240mm Radiator VS. 280mm - How much of a difference does it make?

[Plasma_43]

I am debating whether or not I should purchase a 240mm AIO or a 280mm variant. My question is how much more cooling can 40mm net you? I am planning to use either a Kraken X63 from NZXT or Kraken X53 respectively. I am also planning on using the Meshify C as my case, and I am going to put the radiator in the front. With the Meshify C, if you have a 280mm radiator in the front, there is no space for anymore front case fans, whereas with a 240, you can squeeze one more in there. If i can, I want to squeeze another intake fan in the front if possible, to maximize intake capacity. Bottom line; does 280mm give you a tangible boost in performance when compared to 240mm, or should I simply get the X53, put in an extra case fan and call it a day?

[Sorenson]

it depends on your CPU and the quality of the radiator. Gamers nexus just did a video where a 240 outperformed a 280mm radiator so there are more factors than just size. however, If your case can fit a 280, the 280 doesn’t cost too much more, and you’re using a very hot CPU then i would go with the 280.

[Plasma_43]

17 minutes ago, Sorenson said:

it depends on your CPU and the quality of the radiator. Gamers nexus just did a video where a 240 outperformed a 280mm radiator so there are more factors than just size. however, If your case can fit a 280, the 280 doesn’t cost too much more, and you’re using a very hot CPU then i would go with the 280.

I am currently planning on buying the 3700X and doing a moderate overclock; however, I am planning on purchasing all the parts around boxing day, and depending on the performance and price of the 3800XT, I might opt to purchase that one due to its higher base clock and better overclocking potential. So if I end up going with the higher wattage CPU, I should probably get the 280mm AIO to balance it out, right?

[Sorenson]

10 minutes ago, Plasma_43 said:

I am currently planning on buying the 3700X and doing a moderate overclock; however, I am planning on purchasing all the parts around boxing day, and depending on the performance and price of the 3800XT, I might opt to purchase that one due to its higher base clock and better overclocking potential. So if I end up going with the higher wattage CPU, I should probably get the 280mm AIO to balance it out, right?

idk how the 3800XT will perform heat wise, but my 3800x is cooled extremely well on corsair h100i pro 240mm AIO even with PBO overdrive enabled. I don’t overlook on my 3800x (besides PBO overdrive) since for gaming letting ryzen handle the OC generally gives the best results. Ryzen chips run a little hotter than intel CPUs at idle, but with the 240 aio my 3800x is in the 40 to low 50c and in gaming its fluctuates between high 60s and low 70s. My temps are also a little higher since I’m running my system in an SFX case and have the fans and pump set to quiet mode. 

[Zyleyus]

Actually pretty much, a 280 one can counter a 360 one.

240 cooling area: 240*120=28,800cm²

280 cooling area: 280*140=39,200cm²

360 cooling area: 360*120=43,200cm²

28,800/39,200≈73%

39,200/43,200≈91%

And you use only 2 fans rather than 3, cheaper, quieter.

So you should always go with 140 fans if you can.

[Maticks]

Bigger radiators make for a bigger surface area.

 If you want to run quieter fans more surface area will mean you can turn the speed down.

 Comes down to what your case will fit.

 You can never have a big enough radiator. You just turn your fans down more if it doesn’t get warn enough.

rad costs are pretty low these days

[bowrilla]

On 6/27/2020 at 8:12 AM, Zyleyus said:

Actually pretty much, a 280 one can counter a 360 one.

240 cooling area: 240*120=28,800cm²

280 cooling area: 280*140=39,200cm²

360 cooling area: 360*120=43,200cm²

28,800/39,200≈73%

39,200/43,200≈91%

And you use only 2 fans rather than 3, cheaper, quieter.

So you should always go with 140 fans if you can.

Well ... you have to consider the fans as well and their actual coverage. The results will be different - and actually even more surprising.

 

You have to compare those areas AND include the more or less dead zones beneath the hubs. Let's assume 1,600mm² hub area (should be an ok average but will vary from fan to fan and might be smaller, i.e. beQuiet Silent Wings 3 is ~1400mm² NOTE: that will only change the values below by ~2%) and let's assume the blade diameter is ~10mm smaller than the nominal size of the fan (so 110mm for a 120mm fan, 130mm for a 140mm fan, it's a rough value and will differ from fan to fan). That means the actually vented are of a 120mm fan is ~7,903mm² and ~11,673mm².

 

This results in the following active (!) cooling area values:

240mm: 15,804mm² / 67.7%

280mm: 23,346mm² / 100%

360mm: 23,709mm² / 101.6%

 

The not acvtively vented area is not doing much. If there's no airflow the heat transfer is terribly slow and inefficient.

