is that only a 2 heat pipe cooler? well you can get 6-8 pipe ones... i think gn said something about this but i dont no. its normally not worth the cost of an extra fan vs the performance you get.

[freeagent]

4 hours ago, thrasher_565 said:

is that only a 2 heat pipe cooler? well you can get 6-8 pipe ones... i think gn said something about this but i dont no. its normally not worth the cost of an extra fan vs the performance you get.

It looks like 4 pipes, with the knockout ready to be opened up for the 5th for the model above 

 

 

[Ralfi]

7 hours ago, freeagent said:

Those little pipes can only do so much, you need some more girth

They’re 6mm diameter. Would that be below average?

https://www.deepcool.com/product/cpucooler/2019-11/7_12723.shtml

 

5 hours ago, thrasher_565 said:

is that only a 2 heat pipe cooler? well you can get 6-8 pipe ones... i think gn said something about this but i dont no. its normally not worth the cost of an extra fan vs the performance you get.

Nah, 4 pipes.
 

I might have to check that Gamers Nexus vid again to see if they used coolers with the same pipe thickness or thicker.

[freeagent]

6 minutes ago, Ralfi said:

They’re 6mm diameter. Would that be below average?

No not at all..

 

8mm pipes is where its at 

 

Don't listen to me I am sure its fine 

[Ralfi]

3 minutes ago, freeagent said:

Don't listen to me I am sure its fine

Oh, I’m happy enough with the performance. Just curious as I hadn’t thought about that aspect of things. Thanks

[thrasher_565]

3 hours ago, Ralfi said:

Oh, I’m happy enough with the performance. Just curious as I hadn’t thought about that aspect of things. Thanks

ya we are just nit picking at this point haha talking smak

[thrasher_565]

4 hours ago, Ralfi said:

They’re 6mm diameter. Would that be below average?

https://www.deepcool.com/product/cpucooler/2019-11/7_12723.shtml

 

Nah, 4 pipes.
 

I might have to check that Gamers Nexus vid again to see if they used coolers with the same pipe thickness or thicker.

well 6mm will allow for probably 8 pipes i would think but this is a "cheap" cooler

[Ralfi]

More results from the lab, this time with the addition of a 2nd top fan as intake + the removal of the 2nd CPU fan…

 

2 CPU Fans + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +11.2°

IDLE GPU Temp over Ambient: +10.8°

IDLE GPU (Hot Spot) Temp over Ambient: +13.8°

IDLE Noise: 27dB

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +45.7°

LOAD GPU Temp over Ambient: +53.8°

LOAD GPU (Hot Spot) Temp over Ambient: +66.8°

LOAD Noise: 37dB

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +45.5°

LOAD GPU Temp over Ambient: +54°

LOAD GPU (Hot Spot) Temp over Ambient: +65°

LOAD Noise: 37dB

 

1 CPU Fan (2nd CPU Fan turned off) + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +11.5°

IDLE GPU Temp over Ambient: +11.3°

IDLE GPU (Hot Spot) Temp over Ambient: +13.3°

IDLE Noise: 27dB

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +45.8°

LOAD GPU Temp over Ambient: +53.2°

LOAD GPU (Hot Spot) Temp over Ambient: +64.2°

LOAD Noise: 36dB

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +45.5°

LOAD GPU Temp over Ambient: +52.1°

LOAD GPU (Hot Spot) Temp over Ambient: +64.1°

LOAD Noise: 36dB

 

1 CPU Fan (2nd CPU Fan removed) + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +12.4°

IDLE GPU Temp over Ambient: +11.9°

IDLE GPU (Hot Spot) Temp over Ambient: +13.9°

IDLE Noise: 27dB

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +44.1°

LOAD GPU Temp over Ambient: +52.4°

LOAD GPU (Hot Spot) Temp over Ambient: +65.4°

LOAD Noise: 37dB

 

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +44.1° (Best)

LOAD GPU Temp over Ambient: +51.8° (Best)

