Question Fan Spec Comparisons Useless?

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
With my latest build I'm currently looking at fans, deciding what to buy. There's an aesthetic element involved but I'm most interested in performance.

I'm actually making a comparison chart of all potential 120mm and 140mm fans.

Except for Akasa few fan suppliers seem to provide noise level dB(A) figures for the range of speeds the fan supports. The figures all appear to be single figure dB(A) and CFM ratings at max RPM.

In my primary PC I have Akasa Viper 120mm CPU cooler and case fans capable of 1900 RPM, used along with a couple of 140mm (1600 RPM) case ones. I have the BIOS set to use 75% PWM on all of them and even under heavy load in the height of summer, it rarely reports four figure speeds. Even when it does they're kicking in temporarily to get the temperature down and go nowhere near their 1900 RPM max speed.

So the question: is it fair to judge a 1900 RPM fan like the Akasa Viper 120mm whose published figure is 28.9dB(A) at max speed when it is being used more typically at half that speed?

At the slowest speed supported, 600 RPM the figure quoted is 6.9dB(A).

I realise the relationship between fan speed and noise level is not linear, the figures for the Viper show that clearly but let's assume it is.

Averaging the Akasa Viper figures out using the lowest and highest numbers at 1000 RPM around 44 CFM/13dB(A) as the more typical airflow/noise level.

If you look at the Noctua NF-F12 with its stated 1500 RPM max/55 CFM/22.4dB(A) and do a quick, again linear, comparison with the Akasa Viper at 1500 RPM the Akasa appears significantly better at 66 CFM/17.68dB(A).

The better spec Noctua NF-A12 figures using the same criteria are 45CFM/16.95dB(A) ie. near enough the same noise level but 30% less airflow.

At 1000 RPM the NF-A12 beats the Akasa Viper on noise at 11dB(A) so it is quieter but again the, roughly calculated, airflow at that speed is 30 CFM, 33% less.

Of course this all relies on manufacturers' figures being trustworthy. If you accept they are: surely just comparing fan dB(A) and CFM at max speed is not like for like and therefore useless to potential customers. Wouldn't both max and min CFM/dB(A) figures be more helpful? Why not a mid-range figure like 1000 RPM (for 120mm fans) too?
 
Last edited:

Tech Junky

Senior member
Jan 27, 2022
793
256
96
I've tried a half dozen different MFG's fans at this point and it's mostly just marketing and prestige when it comes down to their claims or recommendations. I've done the Noctua and I'm sure they have their place somewhere int he ecosystem but, they didn't perform any better or worse tan a cheaper option. The biggest difference comes between static 3-pin and pwm 4-pin configurations.

3-pin options run at constant RPM as set by the MFG
4-pin varies from a min - max based on the spec's

Smaller the diameter the louder they tend to be. 80mm fans are louder for a couple of reasons. The space to move air through is more restricted and the blades run faster causing more noise. 120's are efficient in a tapered case that doesn't have room for 140's.

The angle of the blades on the fan also make a difference in how they perform and the noise they make. The more aggressive / pressure sensitive applications attempt to force more air through a concentrated area.

Also, looking at the bearings for endurance is a factor to consider as well. Not so much the noise side but being able to install and not have to mess with for several years.


https://www.arctic.de/us/P12-PWM-PST/ACFAN00120A - can be $35 for 5-pack on Amazon
Fan Speed:200 - 1800 rpm
Airflow:56.3 CFM/ 95.7 m³/h
Static Pressure:2.20 mm H₂O
Noise Level:0.3 Sone

https://www.arctic.de/us/P14-PWM-PST/ACFAN00125A - slightly more for 5-pack on Amazon
Fan Speed:200 - 1700 rpm
Airflow:72.8 CFM/ 123.76 m³/h
Static Pressure:2.40 mm H₂O
Noise Level:0.3 Sone

fan1: 702 RPM (min = 0 RPM)
fan2: 804 RPM (min = 0 RPM)
fan4: 723 RPM (min = 0 RPM)
fan7: 695 RPM (min = 0 RPM)

They have other variations though to fit most applications depending on preference.