 

Comparing the actively vented area with average fan dimensions, the difference between 280mm and 360mm is absolutely negligable.

 

BUT, smaller 120mm fans often have better static pressure performances - obviously because assuming the same gap between the fan blade housing and the fan blade the leakage area grows exponentially. For example, just compare the 120mm and 140mm Noctua Redux fans NF-A14 and NF-F2. Both are PWM, both are 1500RPM max. The 140mm version has a max H2O column of 2.08mm while the 120mm offers 2.61mm. That's about 25% reduction at max rpm, while the fan size only changed ~16-18%. The air volume however went up by ~50%.

 

Then there's noise. The bigger fans with more moved air means a higher base noise at the same rpm. +3dB doubles the energy, +6db doubles the sound pressure, +10db doubles the perceived loudness. Again, the Noctua fans have values of 24.6dB (140mm) and 22.4dB (120mm). Due to the logarithmic nature of the dB scale you can't just simply add them. Using the decibel calculator at noisemeters 2x140mm fans at max rpm have a combined dB value of 27.6dB and 3x120mm add up to 27.2dB. So at the same rpm a 280mm radiator with 2 fans is about as loud as a 360mm radiator with 3 fans - there's some difference, it might be perceivable if you concentrate but 0.4dB difference is really not much.

 

But since you'll have less static pressure you might need to increase fan speed to acchieve the same heat transfer between ambient air and radiator and therefore the same water temps as the 360mm radiator gives you. This will give the smaller fans an edge.

 

tl;dr: Go 140mm but pay more attention to your fans of choice. Fan performance is more important here. 

 

EDIT: added the relative change for the smaller Be Quiet Silent Wings 3 hubs, it's just ~2%.

[Plasma_43]

3 hours ago, bowrilla said:

Well ... you have to consider the fans as well and their actual coverage. The results will be different - and actually even more surprising.

 

You have to compare those areas AND include the more or less dead zones beneath the hubs. Let's assume 1,600mm² hub area (should be an ok average but will vary from fan to fan and might be smaller, i.e. beQuiet Silent Wings 3 is ~1400mm² NOTE: that will only change the values below by ~2%) and let's assume the blade diameter is ~10mm smaller than the nominal size of the fan (so 110mm for a 120mm fan, 130mm for a 140mm fan, it's a rough value and will differ from fan to fan). That means the actually vented are of a 120mm fan is ~7,903mm² and ~11,673mm².

 

This results in the following active (!) cooling area values:

240mm: 15,804mm² / 67.7%

280mm: 23,346mm² / 100%

360mm: 23,709mm² / 101.6%

 

The not acvtively vented area is not doing much. If there's no airflow the heat transfer is terribly slow and inefficient.

 

Comparing the actively vented area with average fan dimensions, the difference between 280mm and 360mm is absolutely negligable.

 

BUT, smaller 120mm fans often have better static pressure performances - obviously because assuming the same gap between the fan blade housing and the fan blade the leakage area grows exponentially. For example, just compare the 120mm and 140mm Noctua Redux fans NF-A14 and NF-F2. Both are PWM, both are 1500RPM max. The 140mm version has a max H2O column of 2.08mm while the 120mm offers 2.61mm. That's about 25% reduction at max rpm, while the fan size only changed ~16-18%. The air volume however went up by ~50%.

 

Then there's noise. The bigger fans with more moved air means a higher base noise at the same rpm. +3dB doubles the energy, +6db doubles the sound pressure, +10db doubles the perceived loudness. Again, the Noctua fans have values of 24.6dB (140mm) and 22.4dB (120mm). Due to the logarithmic nature of the dB scale you can't just simply add them. Using the decibel calculator at noisemeters 2x140mm fans at max rpm have a combined dB value of 27.6dB and 3x120mm add up to 27.2dB. So at the same rpm a 280mm radiator with 2 fans is about as loud as a 360mm radiator with 3 fans - there's some difference, it might be perceivable if you concentrate but 0.4dB difference is really not much.

 

But since you'll have less static pressure you might need to increase fan speed to acchieve the same heat transfer between ambient air and radiator and therefore the same water temps as the 360mm radiator gives you. This will give the smaller fans an edge.

 

tl;dr: Go 140mm but pay more attention to your fans of choice. Fan performance is more important here. 

 

EDIT: added the relative change for the smaller Be Quiet Silent Wings 3 hubs, it's just ~2%.

