Question Surprising results upgrading heatsink fans

[wicked_sticky]

Thermalrights Frost Spirit is ostensibly the best air cooler for ryzen 7000 cpus. But the fans are a little dated. I tried various combos of higher static pressure, rpm, cfm fans. At best I got 0.3°C lower temps running cinebench using (91.9 vs 91.6 with a 7700x using Frost Commander's 1800/1850 rpm fans).

For that 0.3° fan noise is 8db louder than using stock frost spirit fans.

 

[zinkles]

Hm.. Interesting. I think the 0.3°C is well enough margin for error as well. What I mean is that there can be 1-2°C differences between tests and its mostly negligible.

 

[geofelt]

The limiting factor on any cooler(including aio) is the size of the radiator and the design of the fins.
Changing the fans changes the airflow, but not the radiator.
Once you have sufficient airflow to carry away the heat from the fins, more or less does not matter that much.

A secondary design issue is the number of heat pipes which move heat from the cpu to the fins.
The frost spirit has 4 per radiator.
A design compromise that keeps costs low.

 

[wicked_sticky]

Hm.. Interesting. I think the 0.3°C is well enough margin for error as well. What I mean is that there can be 1-2°C differences between tests and its mostly negligible.

These are averages of 5 tests each configuration (and tried a few other combos) but your not wrong, making sure the fans are mounted as low as possible made a bigger difference than the extra 40cfm.
Using 2x140mm fans got the worst results on the FS140, I'm guessing because some air can pass around the heatsink.
The fans being lined up with the upper lip of the heatsink was 1.5° worse than push as close to the mobo as possible.
I tried running cinebench with the 120mm 1850rpm fan in the front and a 120mm 1900rpm schythe kaze maru fan in the center, same temps as the 120+140 fan combo. Dual 120mm fan hit 65db. way too loud even for someone who doesnt care about fan noise. FC stock fans max out at 63db FS stock fans 55db. (all measured from the bottom front corner of the case).

The limiting factor on any cooler(including aio) is the size of the radiator and the design of the fins.
Changing the fans changes the airflow, but not the radiator.
Once you have sufficient airflow to carry away the heat from the fins, more or less does not matter that much.

A secondary design issue is the number of heat pipes which move heat from the cpu to the fins.
The frost spirit has 4 per radiator.
A design compromise that keeps costs low.

Since I own both heatsinks and a plethora of fans I was trying to find the best combo before putting the rest of the build together.

Frost Commanders 5 heat pipes was at least 6° warmer and a lot louder under load

I couldnt run cinebench on the frost commander without thermal throttling
(R23 score was 18600 vs 20300 on with the frost spirit)

My old single tower, 4x 6mm heatpipe cryorig H5 Ult. (using the frost commander's fans) teetered at the thermal limit but was able to run cinebench for 10min w/o throttling.

I bought the black FC140 to match everthing in the case, but it's not worth the hit in performance. The only good thing about it is the 120mm fan. Crazy high pressure and airflow for a budget heatsink.

 

[JohnBonhamsGhost]

At best I got 0.3°C lower temps

i don't see any mention of the case this system is residing in or the overall cooling setup.

if there is too much stagnant hot air circulating within the case without proper cool air intake and warm air exhaust then it is much harder to lower temperatures no matter what cooler+fans may be in use.

 

[wicked_sticky]

i don't see any mention of the case this system is residing in or the overall cooling setup.

if there is too much stagnant hot air circulating within the case without proper cool air intake and warm air exhaust then it is much harder to lower temperatures no matter cooler+fans may be in use.

Its just king 95 with all 4 panels removed... essentially just an upright motherboard tray and a rear exhaust fan. zero chance of stagnant air