I was messing around with them while powered on and clipped one of the blades on one and sent them an e-mail and they shipped a replacement immediately no questions asked.
 
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I've actually been put off Arctic fans by user reviews on Amazon (UK). Too many reports of annoying sounds coming from some of them and suggestions the manufacturing consistency is not as good as it could be.

I do want to go for all 4-pin PWM control fans.

When you're probably going to be buying 6 or 7 (2 x front intake, 1x bottom intake, 2 x CPU Cooler, 1 x rear extract and 1 or 2 x top extract) which will cost in total between £90 - £175 ($115 - $225) (depending on make/model) getting a set of matching 'good' ones is important.

YT video reviews have and and have not helped.

I was gravitating towards the Noctua NF-P12 Redux 1700 RPM and/or P14s 1500 RPM Redux fans. Inoffensive basic grey/dark grey colour which could be colour customised anti-vibration corners (extra cost of course).

But the YT video I found where they tested the NF-P12 Redux against 3 other 120mm Noctua fans at various speeds including 1000 RPM was telling.

The NF-A12 clearly beat all the other for quietness at 1000 RPM but the figures (60.1 CFM/2000 RPM) suggest it is not shifting as much air in comparison to the NF-F12 (55 CFM/1500 RPM). So it should be quieter.

The YT videos also made me aware of the sound differences too - demonstrating that a low dB(A) figure is not the bottom line. The pitch and regularity of the sound is an underappreciated factor involved in the audibility of fan noise both singularly and in conjunction with others.

There is also the matter of the real audio profile of a particular fan too.

Other YT comparison tests showing, as I'd said was probably the case, that the dB(A) is not in a linear relationship to the fan speed. It varies with fans; some do have a more linear relationship than others but many become noticeably noisier above a certain RPM. With others, likely those with a short RPM range, the measured dB(A) change varies much less at all speeds.

Quite frankly if I had the choice I'd go with what I know, the Akasa Vipers. It seems to be the most rational solution: proven reliable, OK price point, good performance and good/acceptable acoustics.

Unfortunately the Vipers only come in the bright yellow at 140mm (black, yellow, white in 120mm) and even those seem in short supply. I really wanted to go for a colour scheme less 'dramatic' this time. Damned pity - all black or all white in both sizes and I wouldn't have bothered starting this thread.

The nearest equivalent are Akasa Apache available in black at 140mm and 120mm but their performance is notably inferior. The new Akasa Alucia fans too, despite looking good and ideal for my original concept of blue/black aesthetics.

The Alucia 140mm ones are rated very noisy at 38dB(A)/1800 RPM even if that figure is adjusted for their higher top speed. Both the 140mm Viper's 26dB(A)/1600 RPM and 140mm Apache 22.2dB(A)/1300 RPM clearly trounce it in all performance respects.

Having just shelled out for the case I can not afford the fans this month so I'm going to keep on researching and hopefully get some more recommendations here too.
 
Last edited:

Tech Junky

Senior member
Jan 27, 2022
793
256
96
Yeah, when you're spending $100 on all the fans it makes it a bit more serious to dig into them a bit more than most people would buying 2-3 of them. Arctic fits the sweet spot for me with the 5-packs being $8-9 /fan and the PST feature keeps the wiring clean by daisy chaining them together.

Warranty / replacement has been a breeze to get them within a week of contacting them.

If they're making funny noises the person reviewing them probably did something. When I was playing with airflow and repositioning them while running one of the fins on one knocked on the edge of the air cooler and cracked near the hob of the fan and even though it was a hairline crack it made a world of difference in the noise profile.