I really appreciate the attention to detail. When it comes to fans, i've heard good things about NZXT's stock radiator fans, but if those don't work, I found an alternative from arctic. Do you know how the NZXT fans fare when paired with a radiator?

[bowrilla]

1 hour ago, Plasma_43 said:

I really appreciate the attention to detail. When it comes to fans, i've heard good things about NZXT's stock radiator fans, but if those don't work, I found an alternative from arctic. Do you know how the NZXT fans fare when paired with a radiator?

Never had them in use so can't speak from experience but looking at their datasheet the AER P fans are on the higher RPM side distorting their H2O pressure value. And at full speed they seem to be awfully loud. Funny enough the Noctua Industrial PPC 140 fans beat the NZXT AER P fans in every discipline: higher significantly higher H2O column, higher m³/h, lower dB yet still higher RPM. They have good bearings though which is a plus - their blade design must be poor then resulting in such a noisy fan.

 

At 1800 RPM and above I'd rather look at the airflow (m³/h or CFM) because even the the radiator's restrictiveness will reduce the airflow to half and maybe even a bit further, the sheer force of enough air volumen being pushed through will get the job done. Static pressure is most interesting at lower RPM. 

 

Generally speaking you'll be aiming for well under 1000 RPM for your radiator fans and probably sub 600 RPM for the case fans. So stay clear from high RPM fans. There's usually no reason to use them in a desktop watercooling scenario imho unless you don't give a damn about noise and want to squeeze out the last bit of performance when overclocking. But then you're definitely not going for AiOs.

 

There's no way of knowing without further data (or a lot more physics) to determine the H2O column at a comparable reference speed. As far as I know and from looking into some fans that are offered in more than one speed setting, the H2O column is not linearly dependant on the rpm (Noctua Redux NF-P14s 900 / 1200 / 1500 with H2O columns of 0.77 / 1.29 / 1.91). If I had to make a guess from 3 datapoints it looks more like an exponential curve but I have no idea about the aerodynamics at play so at this point I'm out without some measuring rig.

 

No idea about the Arctic fans but make sure not to fall for max values. They are pointless. and can only give you a hint on how well it will perform. You won't be using them at max RPM. Why exactly Arctic? Looking at their lineup of case fans I can mostly spot airflow optimized 120mm and 140mm PWM fans? Look at Corsair or Noctua, even the Be Quiet Silent Wings 3 would probably perform better.

 

Corsair at least gives you min/max values (at least for some?). The ML140 PWM fan has a speed range of 400-2000RPM and a static pressure of 0.2 - 3.0 mm H2O. Most people will keep it in the 400-800 RPM range. There's no point in having the radiator fans adapt to CPU temps. If at all reacting to water temps would be a lot more useful.

 

tl;dr A reference speed of 500 or 800, heck even 1000 RPM would give people a lot more information about what they're actually buying. But I guess some manufacturers are just compensating their bad designs with higher revs.

[Plasma_43]

53 minutes ago, bowrilla said:

Never had them in use so can't speak from experience but looking at their datasheet the AER P fans are on the higher RPM side distorting their H2O pressure value. And at full speed they seem to be awfully loud. Funny enough the Noctua Industrial PPC 140 fans beat the NZXT AER P fans in every discipline: higher significantly higher H2O column, higher m³/h, lower dB yet still higher RPM. They have good bearings though which is a plus - their blade design must be poor then resulting in such a noisy fan.

 

At 1800 RPM and above I'd rather look at the airflow (m³/h or CFM) because even the the radiator's restrictiveness will reduce the airflow to half and maybe even a bit further, the sheer force of enough air volumen being pushed through will get the job done. Static pressure is most interesting at lower RPM. 

 

Generally speaking you'll be aiming for well under 1000 RPM for your radiator fans and probably sub 600 RPM for the case fans. So stay clear from high RPM fans. There's usually no reason to use them in a desktop watercooling scenario imho unless you don't give a damn about noise and want to squeeze out the last bit of performance when overclocking. But then you're definitely not going for AiOs.

 

There's no way of knowing without further data (or a lot more physics) to determine the H2O column at a comparable reference speed. As far as I know and from looking into some fans that are offered in more than one speed setting, the H2O column is not linearly dependant on the rpm (Noctua Redux NF-P14s 900 / 1200 / 1500 with H2O columns of 0.77 / 1.29 / 1.91). If I had to make a guess from 3 datapoints it looks more like an exponential curve but I have no idea about the aerodynamics at play so at this point I'm out without some measuring rig.

