Case airflow problems at low RPMs with custom loop

[Raxxath]

I recently noticed after a long gaming session that my case was very warm to the touch. Checked the temperatures and the liquid temp was up at 43c, which is warmer than I would like (ambient was around 25c).

 

Did some testing (in 22c ambient) and found that it was maxing out around 40c, with fans in the 900-950 rpm range, which is actually louder than I would like to be be running them at.

 

Removing the side panel of the case dropped the liquid temperature all the way down to 33c, a 7c difference, without changing the fan speeds. Clearly a lot of heat is getting trapped inside of the case for some reason.

 

I tested increasing the fan speeds to max (1250 on 3 front intakes, 1250 on 2 top exhaust, 1450 on 1 rear exhaust) with the side panel back on, and it dropped to 35c. Better, but way too loud for me.

 

So why is my case airflow such a problem? This is the same case layout and fan layout that countless custom loops are built with.

 

-Front radiator is a Nemesis GTS 360, top is a Nemesis GTS 240, optimized for low RPM fans.

 

(Picture is old and has the top fans as intakes. They are definitely exhausts now)

[Skiiwee29]

What case is that? Looks like a Define S or Define R5 at a guess. You are at about the worst that I see recommended of 20c above ambient which is where you are at. Seems that at the lower RPM levels, your radiators are restrictive enough to not allow proper in take which that is your only intake off the front rad. You will need to sacrifice some noise to improve your water temps. But at 43c water temp, you are still 100% fine with no issues to your thermals on components. I have a similar setup in a different case, with a 280 on top and 240 front. I have intake front and Rear and exhaust only top. Try turning the rear fan around and running it as intake. If you decide to try this, make sure you get a dust filter. Once the water reaches 50c is when you start to have problems, most notably with hoses becoming compromised and starting to soften up. 

[Raxxath]

Forgot to mention it's a Define C.

 

Summer temperatures are coming and I prefer to run the fans at lower speeds, which is why I'm concerned.

 

I might not be able to test reversing the rear fan without taking the whole thing apart. Working with this much hardware in such a compact case is a real pain, actually. But it's an interesting thought, and would certainly help with the top radiator's cooling. I wonder how the lower half of the case would fare, though.

[kevinisbeast707]

Lol if you thought working in that was a pain, do the same thing in a SIlverstone rvz-01. But turning the fan around is a good idea. Also what fans do you have as there may be fans that are better optimized for your specific setup.

[Dschijn]

Try front fans intake and top intake as well. Only the back fan will exhaust. Worked very good in a Dark Base Pro 900.

 

Seems like your front rad is feeding warm air to the top rad and make it perform very badly. With ~1000rpmand a 360+240 you should get water temps at around ~9-10K over ambient.

[Raxxath]

Looks like Jay had a very similar issue in a different case, though his was more about poor case design:

 

 

 

On 6/10/2018 at 6:42 AM, Dschijn said:

Try front fans intake and top intake as well. Only the back fan will exhaust. Worked very good in a Dark Base Pro 900.

 

Seems like your front rad is feeding warm air to the top rad and make it perform very badly. With ~1000rpmand a 360+240 you should get water temps at around ~9-10K over ambient.

 

I actually had it setup with the top fans as intakes when I first built it, but switched it when I saw literally everyone else using front intakes + top exhaust. I didn't do any extended testing sessions then, so I'll have to switch it back and do more temperature testing.

 

I'm also thinking that I may just need a different case. Looking at a PC-O11 Dynamic, I could have three bottom intake fans, with a side radiator and top radiator both exhausting. That way both radiators should be getting fed near ambient air, and I don't see how any hot air could get trapped inside the case with that setup.

[For Science!]

2 hours ago, Raxxath said:

-

I also have had a loop in Define C / Meshify C and agree that in this very small case it is easy to re-heat the internals with the "intake" past a radiator. I have a 360 GTS and 240 GTS on the front and top respectively.

 

Prior to my annual maitenance this year, my Define C build during extended loads would reach a fluid temperature 44 degrees celcius and fan speeds of about 1000 rpm. Similar to your situation, the side panel and the rest of the case internals get very hot which in my case was more severe since i didn't have a very good (any) exhaust system. But anyway, given the crammed nature of my SLI loop in the Define C, I reasoned that any amount of switcheroo wasn't going to help.

 

During the annual maintenance, I flipped all fans of my radiator to exhaust instead and to supplement the "severe" negative pressure, put my unobstructed rear fan as an intake (with a dust filter, of course). This now means that my loop can maintain a fluid temperature of 37 degrees while the fans run at about 750 rpm. Since the fans are very low rpm, even after a few months now, the dust issue is a non-problem. All in all, a reduction of fluid temps by 7 degrees in Realbench while at a lower fan RPM would seem to be worth it in my eyes and system.

