What max coolant temp should I be targeting?

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
I don't personally like targeting deltas over ambient.

There is an ideal temperature for the conpenents to work at regardless of room temperature, which means if you operate in a warmer environment you need a more serious loop than if you operate in a cooler environment.
That doesn't change anything. If you lived in a sauna you'd OBVIOUSLY have to take that into account.
 

VirtualMirage

Limp Gawd
Joined
Nov 29, 2011
Messages
470
That is similar to how I have my radiator fan profile setup, but my fans start out at a higher RPM (1300 RPMs) and ramp up from there. For me and my case setup, 1300 RPMs I can barely hear them and these fans top out at 2200 RPMs. At 35C, they start to slowly ramp up to 1500 RPMs. By 40C they are at 1800 RPMs. If it ever gets too hot, they will max out at 2200 RPMs by 45C. Thankfully, in day to day use, it stays below 40C and I only ever see around 40-41C if I am doing an extended CPU burn test. The fans I am using are BeQuiet! SilentWings 3 120mm High Speed PWM fans. Their static pressure rating and noise levels spec'd out pretty well.
Just giving a slight update to my setup.

Yesterday, I decided to move two of my fans over to the top of my 240mm x 54mm radiator, creating a push/pull setup, to see if it made any difference. Idle and low load coolant temps dropped by around 1-2c while maintaining a fan speed of around 1,300 RPMs, sitting between 29-30c (dropping to as low as 28c at night). Running Furmark CPU Burner and Cinebench for extended periods results in coolant temp staying below 33c while fan speeds don't change much, maybe getting to around 1,400 RPMs (I didn't change my fan ramp profile). And while I do have more testing to do to check for consistency, these results are with my ambient room temp being around 4c higher than when I posted my results earlier (winter then vs late spring now). That is certainly lower temps than before all the while the fan speeds stay lower too (which is expected).

Note: my radiator fans are currently not setup to pull cool air from directly outside the case, they are exhausting air out from within the case. But I do have the radiator in its own chamber behind the motherboard, where the power supply resides, and there are 3 intake fans just on that side that feed air into that chamber (2x120mm fans in front and 1x140mm fan in rear all setup as intakes). Aside from the power supply and the D5 pump, there are no other components on that side. As such I expect the ambient temp that the radiator fans are pushing through the radiator are slightly higher than my ambient temp just outside the case. Current ambient temp just outside the case is around 25-26c. Running all my fans at max RPM doesn't help much in being able to getting my idle coolant than the 29-30c it currently sits at when idle. I was able to squeeze out maybe 1c lower at most, so 28-29c might be the floor of ambient temp from within the case. Even my RTX 3090 FE idling is sitting at that same temp. As such, I am probably shouldn't expect to reach any coolant temp lower than that unless I reverse my fan setup and begin pumping room temp air through the radiator first. And if I did that, it might only drop it 2c at most by my guess at the expense of sucking in more dust and hair through the radiator first (which is one of the reasons I have it setup the way I do, to minimize this).
 

n0ns3ns3

n00b
Joined
Mar 21, 2017
Messages
6
Well guys, here are few points for you to consider:
  • D5 is kinda industrial pump. It accidentally found its way into computer cooling (same as aquarium pumps).
  • The pump will work fine in -10C to +95C range. The problem is the original plastic top which is rated for 0C to 60C. Since we all use fancy tops, they can handle over 60C easily.
    • Fittings have no problem as they are simple brass parts.
    • Tubing could be a problem. For example some PETG can start to deform. On the other hand, soft tubing from EK is rated for -20°C to 80°C range.
  • Delta T have nothing to do with the max or safe temps. It is simply temperature difference between the coolant and ambient. Radiator manufacturers use delta T to describe how much heat can their radiator dissipate. Some even provide 2 points like 10C and 15C.
So basically the limit for the max coolant temp is probably tubing or components (CPU/GPU) way before the pump.

