To write that we have something mapped out to the last detail is perhaps too bold, but after proper preparation, few pieces of hardware are as easy to evaluate as fans. Of course, this had to be preceded by long preparations, developing a methodology, but you already know the story. What you don’t know yet is the first fruit, or rather the results of Akasa, SilentiumPC, SilverStone, Xigmatek or more exotic Reeven fans.
Initial warm-up…
Before we even start measuring anything, we let the fans run “idle” for a few minutes after plugging them in. This is because immediately after a cold start the fans reach different parameters than after a certain amount of short-term operation.
Until the operating temperature of the lubricant is stabilized, a typically lower maximum performance is achieved. This is because at lower temperatures the lubricant is denser, which is associated with higher friction. Therefore, the fans do not reach maximum speed immediately, but only after the first few seconds. Before the first measurements, we therefore leave the fans running for at least 300 seconds at 12 V, or 100 % PWM intensity.
…and speed recording
The speed of the fans is monitored using a laser tachometer, which reads the number of revolutions from a reflective sticker on the rotor. For this purpose, we use the UNI-T UT372 device, which also allows real-time averaging of samples. Thus, we do not record the peak value in the graphs, but the average speed value from a 30-second time period.
However, the speed itself is a relatively unimportant parameter that is often given more attention than is appropriate. This is the case even in many fan or cooler tests, where speed is used to normalize the different modes in which other variables are measured.
However, hyper-focusing on a specific speed is a rather unfortunate decision if only because the fans don’t gain any commonality. At the same speed all other variables are different, there is no intersection. It can be noted that a better normalization would have been by any other variable, whether it be static pressure, flow or noise level, which wins in our case. But more on that in the next chapter.
We only measure the speed so that you can associate a particular parameter (such as the amount of static pressure or some noise level) with something according to which you can adjust the fan yourself. Perhaps for that alone, the information about the achieved speed is useful. As part of the fan analysis, we will also indicate what the fans’ starting and minimum speeds are. Start-up speeds tend to be higher than minimum speeds because more force is required to get the rotor moving than once the fan rotor is spinning, and a minimum power intensity is sought at which the fan does not stall.
Thank you so much for this! I thought that I might have just gotten a bad batch with my CM AIO cooler. Although, in my case, it’s not rattling but an uneven hum that keeps going up and down.
And thanks to your other reviews, I ordered some Noctua NF-A12x25 today. That’ll hopefully solve that. All of my other fans are Noctua too and while not all the new and best, they at least produce steady noise.
Using Noctua NF-A12x25 fans will definitely solve the issue you’re describing. The sound of the tested Cooler Master fans can be perceived differently by everyone. It’s certainly not clean in terms of aerodynamics. There’s absolutely nothing wrong with your ears. 🙂
Yeah, I’m very happy with the A12x25s! Skimming the Noctua website also reminded me that based on their description, what I heard may have been that “beat frequency”. They of course took care of that with the RPM offset in their 2x package.
Using different motors to achieve different RPMs under the same PWM signal is indeed a clever trick to reduce noise while maintaining strong airflow and cooling performance. 🙂