 

I think given the very compact nature of the C-series, I would promote you to consider an all-exhaust on the radiators with a filtered rear-intake. This reduced my component temps substantially too.

 

[Raxxath]

7 minutes ago, For Science! said:

I also have had a loop in Define C / Meshify C and agree that in this very small case it is easy to re-heat the internals with the "intake" past a radiator. I have a 360 GTS and 240 GTS on the front and top respectively.

 

It's funny, because you would think that in a small case, hot air can pass through it easier and be exhausted faster. But in practice it doesn't seem to work that way.

 

Your experience swapping your fans around makes me think that my idea with the PC-O11 Dynamic having three fans feeding the radiators cool air to exhaust would work really well then.

[Mick Naughty]

Just seems like how it should be to me. Can’t have high fpi rads and low fan speed. Even worse when using a rad as an intake. I’d turn the fan speed up and deal with it. Don’t know how you game and can still hear it. 

[Raxxath]

I switched the top fans back to intakes and it helped by a couple degrees, but I'm still not satisfied. I'd like to try all the radiator fans as exhaust, but that would require taking the whole thing apart, due to the lowest radiator fan being inaccessible. Might just have to go with a different case.

 

Trying to balance noise, price, looks, size, and performance all at the same time really is difficult.

[bling]

I guess you have to try all the combination and find the best one for that case.

 

For what is worth I tried front and back as intake and top as exhaust, but the result was similar to your situation because the hot air couldn't get out fast enough.

 

Maybe try to run the outtake fan/fans at a higher speed and slow down the intake ones, see if it changes anything. This may help exhaust the hot air inside the case.

[Raxxath]

After more testing, I have some interesting conclusions.

 

Having all front and top radiator fans as intakes with just the rear fan as exhaust resulted in about a 16C delta between ambient temperature and load water temperature.

 

Switching the front and top radiator fans to exhaust and using the rear fan as an intake resulted in a 10-11C delta between ambient temperature and load water temperature.

 

Much better. And also the case itself didn't heat up like crazy, like it was before. This setup was also slightly ghetto-rigged, and yet still performed much better than all the others. I couldn't access some of the fan screws on the lowest radiator fan, so I disconnected it and installed a fan between the front panel and the radiator. I'm sure it's suffocated in there, so I would expect slightly better performance when it's mounted properly.

 

I'm surprised all this isn't more common and available knowledge. It probably matters a lot more for us who use lower fan speeds, but you'd think that would encompass a lot of water coolers.

 

Like I said before, I think I'm going to switch to a PC-O11 Dynamic, and have 3 intake fans at the bottom with exhausting radiators on the side and top. I would expect superb performance from that setup based on my testing so far.

 

On 6/19/2018 at 5:28 PM, Mick Naughty said:

Just seems like how it should be to me. Can’t have high fpi rads and low fan speed.

 

Nemesis GTS are optimized for low speed fans.

[Mick Naughty]

5 minutes ago, Raxxath said:

Nemesis GTS are optimized for low speed fans.

As any company would claim. Even more so in a specific use case scenario as tests would show. But then again 16 is barely above average. Shouldn't be having an issue with such fancy rads.

[Dschijn]

That is quite interesting!

I would have guessed that front/top intake would be cooler since your back is so open with mesh and not restrictive. To me it proves that cases are not the same and need individual testing to find the best solution.

[For Science!]

11 hours ago, Raxxath said:

-

Glad I was able to help. In most sensible watercooling loops that use low rpm fans, the air most probably stays too long inside the case, and therefore in compact cases like the C, it just needs to be pushed out of the case. Hence reverse intake.

[Raxxath]

6 hours ago, Dschijn said:

That is quite interesting!

I would have guessed that front/top intake would be cooler since your back is so open with mesh and not restrictive. To me it proves that cases are not the same and need individual testing to find the best solution.

2 hours ago, For Science! said:

Glad I was able to help. In most sensible watercooling loops that use low rpm fans, the air most probably stays too long inside the case, and therefore in compact cases like the C, it just needs to be pushed out of the case. Hence reverse intake.

 

I don't think that the Define C is all that unique, though. It has the same basic layout as 90% of the ATX tower cases out there, just a bit tighter inside.

 

To me it proves that there is a need for specifically designed cases for water cooling, and that a lot of water cooled setups out there are running inefficiently by recycling heated air through radiators.

 

For a while now I've been trying to design a case that's relatively compact and allows for multiple radiators to intake fresh air from outside the case, and then exhaust it outside the case (even though I don't have the means to build it myself). That would be the ideal setup.