Since I joined the mini-ITX cult, I have years of running the loop with coolant at over 40C. Sometimes because that what needed for dumping 500W of heat with 280 rad without jet noise. And sometimes because I simply like my system quiet.
 

hititnquitit

[H]ard|Gawd
Joined
Apr 16, 2018
Messages
1,745
Well guys, here are few points for you to consider:
  • D5 is kinda industrial pump. It accidentally found its way into computer cooling (same as aquarium pumps).
  • The pump will work fine in -10C to +95C range. The problem is the original plastic top which is rated for 0C to 60C. Since we all use fancy tops, they can handle over 60C easily.
    • Fittings have no problem as they are simple brass parts.
    • Tubing could be a problem. For example some PETG can start to deform. On the other hand, soft tubing from EK is rated for -20°C to 80°C range.
  • Delta T have nothing to do with the max or safe temps. It is simply temperature difference between the coolant and ambient. Radiator manufacturers use delta T to describe how much heat can their radiator dissipate. Some even provide 2 points like 10C and 15C.
So basically the limit for the max coolant temp is probably tubing or components (CPU/GPU) way before the pump.

Since I joined the mini-ITX cult, I have years of running the loop with coolant at over 40C. Sometimes because that what needed for dumping 500W of heat with 280 rad without jet noise. And sometimes because I simply like my system quiet.
Whats your question buddy? Or are you just making an observation?
 

bal3wolf

Limp Gawd
Joined
May 14, 2008
Messages
471
Well guys, here are few points for you to consider:
  • D5 is kinda industrial pump. It accidentally found its way into computer cooling (same as aquarium pumps).
  • The pump will work fine in -10C to +95C range. The problem is the original plastic top which is rated for 0C to 60C. Since we all use fancy tops, they can handle over 60C easily.
    • Fittings have no problem as they are simple brass parts.
    • Tubing could be a problem. For example some PETG can start to deform. On the other hand, soft tubing from EK is rated for -20°C to 80°C range.
  • Delta T have nothing to do with the max or safe temps. It is simply temperature difference between the coolant and ambient. Radiator manufacturers use delta T to describe how much heat can their radiator dissipate. Some even provide 2 points like 10C and 15C.
So basically the limit for the max coolant temp is probably tubing or components (CPU/GPU) way before the pump.

Since I joined the mini-ITX cult, I have years of running the loop with coolant at over 40C. Sometimes because that what needed for dumping 500W of heat with 280 rad without jet noise. And sometimes because I simply like my system quiet.

Some good info there i always thought my water temps sucked i sit at around 20c amb and max my water gets up to is 30c most of the time its around 27 tho but after reading some of this looks like my temps arent to bad, if im not doing anything my water temp will drop down to 22-23c. I do have a 30mm 360 rad in top a ek 420 rad in front and a 60mm 360mm rad in bottom with dual ddc pumps in a bay res cooling a 5950x with a TechN block and a 3090 Kingpin HC with a mp5works active backplate on it gpu never passes 45c usualy under 40c and cpu gets up to 77c@ 4.75/4.65Ghz on 1.24-1.269 one thing to note is i dont think the ek rad is performing very well.
 

n0ns3ns3

n00b
Joined
Mar 21, 2017
Messages
6
Some good info there i always thought my water temps sucked i sit at around 20c amb and max my water gets up to is 30c most of the time its around 27 tho but after reading some of this looks like my temps arent to bad, if im not doing anything my water temp will drop down to 22-23c. I do have a 30mm 360 rad in top a ek 420 rad in front and a 60mm 360mm rad in bottom with dual ddc pumps in a bay res cooling a 5950x with a TechN block and a 3090 Kingpin HC with a mp5works active backplate on it gpu never passes 45c usualy under 40c and cpu gets up to 77c@ 4.75/4.65Ghz on 1.24-1.269 one thing to note is i dont think the ek rad is performing very well.
I have no personal experience with EK rads (only alphacool, swiftech and Hardware Labs), but AFAIR from reviews, they are far from best. And yet you have way more than enough of them to cool anything at reasonable noise. You have not mentioned fan's speed, so I can only guess the actual noise levels. The temperature of the GPU is excellent. CPU is fine. Though with Ryzen it is better to specify what sensor it is.
It is important to note, that the effectiveness of radiators is greatly reduced if they placed one after another. For example, if in your setup the front and bottom are used as intake and the top as exhaust, the top rad doing very little for cooling. That happens because it uses much hotter air. This statement is correct for low fan speed. Higher airflow increases the effectiveness of the last rad. It works similar to radiator thickness. Thicker radiators need higher airflow to show advantage over thinner ones. For me and my usecase, there is no point in over 45mm thickness of the rad.
 

