Alphacool 120 mm Apex Stealth Metal Fan Reality Check – Exact error and root cause analysis and reworking including perspectives

1 - Introduction 2 - Measurements with different results 3 - Measurement protocols 4 - Variance comparison 5 - Summary and conclusion

The measurements of the Apex Stealth Metal Fan

Here’s what I did: I tested the Apex under various conditions. I used a Noctua as a comparison in order to maintain plausibility. I only used the 60 mm radiator, as it was already clear where the journey would take me. There has already been a lot of discussion about the measurements in the box and there are various opinions on the measurement setup. I’ll say something about that later. For now, however, we’ll focus on this measurement setup. Let’s take another quick look at it. Here is a sketch to help you understand it better:

The air is drawn in via the preflow chamber through an inlet opening with a diameter (D) of 14 mm. As a 120 mm fan is being tested here, I have installed a reducer at the inlet. This means that the diameter is only 12 cm. The air now flows through the fan and is conveyed to the rear to the anemometer, where I measure the volume flow in m³/h. The outlet opening is D 14 cm. You have to remember this, as this variable will play an essential role in relation to the measuring software.

I first installed the Apex fan as it should be (see picture above). Namely in the preflow chamber on the left in front of the board in the middle. Then I took measurements with and without the flow straightener (I’ll call it a honeycomb later on). To do this, I parameterized the exhaust air measurement in the Testo measurement software according to the conditions. The diameter at the outlet of the box is 14 cm and 100 percent of the area is available. I also left the correction factor at 1.00. This is because the ambient conditions with regard to atmospheric pressure were 1010 hPa. You can see the details in the logs.

Example of a measurement protocol (you can see the software parameters at the top):

Install the fan in the box, set the speed to the maximum approx. 3035 rpm. Then close the lid and wait 30 seconds to allow the air to calm down. I then carried out the measurements for approx. one minute each. This results in around 61 measurements, which the software saves automatically. This results in an average value as well as the min. and max. values of the respective fan. The measurements with and without honeycomb have already shown clear differences. However, this can be explained technically.

This is because the rectification of the air by means of the honeycomb directs the air flow to the anemometer in a more targeted manner. It can also be assumed that the honeycomb already results in a higher flow velocity – due to the smaller holes. The fact that the fan speed also drops to approx. 2900 rpm indicates a certain counterpressure and makes this assumption seem plausible.

Simplified explanation: According to Bernoulli, a reduction in diameter at the same pressure results in the volume flow behind the constriction becoming faster, but the pressure drops. Due to the conservation of mass or volume, what goes in must also come out at the end. This generally valid statement is also the basis here. But here’s the problem: air is compressible! You need to know this, which means that Bernoulli’s equation as such cannot be applied one-to-one. This is because we normally work with incompressible fluids. I don’t want to drift any further into physics at this point. The fact is that the air that goes in also comes out at the end. Depending on the structure, the pressure and flow velocity change in such a way that the volume is (approximately) maintained at the end. We will not consider the compressibility of the air at this point. I will also ignore the temperature factor here. Because this can also play a role. Please bear with me, otherwise it will only get more confusing.

You can also install the fan on the other side – to the right of the board in the outlet chamber. I have also looked into this, as it also changes the measured values. The mere fact that this is possible means that the fans are treated differently. In other words, I could make one fan look better than the other. I can also do the same in conjunction with a radiator. Install on the left or right, with or without honeycomb, push or pull, etc. This results in various variations which I have named as follows:

The Noctua NF-F12 without honeycomb is already better than the Apex with honeycomb. The Noctua achieves a maximum value of 182.3 m³/h with the honeycomb installed. According to Noctua’s specifications, the fan should achieve 186.7 m³/h. This seems quite plausible. The situation is different with the Apex. According to the article on the fan revolution, it was measured at 198.6 m³/h at approx. 2950 rpm. Whether this is the peak value or the average? Only the former tester himself knows.

With the adjustment in the software to the outlet diameter D16 cm, I get an average value of 196.3 and thus very close to the fabulous values. Making the software believe that it is a larger diameter than 14 cm means: de facto a higher flow velocity – which does not actually exist. This is because the software cannot see that the D16 is not one at all. The same also applies to the measurements on the 60 mm radiator. According to the article at the time, the Apex is supposed to achieve a flow rate of 190.4 m³/h. With an adjustment in the software to D21, I get an average value of 189.9 m³/h. Otherwise, I don’t even come close to the measurement results from the article.

The fact is, the Apex Metal Fan clearly loses airflow on a 60 mm radiator. At 3000 rpm a whole 44.3 m³/h and at 2200 rpm even 45.5 m³/h, if you take the measurements Apex left with honeycomb against the Apex left push config with honeycomb. If you were to compare these values in the database with the Be Quiet! Silent Wings Pro 4, the Apex would be right around the same level. However, the question remains as to whether the values from the Silent Wings Pro 4 are correct. I can’t answer that here and now.

I also looked at the static pressure measurements (see picture above). Unfortunately, I despaired as to how it was even possible for the former tester to measure such values. Do-it-yourselfer vs. craftsman and a lot of trust in God – from my point of view, this is not at all feasible for the following reason:

You can’t position the fan over the opening in such a way that the area is completely covered and sealed. This means that the pressure can escape to the rear through the narrow openings. At least I wasn’t able to do this with the equipment that the former tester gave me. Interestingly, I can measure a higher static pressure with a radiator than without. Why is that? The radiator seals the outlet opening! In addition, the cable from the aero to the fan was too short to connect it directly.

So I needed another extension to measure in the push installation. As a result, the fan speed was only 2200 rpm at 100 percent. You don’t have to understand this, but unfortunately it was not possible to do otherwise in the short time available. But here, too, you can conclude from the values that the values in the database (or in the article) are to be doubted. Under the circumstances, I arrive at a maximum of 0.29 hPa, which corresponds to approx. 2.9 mmH2O. Unfortunately, I can’t tell you how the tester got to over 0.6 hPa here. The tester has also been unavailable since Roman’s last video. Coincidence?

I didn’t take any measurements in terms of volume vs. speed. In the end, I didn’t have the software. I’ll have to make up for that. The fact is, the fan is quiet, no question about it. Compared to the Noctua NF-F12 industrialPCC-3000 PWM, it is even significantly quieter. Using my dB (A) measuring tool, I took a hands-free measurement from a distance of approx. one meter with the case open. Noctua NF-F12 at 3000 rpm approx. 45 db (A) and the Apex at 3000 rpm approx. 40 dB (A). This is a clearly perceptible difference and the subjective sensation of the Apex is also much more pleasant. The Noctua at 3000 rpm really blows your ears off. So much for my purely subjective opinion on the subject of volume. Don’t worry, everything will be measured again properly at a later date. Separation!

On the next page but one, we’ll take a look at my measurement system analysis. If you’re still interested, there are a few protocols for the Apex/Noctua on the next page.