RPM's / DBA make a difference at 100% but you have to have a pretty crappy case w/ no air flow to hit those speeds. Think more industrial metal sealed case with an exhaust port to keep dust / water out. Any decent consumer case should have adequate ventilation with a few fans n the right position.

You could get some cheap $5 fans and they could be the best until they die or you could spend $25/fan and they're just pure junk, don't move air, and make tons of noise. If you keep the CPU cool then they won't ramp up as the calculations are always looking at the CPU temp to determine the speed they need to run at.

I took my cheap $45 air tower and put 2 Arctic 120's on it, skipped the one that came in the box, and mounted it to the CPU with a graphite pad. Most of the time it's sitting at 75-77F which is slightly higher than room temp. When it's cooking something like a microwave though it only goes to about ~130F / ~53C. Even at those temps I don't notice the fan noise at all and they don't need to run for very long to bring the temps back down. I have 3x140's in he front, 2x140's on top, 1x140 on the back and the 2x120's on the cooler. I have room for more but, haven't needed to add them at this point and it's been 6 months since building it. On prior builds I would have repositioned things several times by now but, that case was more compact and only held 120's in which I had ~11 of them installed between the case / CPU. Somewhat overkill but, redundancy built in Temps stayed down and never heard it when they ramped up.

Noctua fans are ugly but, then again if you don't have a window on the side they don't matter. They have a name for themselves and plenty of fan boys for the brand. Akasa might as well. I haven't heard of them until now.


this is for "bulk" packs / 4-pin PWM and there's not really a huge amount to choose from. Of course PCP doesn't list 100% of what's on the market either. It's a good place to start though to get a feel for price / availability. Buying 4-6 fans per package makes for a better investment per unit. going to the 3pack options bumps them over $10/fan and introduces the flaky RGB junk.
 
  • Like
Reactions: Mantrid-Drone

Paperdoc

Platinum Member
Aug 17, 2006
2,106
195
106
I agree with OP - the published noise figures are usually limited to max speed only and it would be nice to have more info.

Small point maybe (since OP plans to get only PWM fans) on speed control. Speed CAN be controlled on BOTH 3- pin and 4-pin fans, and the means to do that usually are built into mobo headers now. HOW to control a fan's speed is the issue. One can control the speed of a 3-pin fan ONLY by varying the VOLTAGE fed to it on Pin #2 of the connector. It can range from 12 VDC for full speed down to about 5 VDC for min speed without stalling. A new 4-pin fan is controlled differently. It is fed a constant 12 VDC on Pin #2, and then the new PWM control signal on Pin #4. This fan type has a new component inside its case - a chip that uses the PWM signal to modify the flow of current from that DC power supply on Pin #2 through the windings to achieve speed control. It IS possible to control such a fan's speed using the older method of varying Voltage supplied, but not ideal from a technical perspective. The difference in design is reflected in the labels used in mobo fan header configuration settings: Voltage Control Mode (aka DC Mode) for 3-pin fans, and PWM Mode for 4-pin fans. You set the header according to the fan type connected. Since their designs are different electrically, you should NOT mix the two types on one header. NOTE that almost all fan HUBS are designed only to distribute to their fans the electrical signals from a 4-pin PWM fan system, so you can NOT control the speed of a 3-pin fan using a common HUB.
 
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I've been using PWM fans/PWM control since I finished building my first desktop back in 2016. I actually decided on using PWM fans about a year before as they sounded like a superior method of controlling fans semi-automatically

I've never had any problems with them. I have them set at 75% in the Gigabyte MB settings and they keep my temperatures, in summer, down well below 50°C even when the CPU is working reasonably hard. Idling/low stress I get between 30°C and 45°C, depending on the ambient temperature, all year round.

I tried swapping the MB setting to 'Silent' mode recently and whilst it was indeed quieter I did see the temperatures go over 50°C much more often than usual so I reset it back.