bal3wolf

Limp Gawd
Joined
May 14, 2008
Messages
471
I have no personal experience with EK rads (only alphacool, swiftech and Hardware Labs), but AFAIR from reviews, they are far from best. And yet you have way more than enough of them to cool anything at reasonable noise. You have not mentioned fan's speed, so I can only guess the actual noise levels. The temperature of the GPU is excellent. CPU is fine. Though with Ryzen it is better to specify what sensor it is.
It is important to note, that the effectiveness of radiators is greatly reduced if they placed one after another. For example, if in your setup the front and bottom are used as intake and the top as exhaust, the top rad doing very little for cooling. That happens because it uses much hotter air. This statement is correct for low fan speed. Higher airflow increases the effectiveness of the last rad. It works similar to radiator thickness. Thicker radiators need higher airflow to show advantage over thinner ones. For me and my usecase, there is no point in over 45mm thickness of the rad.

Yea my old case layout was front rad intake top rad exhaust bottom rad intake it waset cooling as good as it could cause i think i was adding more heat to the loop with top rad getting hot air i was lazy and didn't wanna spend the time to redo the loop. But last night i couldnt sleep so im trying all rads to exhaust water temp will still rise up to 30c but takes 5-6hrs for it to do it and if i turn my fans up i dont go past 27c on water temp now so changing fans around did improve temps.
 

n0ns3ns3

n00b
Joined
Mar 21, 2017
Messages
6
Yea my old case layout was front rad intake top rad exhaust bottom rad intake it waset cooling as good as it could cause i think i was adding more heat to the loop with top rad getting hot air i was lazy and didn't wanna spend the time to redo the loop. But last night i couldnt sleep so im trying all rads to exhaust water temp will still rise up to 30c but takes 5-6hrs for it to do it and if i turn my fans up i dont go past 27c on water temp now so changing fans around did improve temps.
Interesting. If all three rads are exhaust, where is intake?
I can only think of mesh side panel or even better no side panel at all. Otherwise the pressure within case will be reduced. That will also negatively affect cooling performance and noise levels.
 

bal3wolf

Limp Gawd
Joined
May 14, 2008
Messages
471
Have A 120mm in back pulling in air now and the x71 has mesh both sides of bottom bay so the bottom left rad fan get air. Heat coming out of rads is much cooler now heres my case the core x71 it has quite a few vents to let air in the top left spot is doing intake its working might not seem perfect but best temps i have seen.
 
Last edited:

n0ns3ns3

n00b
Joined
Mar 21, 2017
Messages
6
Have A 120mm in back pulling in air now and the x71 has mesh both sides of bottom bay so the bottom left rad fan get air. Heat coming out of rads is much cooler now.
The bottom rad gets enough air for sure. Just curious, if you fix the fan speed and compare side panel open with closed?
My guess there will be some difference because the front and top rads are kinda in less than ideal intake as for them the fresh air comes mostly from rear and it is way less open area vs top + front.
 

bal3wolf

Limp Gawd
Joined
May 14, 2008
Messages
471
The bottom rad gets enough air for sure. Just curious, if you fix the fan speed and compare side panel open with closed?
My guess there will be some difference because the front and top rads are kinda in less than ideal intake as for them the fresh air comes mostly from rear and it is way less open area vs top + front.
I cant run with side panel off 24/7 i don't sensitive hearing but will give it a shot i tune my fans to a noise profile i can handle and my german shepard is kinda noisey shes tryings to see what the computer looks like lol. But i took side panel off to check air flow on the new fan and panels been off for 15-20 mins temps might changed 1-2c with it off not a huge swing.
 

