Talk about an embarrassment of riches! In short order, Noctua banged out a bunch of PWM fans at the end of 2011. I thought you’d like to look at three of them. Here are the first two:
What’s that you say? These fans look just like Noctua’s old fans? Well of course they do. They are our faithful old friends, the NF-P14 and the NF-P12 with PWM circuits. Apparently, it took Noctua engineers years to figure out how to make PWM PCB’s the Noctua way, which is essentially perfect. I’ll tell you how well they did later when I test these fans.
So let’s look at the third fan, the NF-F12 PWM:
Those vanes are called stators. When a fan’s output is blown through properly designed stators, not only does it focus the airflow but it increases the static pressure. Noctua’s PR guy tells me I shouldn’t expect more CFM from this fan, or any better cooling if we put it on an NH-D14, their top cooler. The NF-F12 PWM is designed to push more air through constricted settings, like radiators. But for us air coolers, higher static pressure can overcome the restrictive effects of air filters. That means the NF-F12 PWM should be an excellent case fan.
Well, let’s start with the basics: the NF-P12 PWM and the NF-F12 PWM are being sold at retail. The NF-P14 PWM is not. It only comes with Noctua heatsinks, and you can’t buy it separately. So, you’ll see the NF-P14 PWM a little later in the article.
NF-P12 PWM Nine-Bladed Fan
The nine-bladed NF-P12 PWM comes with a few accessories:
- On the left you can see a PWM Y-cable. Both limbs have the ground, 12 V and PWM control lines, but only one has an RPM reporting line. Like the 3-wire that came with the original NH-D14, this Y-cable will not confuse the motherboard with two sets of RPM signals.
- Then comes the 30 cm (~1 ft) extension cable. This makes up for the fan’s cable only being 20 cm long (eight inches). Most of us don’t need a longer fan cable, but for those who do, Noctua has you covered.
- There are four “Vibration-Compensators.” They come with all Noctua fans, and you are supposed to use these to fasten the fan to a case. Let me know if you can make them work. I want to know how.
- There is a packet of four standard fan screws, not shown.
- In front of the fan is the Low Noise Adapter (L.N.A.), part NA-RC6. It will drop the fan speed from 1300 RPM to 900 RPM — and still lets the PWM signal through.
We’ll do some testing further on. But for now, we have the product page.
NF-F12 PWM Seven-Bladed Fan
Here is the Product Page.
- 30 cm PWM extension cable — check.
- Screws — check.
- Low Noise Adapter — Hey! This one is an NA-RC8. It only drops the fan from 1500 rpm to 1200 rpm. Hmm. On the D14 the RC8 dropped my P14 to 900 RPM. The RC6 dropped it to 700 RPM. So they do have different resistances.
- The Y-cable is the Noctua now-standard (and excellent) PWM Y-cable — check. Note that all the cables are beautifully sleeved.
- And finally the Vibration-Compensators.
First, I plugged them in to my PSU’s 12 Volt line to learn what their max speeds were. Then I poked a grounded line into each fan’s PWM channel. All three fans went to zero. This is unusual among PWM fans. Most go to some minimum speed. And then on most PWM fans you see a variety of duty-vs-rpm curves where the rpm increases as the “duty” increases (the term “duty” in PWM situations means the percent of time the PWM line is providing 5 Volts to the fan). These curves are often non-linear. We will see what the Noctua curves look like.
I plugged the fans into my new Zalman PWM controller (link). I could then provide approximately 25%, 50% and 75% PWM duty to each fan. To observe the speed of the fan, I sent the RPM line to a fan header on my motherboard and recorded the number shown by ET6, the motherboard utility that measures such things. I also recorded the sound pressure level with my tried and true SPL meter from 10cm directly in front of the fan. Subtract 20 dB and you have the equivalent sound pressure level of what you would hear one meter away.
|Noctua fans||PWM duty||Noctua|
|NF-F12 PWM RPM||-0-||363||757||1123||1470||1500|
|NF-F12 PWM SPL||-0-||<12 dBA||15 dBA||22 dBA||28.5 dBA||22.4 dB|
|NF-P12 PWM RPM||-0-||330||686||992||1271||1300|
|NF-P12 PWM SPL||-0-||<12 dBA||13 dBA||19 dBA||25 dBA||19.8 dB|
|NF-P14 PWM RPM||-0-||381||721||991||1225||1200|
|NF-P14 PWM SPL||-0-||<12 dBA||17 dBA||24 dBA||29.5 dBA||19.6 dB|
First of all, none of the fans clicked at any speed. That’s amazing. Mission One accomplished.