 

No idea about the Arctic fans but make sure not to fall for max values. They are pointless. and can only give you a hint on how well it will perform. You won't be using them at max RPM. Why exactly Arctic? Looking at their lineup of case fans I can only spot airflow optimized 120mm and 140mm PWM fans. Look at Corsair or Noctua, even the Be Quiet Silent Wings 3 would probably perform better.

 

Corsair at least gives you min/max values (at least for some?). The ML140 PWM fan has a speed range of 400-2000RPM and a static pressure of 0.2 - 3.0 mm H2O. Most people will keep it in the 400-800 RPM range. There's no point in having the radiator fans adapt to CPU temps. If at all reacting to water temps would be a lot more useful.

 

tl;dr A reference speed of 500 or 800, heck even 1000 RPM would give people a lot more information about what they're actually buying. But I guess some manufacturers are just compensating their bad designs with higher revs.

I appreciate the attention to detail. When it comes to why I chose an Arctic fan as an alternative, I found a well-performing static-pressure fan at a very reasonable price. This is just one I have in mind though, and I will look into the Noctua Redux NF-P14s. I want to avoid airflow fans when looking for radiator fans, as they simply cannot mesure up to static-pressure fans when used on a radiator. 

[bowrilla]

Just now, Plasma_43 said:

I appreciate the attention to detail. When it comes to why I chose an Arctic fan as an alternative, I found a well-performing static-pressure fan at a very reasonable price. This is just one I have in mind though, and I will look into the Noctua Redux NF-P14s. I want to avoid airflow fans when looking for radiator fans, as they simply cannot mesure up to static-pressure fans when used on a radiator. 

Well, they can if you run them at high RPM. A fan with 2-3mm H2O is not really a full blown static pressure fan but rather a hybrid one. As mentioned before: there's not much to running a static pressure fan at 2000 RPM. At that speed you can basically also use an airflow fan and just force enough air through. Corsair's ML140 fans can come in a dual package for I think 30-ish bucks MSRP. That's pretty decent. They are hybrid fans as well and their min RPM of 400 is okay. They are a bit noisy at max RPM but that's also 2000 RPM, no idea how they perform at lower speeds.

[oeci]

Leaving all the science aside it's important to realize what CPU we are dealing with, it's a 3700X, possibly a 3800XT. My AIO with a 240 mm radiator runs constantly at 50 % fan speed (900 RPM), is barely hearable and keeps the 3800X very cool, the highest temps I've seen so far at 100 % CPU usage (while exporting photos in my RAW converter or while stressing the CPU using the Aida64 stress test) is 70 °C with a lot of headroom for either setting the fan speed even lower during idle times or for overclocking. I'd go for a 240 mm rad. because it's more than enough for the CPUs you've mentioned.

[Plasma_43]

2 hours ago, oeci said:

Leaving all the science aside it's important to realize what CPU we are dealing with, it's a 3700X, possibly a 3800XT. My AIO with a 240 mm radiator runs constantly at 50 % fan speed (900 RPM), is barely hearable and keeps the 3800X very cool, the highest temps I've seen so far at 100 % CPU usage (while exporting photos in my RAW converter or while stressing the CPU using the Aida64 stress test) is 70 °C with a lot of headroom for either setting the fan speed even lower during idle times or for overclocking. I'd go for a 240 mm rad. because it's more than enough for the CPUs you've mentioned.

Very good to know. What make and model is the AIO you are using?

[oeci]

My AIO is the Deepcool GamerStorm Captain 240 Pro.

[CanNuckingfutzman]

I've been running water coolers for a very long time. I've run 120, 240, and 280...from cheap rigs to very expensive, massive copper cores. Long story short, as others have mentioned, there's enough variables to make this hard to answer with a simple, "You should run a...".  But, my experience is that good 280mm radiator even with marginal 140mm fans is "better" in the real-world than a 240mm equivalent. Thsi is only because of how I define "better".  I find that the sound produced by 120mm fans at higher RPMs to be more annoying than 140mm fans at lower speed, ergo 140mm fans at low speed is a "better" setup for me. For example, I use a 280mm thick copper radiator (45mm) with 140mm Artic fans, which run whisper quiet all the time. This setup just has too much mass to get very hot and the fans basically run at nearly idle speeds all the time. Then again, I have a cheap 240mm Enermax AIO in another rig that run push/pull configuration (4x120mm fans) that is also whisper quiet because it's totally overkill for the I5-9600K it's cooling. In this case, also never going over 50% of full speed.      

[Ravendarat]

Dude, 11 months? Really think this needed to be bumped?