I'm gathering info on a whole slew of fans just so I can compare them. It is confirming my suspicions that most of the 'silent' (120mm) fans are only 'silent' because the rpm is capped, usually at <1500rpm. The specs for some makes are otherwise identical to those with a higher top speed and consequently higher dBA noise level rating too.

That makes raises the question about the necessity of many of those 'silent' fan versions if fans with a wider rpm range may be just as quiet if used at the same rpm ie. <1500rpm.

One other thing: some fan manufacturers' specs are being misquoted by even specialist PC sellers and a reason for that is the manufacturers web site info is sometimes confusing (noise level rating in particular), occasionally even wrong or misdirect links for the specs of a particular fans to a different model.

I've had to recheck to confirm the specs of Arctic F12/F14 and P12 /P14 PWM PST and the CO (Continuous Use) versions using multiple sources because of those sorts of problems.

The most reliable are Ebay sellers who include shots of both front and back of the original fan box. Even then that does not always include all the relevant info a buyer might be interested in, particularly static air pressure ratings (Arctic again).
 
Last edited:

Paperdoc

Platinum Member
Aug 17, 2006
2,106
195
106
I agree about the need (or lack thereof) for "low speed" fans, other than as a marketing gimmick. We should recognize that the SPEED of the fan is not of primary importance. What IS important is the AIR FLOW it delivers. Of course there IS a relationship for any one particular fan design. Then realize that you can use a mobo's automatic fan speed control system. It actually is a TEMPERATURE control system. Its sole aim is to keep the temperature as measured by a relevant sensor (inside the CPU chip for CPU cooling, or on the mobo for case ventilation) at a defined target. In today's common practice, that really is a "fan curve" of what fan speed (as a % of full power, whatever the fan) to run for what measured temperature. (Even that curve can be changed by the user on many mobos.) The means of doing temperature control is to manipulate the fan speed. Although a fan header measures fan speed by a signal sent back by the fan and monitors that for possible failure, it does NOT use that speed for its control work at all. So IF you use that system it will change the fan speed signal to whatever it takes to achieve the TEMPERATURE its is supposed to target. If you have installed a "low speed" fan which merely limits the max speed, the system will run it at the SAME speed as a comparable "normal" fan until it gets to its max speed, and then the "low speed" fan simply cannot do anything more. A "regular" fan model would still have reserve cooling capacity for high workloads. Since the low- and regular- speed models from any maker tend to be identical except for the motor speed limit, the actual noise generated by running them to achieve equal air flow rates is also identical, so there's no advantage.

Using the automatic scheme does have small advantages over a fixed speed: if the cooling needs are large due to high workloads for some time, the fan will speed up to meet that need; if the cooling needs are low, the fan will slow down and create marginally less noise. None of this requires attention or intervention by the user.
 
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
The corollary to that must surely be the lowest supported speed might actually be more important than the highest speed when considering noise levels.

If a fan has a lower limit of say 650rpm (assuming PWM control does not bypass that - does it?) it is going to be that lowest speed which determines how quiet/noisy it is, particularly if the case/CPU is well cooled.

If all the fans only have a lower limit of 650rpm whatever temperature they are maintaining that is the lowest noise level they can ever be even though they could be maintaining an acceptable <50°C temperature at half that speed or less.

I had not considered Cooler Master fans before as the ones I'd come across in the Cooler Master case I have were of lower spec than the fans I eventually used in that build. But in my recent researches I see that their 'top end' Masterfan Pro range fans have an interesting S/Q/P switch option ie. Silent/Quiet/Performance settings which effectively allow you to change the fan spec.

The specification sheets shows lowest speed supported is rather high at 650rpm (and why I used it as an example above) in all cases; it is only the upper speed which changes with use of the S/Q/P switch.

Anyone have experience with these? Opinions?
 