LOAD GPU (Hot Spot) Temp over Ambient: +61.8°(Best)

LOAD Noise: 37dB

 

So, to conclude, this is the best fan setup for my particular case…

 

[thrasher_565]

15 hours ago, Ralfi said:

More results from the lab, this time with the addition of a 2nd top fan as intake + the removal of the 2nd CPU fan…

 

2 CPU Fans + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +11.2°

IDLE GPU Temp over Ambient: +10.8°

IDLE GPU (Hot Spot) Temp over Ambient: +13.8°

IDLE Noise: 27dB

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +45.7°

LOAD GPU Temp over Ambient: +53.8°

LOAD GPU (Hot Spot) Temp over Ambient: +66.8°

LOAD Noise: 37dB

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +45.5°

LOAD GPU Temp over Ambient: +54°

LOAD GPU (Hot Spot) Temp over Ambient: +65°

LOAD Noise: 37dB

 

1 CPU Fan (2nd CPU Fan turned off) + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +11.5°

IDLE GPU Temp over Ambient: +11.3°

IDLE GPU (Hot Spot) Temp over Ambient: +13.3°

IDLE Noise: 27dB

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +45.8°

LOAD GPU Temp over Ambient: +53.2°

LOAD GPU (Hot Spot) Temp over Ambient: +64.2°

LOAD Noise: 36dB

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +45.5°

LOAD GPU Temp over Ambient: +52.1°

LOAD GPU (Hot Spot) Temp over Ambient: +64.1°

LOAD Noise: 36dB

 

1 CPU Fan (2nd CPU Fan removed) + 1 Top Intake Fan

 

IDLE CPU Temp over Ambient: +12.4°

IDLE GPU Temp over Ambient: +11.9°

IDLE GPU (Hot Spot) Temp over Ambient: +13.9°

IDLE Noise: 27dB

 

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +44.1°

LOAD GPU Temp over Ambient: +52.4°

LOAD GPU (Hot Spot) Temp over Ambient: +65.4°

LOAD Noise: 37dB

 

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +44.1° (Best)

LOAD GPU Temp over Ambient: +51.8° (Best)

LOAD GPU (Hot Spot) Temp over Ambient: +61.8°(Best)

LOAD Noise: 37dB

 

So, to conclude, this is the best fan setup for my particular case…

 

makes sens

[Allyxxx]

I thought for certain the 2 CPU fan setup would be slightly better or about the same. Considering that the rear and front fans on the heat sink are different, it is possible the performance could have been reduce.

 

15 hours ago, Ralfi said:

So, to conclude, this is the best fan setup for my particular case…

 

The top fan on the front looks like it is exhausting fresh air before it gets to the CPU.

[freeagent]

Atta boy, its good to see people playing with their setup to see what works best, instead of taking the cookie cutter approach

[Ralfi]

35 minutes ago, Allyxxx said:

The top fan on the front looks like it is exhausting fresh air before it gets to the CPU.

Yeah I know it looks that way, but what you’re seeing of that fan is the frame - the frame design of that particular Corsair fan looks like fan blades. I can assure you that it’s an intake .

 

I was actually going to do another test with it flipped to be an exhaust, mainly to see how it impacted GPU temps, but as you mentioned, it’ll most likely make things noticeably worse for the CPU, so I’ll keep things as is for now, as I’m happy that I now have the best fan orientation for overall cooling performance.

 

36 minutes ago, Allyxxx said:

I thought for certain the 2 CPU fan setup would be slightly better or about the same. Considering that the rear and front fans on the heat sink are different, it is possible the performance could have been reduce.

Yeah I think if I had a proper 2-fan cooler, the results would've favoured 2 'in-sync' fans. Despite me trying to match RPMs between the Push/Pull fans with the fan curves, it still didn't work. & when I turned off the Pull fan for the 1 CPU Fan testing, the airflow was blocked by the still blades of the Pull fan, so I didn't hesitate to remove the Pull Fan completely for the remaining tests & wasn't surprised with the improved results.