n0ns3ns3

n00b
Joined
Mar 21, 2017
Messages
6
I cant run with side panel off 24/7 i don't sensitive hearing but will give it a shot i tune my fans to a noise profile i can handle and my german shepard is kinda noisey shes tryings to see what the computer looks like lol. But i took side panel off to check air flow on the new fan and panels been off for 15-20 mins temps might changed 1-2c with it off not a huge swing.
Seems like there is no point in leaving the side panel open in your case.
I had a bit different experience with 2x240 rads as exhaust in Fractal's Define S nano. Opening side panel made much larger difference of 5-6C.
Eventually, I ended with a single 280 that did better job due to much less restriction on fresh air intake with closed panels. Since then, single 280 rad loop is my favorite. It can handle 300-350w of heat at very low noise (fans below 1000rpm) and 500-600w at very reasonable noise (fans at 1400-1500rpm) with room temperature of 26C or lower. Though I only had ~200w CPU and 330w GPU to play with.
 

bal3wolf

Limp Gawd
Joined
May 14, 2008
Messages
471
Seems like there is no point in leaving the side panel open in your case.
I had a bit different experience with 2x240 rads as exhaust in Fractal's Define S nano. Opening side panel made much larger difference of 5-6C.
Eventually, I ended with a single 280 that did better job due to much less restriction on fresh air intake with closed panels. Since then, single 280 rad loop is my favorite. It can handle 300-350w of heat at very low noise (fans below 1000rpm) and 500-600w at very reasonable noise (fans at 1400-1500rpm) with room temperature of 26C or lower. Though I only had ~200w CPU and 330w GPU to play with.

Im not 100% sure how much heat im dumping into the loop the 5950x iv seen use up to 270Watts of power and gaming the 3090 kingpin will go up to 520 watts and then i have the mp5works active back plate cooler on it also so i figure i could be dumping 600-1000 watts of heat into the rads im keeping a 5-7c delta now looks like as bubbles leave the loop pretty happy with results its really hard to find perfect water cooling cases most it seems.
 

Tsumi

[H]F Junkie
Joined
Mar 18, 2010
Messages
13,551
I have no personal experience with EK rads (only alphacool, swiftech and Hardware Labs), but AFAIR from reviews, they are far from best. And yet you have way more than enough of them to cool anything at reasonable noise. You have not mentioned fan's speed, so I can only guess the actual noise levels. The temperature of the GPU is excellent. CPU is fine. Though with Ryzen it is better to specify what sensor it is.
It is important to note, that the effectiveness of radiators is greatly reduced if they placed one after another. For example, if in your setup the front and bottom are used as intake and the top as exhaust, the top rad doing very little for cooling. That happens because it uses much hotter air. This statement is correct for low fan speed. Higher airflow increases the effectiveness of the last rad. It works similar to radiator thickness. Thicker radiators need higher airflow to show advantage over thinner ones. For me and my usecase, there is no point in over 45mm thickness of the rad.

I was going to respond to this a while ago but never got around to it.

Your statement that the exhaust radiator is doing little for cooling in a low airflow setup is not universally true. By plumbing things correctly, you can take advantage of countercurrent flow, where the hottest water hits the exhaust radiator first and then the cooler water hits the intake radiator. The air hitting the exhaust radiator will therefore always be cooler than the water unless other components inside the computer heat it up.

Several Black Ice radiators are set up this way in that there is an optimal flow direction. Those Black Ice radiators have the flow going down half of the thickness of the radiator before returning on the other half, unlike most radiators where the flow is split along the width. Making sure the inlet goes to the exhaust side of the radiator has a measurable impact on performance, especially at lower fan speeds.
 
Top