Paperdoc

Platinum Member
Aug 17, 2006
2,106
195
106
A fan with those three-speed switches is exactly the same as different fan models with different max speeds. Only the MAX speed is impacted. That feature is really only suited for a fan that can NOT be plugged into some automatic fan speed control system and must be powered by a fixed 12 VDC line from the PSU. The feature would allow manual setting of a fixed fan speed less than max.

A fan's spec for min speed is determined by the fan. What the actual min speed of the fan is can be set by customizing (if necessary) the "fan curve" used by the header. Then it is up to YOU in making that setting to ensure that the fan actually never does STALL by trying to run it too slow. Also be aware that, as a fan ages and its bearing wear, that min speed without stalling MAY actually increase.

The impact of PWM vs. Voltage Control Mode is really in what the stall speed is. It is claimed (from a technical perspective) that the real non-stalling minimum speed a fan can run is LESS with PWM control because each tiny brief current surge in the motor gives it a tiny "kick" that prevents stalling at a low speed. But still even in that mode of control, the fan can stall if its electrical feed is set too low. Now, the NORMAL operation of an automatic fan control header is to monitor the fan's speed signal for NO signal, indicating the fan has stalled. In response to such an event, the system will send the fan a full speed signal and check whether it re-starts. If it does not, you get a warning message of fan failure. If it does re-start, the fan's signal is re-set to whatever it was before. BUT that was the signal that allowed it to stall in the first place, so this may become a repeat event.

In selecting a fan there are several factors to consider. Among them are noise and max air flow rate, which are correlated but we want one up and one down. So you want to find a fan with slightly MORE air flow capability than your system ever will need at max workload, and lowest noise at ANY speed, including at minimum speed. It is quite unfortunate, as you said at the start, that the noise at MIN speed is never specifed. We are all left to guess in comparing two fans whether the one with lower noise at max speed also will yield lower noise at ALL speeds.
 
Last edited:
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I think me saying earlier manufacturers 'never' quote the lowest speed noise rating was a bit premature.

Today I was looking at Scythe fans and getting almost as annoyed as I was with the Arctic web site and the quoted specs for various Scythe fan versions sold elsewhere too. But they do often quote the lowest fan speed dB(A) ratings too and for each of the fans when they make a set of them with different speed range specs.

The reason for my annoyance was principally because they do often have three or even four versions available, some PWM, some not and often that mix of fan speed ranges for each fan name type. The Scythe Slip Stream range is further confused by the fact there are older versions of both PWM and non-PWM fans still available from new

I've found sellers wrongly quoting the specs for the non-PWM fans in a PWM fan listing and quote the mid-range 'Regular' stats when the listing is for the 'Quiet', capped upper speed model.

There's even confusion over the term "4-pin" which suggests the listing is for a PWM fan but it is not. Scythe helpfully supply a 3-pin fan (F) to 4-pin Molex connector with several models both PWM and non-PWM.

So if you're buying a Scythe fan you really do need to make sure of the precise model ID number of the fan being sold.
 

aigomorla

Cases and Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
19,497
1,824
126
You can only follow one tester when he uses the same tools on his bed.

Too many things can invalidate the testing, which is why GamersNexus dropped a whole bunch of money on testing equiptment.
(and i mean like 6 figures just to test fans)

Also i will not trust a cheap independant chinese report.
I have had many many off reports, like the cheap 5 dollar fans you get on amazon, to ones which under rated its performance.
(yate loons were actually better then the reports originally stated)

So unless the report is in short documented by Nidac / Delta / SanAce / (any other vendor dell / supermicro) uses , take all reports with a heavy grain of salt.
 

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I tend to believe the bigger name fan manufacturers' testing results because, if they're exaggerating, manipulating or massaging the figures to make their products' performance seem better than it is........................they're going to get found out.

As you say: avoid the cheap, obscure brand fans you find on Amazon and Ebay like the plague. But you do have to at least go some way down the road and trust the likes of Noctua, Corsair, Cooler Master etc quoted fan specs are not bogus. They should at least be using the same criteria to measure their own fans' performance.