[thrasher_565]

6 minutes ago, Ralfi said:

Yeah I know it looks that way, but what you’re seeing of that fan is the frame - the frame design of that particular Corsair fan looks like fan blades. I can assure you that it’s an intake .

 

I was actually going to do another test with it flipped to be an exhaust, mainly to see how it impacted GPU temps, but as you mentioned, it’ll most likely make things noticeably worse for the CPU, so I’ll keep things as is for now, as I’m happy that I now have the best fan orientation for overall cooling performance.

ya nice to see these test. there are other temps to consider too like vrm.

 i was testing with me lian li 011 air as i wanted to no if side fans vs front fans but i got lazzy haha. problem with that case is the back only can do 2 80 mm fans so its would be hard to put that test up[ ageist a case with a 120mm fan. maybe ill look in the an 011 mini evo  air or something. 

 

but case with bottom take seem to be better fro gpu temps as it help to push the gpu exhaust up at the back fan instead of it being sucked back in the gpu.

[Ralfi]

16 minutes ago, thrasher_565 said:

problem with that case is the back only can do 2 80 mm fans

Yeah, I sometimes feel cases need more than 1 x 120mm fans as standard, to help GPU exhaust fumes. Normally the rear fan's there to help CPU more so. But with the PCI slots where they are, it makes it hard.

16 minutes ago, thrasher_565 said:

but case with bottom take seem to be better fro gpu temps as it help to push the gpu exhaust up at the back fan instead of it being sucked back in the gpu.

Cases with bottom intakes are great to help partially solve what I mention above (lack of rear exhaust focussed on dispensing GPU air). I've been trying to figure out how I could do it in my case, by cable-tying a fan to the PSU shroud, but I don't have enough clear room beneath the shroud that lines up with the rear GPU fan, & the front GPU fan has my 'feature' GPU power cables in the way  . Maybe a fan bracket fixed to the 2nd or 3rd PCI slot from the bottom could hold one or two fans directly beneath the GPU & pick up the cool air coming in from the front to push into the GPU, but it may not be much of a benefit for the effort...

 

What i've liked seeing by others are experiments with GPU shrouds - they seem to work very well. It'd be good if Corsair made them for their cases. I'd grab one for sure.

 

16 minutes ago, thrasher_565 said:

ya nice to see these test. there are other temps to consider too like vrm.

 

Thanks man. Yeah I wanted to record all Mobo temps, so added the Gigabyte Hardware Monitor App in my shots. Looks like the VRM got to 27 degress above ambient in that last test. I'm not sure how that fares, as i've not looked into VRM closely. I think the rest of the motherboard temps appear ok right?

 

45 minutes ago, freeagent said:

Atta boy, its good to see people playing with their setup to see what works best, instead of taking the cookie cutter approach

Thanks bud. Good to know others appreciate it

[Allyxxx]

6 minutes ago, thrasher_565 said:

ya nice to see these test. there are other temps to consider too like vrm.

VRM seem to get left out, but varies so much from one motherboard to another. I try find ways to replace 120mm fans with 140mm fans to blow air under the heatsink and pull it and hope some cool air gets push under the VRM.

 

Having the top fan pushing some fresh air on the memory and VRM can help, but may need some moding to offset the mounting to directing more effectively.

[thrasher_565]

10 minutes ago, Ralfi said:

Yeah, I sometimes feel cases need more than 1 x 120mm fans as standard, to help GPU exhaust fumes. Normally the rear fan's there to help CPU more so. But with the PCI slots where they are, it makes it hard.

Cases with bottom intakes are great to help partially solve what I mention above (lack of rear exhaust focussed on dispensing GPU air). I've been trying to figure out how I could do it in my case, by cable-tying a fan to the PSU shroud, but I don't have enough clear room beneath the shroud that lines up with the rear GPU fan, & the front GPU fan has my 'feature' GPU power cables in the way  . Maybe a fan bracket fixed to the 2nd or 3rd PCI slot from the bottom could hold one or two fans directly beneath the GPU & pick up the cool air coming in from the front to push into the GPU, but it may not be much of a benefit for the effort...

