Internal radiator blowing into the case - bad?

[VioletGiraffe]

I have a Corsair 750D case. I have two radiators inside: a thin XSPC EX280 on the front panel blowing in, and a thick 360 in the top blowing out. I don't have any pictures of my own system, but the loop layout is very much like this, except I have no bottom radiator (this image is not mine!):

 

 

My question is: how bad for the overall system cooling performance is it that the front radiator is heating the air inside the case (and, thus, the air coming into the top rad that's blowing outwards)? Did anyone experiment with such layouts and measured them? Obviously, moving the front rad outside the case would help. But could this layout be so bad that it's better to drop the front 280 mm rad altogether? Or does it still bring some net benefit?

Of note: I have very low fan speeds (can't stand fan noise, and it's part of the reason why I like liquid cooling in the first place - good heat dissipation at low airflow).

[Himommies]

Purely performance wise it isn't worth the effort but dust on the other hand.You might wanna change it to push

[VioletGiraffe]

1 minute ago, Himommies said:

Purely performance wise it isn't worth the effort but dust on the other hand.You might wanna change it to push

Quite on contrary: there is a dust filter in front of this rad (on the outside of the case), and there's already more fans blowing out than in (1x140 + 3x120 out, 2x140 in), and no space to add more intake fans.

Hm. I guess the easiest way to kind of test the system without the front rad is to disable this rad's fans. Problem is, the ambient temperature is always different, and it moves up and down, never stable (usually it moves up slowly after I turn the PC on). Such measurements need a thermo-regulated  lab, my results will be inconclusive. Unless I'll see a difference of over 5 degrees C, but I doubt it'd be that drastic.

[Himommies]

Just now, VioletGiraffe said:

Quite on contrary: there is a dust filter in front of this rad (on the outside of the case), and there's already more fans blowing out than in (1x140 + 3x120 out, 2x140 in), and no space to add more intake fans.

Hm. I guess the easiest way to kind of test the system without the front rad is to disable this rad's fans. Problem is, the ambient temperature is always different, and it moves up and down, never stable (usually it moves up slowly after I turn the PC on). Such measurements need a thermo-regulated  lab, my results will be inconclusive. Unless I'll see a difference of over 5 degrees C, but I doubt it'd be that drastic.

Assuming dust isn't a issue it's not worth the hassle especially if it's a custom loop

[VioletGiraffe]

It is a fully custom loop. It's already assembled, has been running for almost a year. Lately I started to wonder whether I need to tear it down to actually improve performance. That's essentially what I'm asking.

[SaviorSelf_666]

I have a 760T and have a 280 rad in the front and 360 rad top, both are intake. I have a 140 fan in the rear as exhaust and 120 fan in the bottom as exhaust. Works good for me. If I have the top as exhaust then I would have warm air going through my rad, I'd rather have both my rads get fresh air than used warm air.

[NoxiousOdor]

2 minutes ago, VioletGiraffe said:

It is a fully custom loop. It's already assembled, has been running for almost a year. Lately I started to wonder whether I need to tear it down to actually improve performance. That's essentially what I'm asking.

I would only change it if you aren't happy with its current performance

[VioletGiraffe]

10 minutes ago, WhisperingKnickers said:

I would only change it if you aren't happy with its current performance

Eh. It's overkill for the current setup (big overkill). But I wonder whether I'm not getting the best possible performance from the cooling setup. Besides, I'm planning to upgrade the system this autumn or thereabouts, and I aim for some serious hardware - hopefully a Threadripper 12-core CPU (failing that, a Skylake-X 8-core) plus a 4K-capable graphics card (at least a GTX1080). I estimate the current system's peak heat production at 300W-350W, and the one I have planned will be ~550W, maybe even 600.

[ZenMonkey]

You won't see a difference in performance or temperature. Also, you need intake fans, so unless you want to eliminate the front rad (which would affect your heat and performance much more), it doesn't seem like you have a choice.

 

Like others have said, not worth it.

[NoxiousOdor]

1 minute ago, VioletGiraffe said:

Eh. It's overkill for the current setup (big overkill). But I wonder whether I'm not getting the best possible performance from the cooling setup. Besides, I'm planning to upgrade the system this autumn or thereabouts, and I aim for some serious hardware - hopefully a Threadripper 12-core CPU (failing that, a Skylake-X 8-core) plus a 4K-capable graphics card (at least a GTX1080). I estimate the current system's peak heat production at 300W-350W, and the one I have planned will be ~550W, maybe even 600.

If it is already overkill and you are planning on upgrading it in the somewhat near future anyway I wouldn't worry about it right now. It sounds like it is fine and unless you want to do it just for fun, it probably won't be worth the time

[VioletGiraffe]

12 minutes ago, AdaptBlood said:

If I have the top as exhaust then I would have warm air going through my rad

Exactly my concern. However, if you have air through both rads blowing in, then you have optimal cooling performance for the LCS loop, BUT you're dumping all that heat into the case, rising the temperatures of all the air-cooled components inside. Also not ideal, and could even shorten the life span of your mainboard (think of the CPU MOSFETs, esp. if you have a multi-core HEDT overclocked CPU that consumes 250-300W).

[VioletGiraffe]

Thanks. If I was getting fun out of rebuilding this stuff I would just do it and see for myself, rather than ask. As it happens, I enjoy using the system, I enjoy planning the build, but I hate actually building it. That's why I'm asking instead of just doing and experimenting. So thanks for your replies, looks like you have unanimously put my concern to rest

[SaviorSelf_666]

True but having good flow throughout the case, keeps air moving consistently. The air flow even at that temp should be cooler than what the VRM gets at, i assume. I personally have my VRM's in my loop. Mosfets are rated pretty high for temp, no?

[xHadrian]

This video by JayzTwoCents should help you:

 

 

[VioletGiraffe]

1 minute ago, AdaptBlood said:

Mosfets are rated pretty high for temp, no?

Yes, you're right. They should have normal operating temps of about 100 C, possibly 125 C. The VRM capacitors in all the decent boards are 105 C, and chokes are probably not sensitive to temps in this range, either. But MOSFETs are very small surface area pieces with relatively small heatsinks so they may still get hot (whether or not they easily get dangerously hot, I do not know).

Having them in the loop makes all the difference, of course.

[VioletGiraffe]

23 hours ago, xHadrian said:

This video by JayzTwoCents should help you:

Interesting, but totally inconclusive: he uses a low TDP CPU, and he only includes this CPU in the loop. The air coming out of my top 360 mm rad is pretty hot, around 36-38 C with ambient being 24. I have a very cool CPU and reasonably hot GPU in the loop, for the grand total of 250-300W.