So, if correct, you can identify the 'best' fan from each maker at the very least.
 

mindless1

Diamond Member
Aug 11, 2001
6,968
973
126
You guys must be running recording studios or 200W+ intel CPUs (I mean besides video card fan noise, which depends more on the GPU vs heatsink design)... I can pick just about any random fan, as long as it's not imbalanced, and achieve low *enough* noise if on a fan controller, whether PWM or the older linear control for 3 pin fans.

However, I'm not so picky about precision in ramping up fan speed to keep temp as constant as possible. Before there were motherboards that could control this, I'd just target a fan speed only high enough to prevent overheating.

I looked through some archived files and found something I made about 30-40 years ago, was just a series of diodes that could be switched in or out of the supply circuit. Today it is hilarious to me, that I decided to coat the diodes in molten plastic. I don't know why, it was a long time ago! Well.. at the time I had limited income and all materials were scavenged from old equipment or scraps of whatever. Nobody made fan controls for consumer applications and I didn't know about electronics supply houses like Digikey and there was nobody selling on the internet. Radio Shack was pretty much all I knew.

Regardless, I'd be fine using a series of diodes to keep fan noise at bay today, or what I did more often in preceeding years, hook a fan up to a rheostat limiter dongle I made, then once I dialed in the speed I wanted, just put a 2W resistor of similar ohm value in series on the fan lead. That won't allow for AS low a fan speed, but I'm just not that picky if I can get it below, say 600RPM or so if that's all the cooling that's needed.

I guess I just don't see the holy grail of working towards a very slight decrease in noise. There is not that much difference between various fans (of same dimensions) when they are running at the same, very low RPM.

If you live in the country, animal sounds (birds mostly) exceed the noise of fans, or in the city, vehicles, or wind, or other things in your home. I cannot hear my computer fans at all, and have never paid extra trying to reduce noise rather than just achieve good lifespan.

Granted, the computer is under my desk and that helps a lot. Why take up desk space anyway?

fan speed controller.jpg
 
Last edited:
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
Well I can hear my fans or more correctly the air being pushed out of the Corsair Carbide case I use which has no sound deadening at all. It is only 2 feet away from my left ear but I've nowhere further away to locate it that is going to make much difference.

I just want my new build to be as good at cooling running as that: PWM set to 75% and giving <40°C most of the year with the fans going nowhere near full speed. But also with enough in reserve so that when I can afford to fit a decent GPU it maintains that sort of performance.

However I'd like it noticeably quieter too - hence the Fractal R2 Define case I've bought which has plenty of sound insulation amongst its other other features. However as with my current desktop it is the fans I hear most (the two HDDs storage are set to spin down after 5 mins) the fan choice dB(A) to CFM ratios are quite important to me.

Currently researching Phanteks - not ones I'd considered (or heard of TBH) before but several of their models, the F120T30 in particular, seem rather good on paper although appallingly expensive.

I had to search hard to find out the bearing type and SAP figures as its not included in the Phanteks web site spec sheets.
 

Tech Junky

Senior member
Jan 27, 2022
793
256
96
I've had a couple of Phanteks cases and they're nice but, Fractal is better. I've had a couple of FD's as well and they're easier to work in and roomier for drives / fans. FDs you can typically fit in 8-10 fans and PH's you can maybe do 5-7.

140's will be quieter when they have to ramp up beyond 50% but, the right air cooler on the CPU makes a difference too.

Now that I looked up the part # I see you're talking about fans and not the case itself. I haven't used PH fans before as they didn't come on my radar for performance. For the price you can't beat the Arctic's I mentioned before. If you want to spend money though there's plenty of pricey fan options out there.

I typically set a cap of $100 for a full set of case fans + cooler fans though. I've done the Noctua setup before and they weren't all that impressive for twice the cost of Arctic.

Go on pcpartpicker.com and look through the reviews and options as you can sort by decibels and narrow it down.