The next thing to note is that, within the broad limits available to test the RPM and the PWM percentage (“duty”), the speeds of these fans scaled linearly with the PWM duty. Again, that represents hard work on the part of the engineers working for Noctua.
Finally, they were quiet. Even the F12, spinning at 1470 RPM, was quiet. Bravo! Note that some of the SPL results could have been quieter than measured. Twelve dBA is the lower limit of the equipment and the environment. Also note that discrepancies between what manufacturers measure and what we measure are normal. More important is that the speeds and sound pressure levels for the PWM P12 and P14 are the same as for their non-PWM predecessors.
Now it’s time to saddle up and test these fans on the NH-D14. Now, where’s that NF-P14 PWM? Ah! Here it is:
Since the P14 PWM doesn’t have its own box, I put it in front of the P12’s box. They are similar fans, after all.
I have previously posted the particulars of my test rig. It is a fine test of high performance heatsinks and their fans. So how well did these do?
The plain P12 and P14 (stock setup) hold the hottest core of the CPU to a net temperature (CPU core temp minus ambient temp) of 51°C at a cost of 26 dBA, measured from the side. The PWM P12 + P14 hold the net temps to 50.6 °C at a cost of 26 dBA. These numbers are essentially identical, which is astonishing. Those PCB boards for the PWM fans produce the same results as the non-PWM fans. Getting those fans dialed in to work the same as the old fans took a lot of work.
Now, let’s have some fun. Let’s put an NF-F12 PWM on the front as push fan in place of the P12. The F12 + P14 PWM combination holds the net temps to 50.5 °C with an SPL of 26.5 dB, despite the fact that the F12 was running at 1477 RPM, more than 200 RPM faster than the P12. That was hard enough to believe that I ran the test again: 50.4 °C this time.
Here is the chart:
So, not only do the PWM fans essentially replicate the non-PWM fans, the F12 cools the same as the P12 on the D14. Is this coincidence or engineering? One thing I did note was that the air I normally feel spilling out the sides of the fin stacks of the D14 (and demonstrate with a tissue) was absent in the front fin stack when the F12 was mounted there. It appears that the F12’s airflow truly is focused, as advertised.
The NF-P14 PWM and the NF-P12 PWM are worthy successors to their non-PWM versions. And the advent of these PWM fans means that the last problem with the NH-P14 has been solved. We have also seen that the NF-F12 PWM is a fine fan in its own right. You water coolers can use this on your rads, and we air coolers will put it in our cases.
I won’t use the Noctua “Vibration-Compensators.” The Noctua pegs are a very tight fit on the F12. I was afraid to put one in because I wasn’t sure I could get it out again without breaking it or cutting it off. The main problem is the length of the pegs: they are way too short. Unless you are willing to grab the end of one of these with needle nosed pliers you will have trouble pulling them into position. It is hard to get your fingers into the fan’s corner to pull the peg through. Look at this:
That picture shows you where the peg will end up once it is in. You can see that grabbing enough of it to pull it to that position might be quite the task. What you need is a longer peg, like this one:
This peg is long enough that you can grab it on the other side of the fan from the case. You won’t have to try to get your fingers into some far corner and struggle to get the peg in. This is a small issue. Just ignore these “Vibration-Compensators.” When you buy the fan, get aftermarket pegs.
In conclusion, these are solid fans:
- They are quiet
- They do not click
- They perform well
- Their PWM/RPM curves appear to be linear
These fans include excellent accessories:
- Their power cables are just the right length
- If you need a longer cable, one is included
- They include a PWM Y-cable that is well-designed, and not excessively long
- All cables are handsomely sleeved
Thanks to Noctua for supplying these PWM fans.