 

What i've liked seeing by others are experiments with GPU shrouds - they seem to work very well. It'd be good if Corsair made them for their cases. I'd grab one for sure.

 

Thanks man. Yeah I wanted to record all Mobo temps, so added the Gigabyte Hardware Monitor App in my shots. Looks like the VRM got to 27 degress above ambient in that last test. I'm not sure how that fares, as i've not looked into VRM closely. I think the rest of the motherboard temps appear ok right?

 

[Allyxxx]

I have seen that Major Hardware video and many others by that guy. That's how got interested in trying to make some custom mounting solutions for fans and cases.

[Ralfi]

4 hours ago, Allyxxx said:

I have seen that Major Hardware video and many others by that guy. That's how got interested in trying to make some custom mounting solutions for fans and cases.

Yeah that's the main video I was thinking of. Very practical thinking.

 

How did your custom solutions turn out?

 

edit...Ooh I just found an old 80mm Noctua NF-R8-1800 fan which just fits under the GPU. It even magnetises to the outer side of the PCI slot cover strips, so I may do some brainstorming to get more cool air into the GPU...

 

Here’s where it stuck to the outside of the case, so I thought I’d mount it there initially…

 

 

Airflow was a bit restricted by the PCI slot covers, so I then moved it to the inside…

 

 

[Ralfi]

^ Installing the Noctua didn’t improve things though, so it’s coming out. Kind of glad because it didn’t suit the theme of the case anyway & it put a dent into my cable management  

 

1 CPU Fan + 1 Top Intake Fan + 1 Rear Intake Fan

 

IDLE CPU Temp over Ambient: +12.8°

IDLE GPU Temp over Ambient: +14.7°

IDLE GPU (Hot Spot) Temp over Ambient: +16.7°

IDLE Noise: 28dB

FurMark Stress Test, 30 minutes...

LOAD CPU Temp over Ambient: +45.1°

LOAD GPU Temp over Ambient: +53.7°

LOAD GPU (Hot Spot) Temp over Ambient: +64.7°

LOAD Noise: 37dB

FurMark Stress Test, 60 minutes...

LOAD CPU Temp over Ambient: +45.3°

LOAD GPU Temp over Ambient: +52.8°

LOAD GPU (Hot Spot) Temp over Ambient: +63.8°

LOAD Noise: 37dB

[fonzz1e]

afaik usually cpu cooler with 2 fans, they are using the same model fan with high static pressure

example, noctua nf-u12a using two nf-a12x25

[thrasher_565]

gn was talking a bit about cpu coolers and how thick they are, fin densaty, and its probly not a simple as adding another fan. what fan you add is also imported.

 

the 80mm fan on the bottom was there for when gpus were hitting 90c. but  there a few gtx 3090s that are lemons well others are fine so i still dont really no. i thought it was the case but now i dont no as others have had 3090s and not had temps problem and used worse air flow cases. there were talks about vrm and thermal pads being a problem so maybe that it?

[jones177]

On 3/17/2022 at 11:36 PM, Ralfi said:

 

 

edit...Ooh I just found an old 80mm Noctua NF-R8-1800 fan which just fits under the GPU. It even magnetises to the outer side of the PCI slot cover strips, so I may do some brainstorming to get more cool air into the GPU...

 

Airflow was a bit restricted by the PCI slot covers, so I then moved it to the inside…

I use 60mm Noctua fans below the GPUs on my cases that have PSUs on the bottom. 

I started using them with RTX 3080 ti since they put a lot of heat into the cases.  

 

On my CM H500P Mesh and CM H500 ARGB they are good for 3c on the GPU but only 2c on my Corsair 5000d. 

 

My main gaming PCs now use Lian Li o11 Dynamics that have all the intakes below the GPU. They fix the issue of hot air getting trapped below the GPU but make the air that the CPU cooler intakes hotter. Not a big issue with a 360mm AIO but could be with less.