If you want quiet and moving a lot of air look at the CFM values as well.
 
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
The thing is that the whole idea of this research and comparisons of fans that I had not have considered before is to find something better, at least on paper, than the Akasa Vipers I've used in my previous two builds.

The problem, without even needing to put the 'best' from each manufacturer up against each other in a comparison table, is that the Akasa Viper 120mm and 140mm already look to be the ones in the sweet spot with the best balance of CFM, SAP and dB(A) figures..

Add in a reasonable price and my experience with them ie. after 6 years use they're still all good and they're going to be hard to beat.

If that is finally the case I'll have to use them again even though I wanted to get away from their bright yellowness this time. They do a 120mm in black and a three pack of 120mm white ones I'd not come across before.

However in 140mm, what I'd prefer to use for this new build's case, they only do the Vipers in yellow, a standard 140mm fan mount and a 140mm with 120mm mount. Why no all black or all white? How many PC builders want a bright yellow 140mm fan? It seems to me a mad marketing decision not to do their best 140mm in all black too.

Look like I might be doing a half blue, half yellow lightly colour themed build. Even though not what I wanted the colour combo is strangely appropriate at this time.
 

Tech Junky

Senior member
Jan 27, 2022
793
256
96
It's called branding! I had some Cougar fans in one case and they were Orange. They do silly crap to brand things to stand out from other companies. Look at the baby turd brown Noctua uses.

If they're quiet but not pretty then you have to decide if it's worth the hassle of swapping them out and selling the Viper's on ebay to get some cash out of them. The easiest solution is get a case where you don't see them through a panel ;)

Stats and specs are always going to vary between models. I look for something with a wide range from 200rpm - 2000rpm and then sift through the reviews to see what people are saying. Price is usually a concern though as it's something you don't see returns on other than not frying things due to high temps.

Getting a 3 or 5 pack tends to be more economical though but, on avg $10/fan is where the sweet spot is for price vs performance. / durability.
 
  • Like
Reactions: Mantrid-Drone

mindless1

Diamond Member
Aug 11, 2001
6,968
973
126
^ Your build, your money to spend on what you want.

To me it would depend a lot on what kind of deals you can get on a particular model. I have several fans that normally cost over $20 a piece, but I only bought them because they were on sale at the time for $2-8 a piece.

I can't ever recall spending even $10 ea. for a fan, always buying brands like Panaflow, NMB, Sunon, Papst, Delta, etc. I just don't think there is enough difference to spend $10 for a fan, if a typical case wall or heatsink mounting. Sometimes on a PSU, the flow direction matters a bit more. A few CFM or decibels here or there though, is very little to fuss over when we're talking low digits.

Heh, I remember this one time, I bought an 80 pack of NMB 80mm fans for... maybe $30. It might've been an anandtech hot deal or something. This was around 20 years ago, still running some of those today.
 
Last edited:
  • Like
Reactions: Mantrid-Drone

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I sort of agree with that but the problem with cheap components tends to be that they will eventually go wrong and need replacing.

I just want to fit them and forget but, with fans and to a lesser extent HDDs, you can hear them and if fans go wrong, unless you're a monitoring pedant or have MB BIOS options to warn you there's a possibility you're going to be cooking your expensive CPU/MB and all for the far smaller amount you've 'saved' using cheap fans.

I have been a bit surprised in my research to find PC fans with only a one year warranty despite their quoted longevity figures. If the manufacturer doesn't think they're going to last beyond a year's use that's a red flag to me.
 
Last edited:

mindless1

Diamond Member
Aug 11, 2001
6,968
973
126
I sort of agree with that but the problem with cheap components tends to be that they will eventually go wrong and need replacing.

I just want to fit them and forget but, with fans and to a lesser extent HDDs, you can hear them and if fans go wrong, unless you're a monitoring pedant or have MB BIOS options to warn you there's a possibility you're going to be cooking your expensive CPU/MB and all for the far smaller amount you've 'saved' using cheap fans.

I have been a bit surprised in my research to find PC fans with only a one year warranty despite their quoted longevity figures. If the manufacturer doesn't think they're going to last beyond a year's use that's a red flag to me.
Note the brands I listed, they were high quality fans, just found at a deep discount. Warranty doesn't mean a lot, IMO, rather the type of sleeve bearing and mounting orientation, or using dual ball bearing and at least 20mm thick... as long as it's a major fan manufacturer, brand.

I haven't had a fan failure in a very long time and back then it wasn't a fan I picked, rather some off brand junk included with the product whether a heatsink, case, PSU or whatever.
 

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
Manufacturers now seem to favour rifle bearings or some other variation on the FD bearing or magnetic levitation and sometimes even a mix of bearing technolgies in preference to sleeve and ball bearings.

Most of the exclusively PWM fans I've been looking at don't use sleeve bearings and as for ball bearings the only make that uses them for that type of fan is Arctic. Even then they only use them in their 120mm/140mm PWM PST CO (Continuous Operation) 'F' and 'P' range.
 

mindless1

Diamond Member
Aug 11, 2001
6,968
973
126
^ Rifle, FD, Magnetic, etc. are all sleeve bearings. There's only 3 real options for a 20mm or thicker fan. Sleeve bearing, dual ball bearing, or one ball and one sleeve which is the worst of both worlds and hardly ever done except on smaller, thinner fans like found on some crappy $5 Coolermaster or generic heatsink.

Why do they market them like this? Because it's cheaper than the time proven solution for a sleeve bearing fan which is put in a reservoir with some felt/material to hold an oil reserve.

Any fan intended to last over 10 years that's not dual ball (or needle, etc) bearing will have this, or any motor really whether it be the HVAC fan in an automobile or home, or the motor spinning a treadmill belt.
 
Last edited:

Mantrid-Drone

Senior member
Mar 15, 2014
278
18
81
I think you'll find that usually when they refer to a sleeve bearing they're talking about just that: a simple machined bearing which usually does not have sealed in lubrication or made of/lined with a self-lubricating material either.

As I understand it whilst rifle bearings are, of course, sleeve bearings too they're a step up from that basic form as they do have 'rifle like' lubrication channels. They're cited by some sources as an unpatented alternative to FDB.

With FDB or other similar named lubrication systems the lubrication is, apparently, better sealed in and has some sort of design feature that, unlike the rifle bearing ensures it doesn't pool if the fan is used horizontally.

It is interesting, sort of, that one fan manufacturers I looked at used rifle bearings in their earlier fan models but then swapped to the patented FDB system for their more recent fans.

If they're willing to pay the extra for that it does suggested the rifle bearing is inferior.

FDB fans are always cited as having the greatest longevity with their current iterations claiming 100,000hrs or more compared to rifle bearing ones which top out at 50,000hrs. That is very likely the reason the manufacturer concerned swapped to the more costly design.
 
Last edited:

mindless1

Diamond Member
Aug 11, 2001
6,968
973
126
^ It is just marketing. Lubrication channels can only help if there is a sufficient reservoir to pump the lube from. If there is, there doesn't need to be a channel, as it is capillary anyway.

It might be fair to assume that if their marketing dept went the extra mile to market the fancy bearing name, that they also went the extra mile to provide more of a lube reservoir, but this is not necessarily true, and the longest lasting fans for more valuable goods, don't fall into this game, instead they just build in the oil reservoir or use dual ball bearings.

It is a trade off. In order to get a higher airflow to noise ratio, you have to pick a lower quality fan. That doesn't mean the fan is "low" quality, just nowhere near as robust as those meant for industrial applications and don't have to use marketing buzzwords that are only popular in the PC builder scene.
 
  • Like
Reactions: Mantrid-Drone

ASK THE COMMUNITY