The Arctic-Cooling PWM fans are in and they work!

Felger Carbon · Post by **Felger Carbon** » Tue Jul 03, 2007 4:41 pm

Just got the Arctic Cooling PWM fans in this afternoon. 2ea 92mm, 2ea 120mm. I've run one 120mm fan.

The fan itself is unremarkable. 166g, larger hub than usual but not nearly as large as the Panaflo 120s. Spec sez .13A, 37CFM, 400-1500RPM which would be at 30% and 100% PWM (you really should read the PWM spec - the latest I have is 1.3). CFM spec seems low (to me) for a 1500RPM fan.

For SpeedFan 4.32 to control the PWM speed, you have to make sure Quiet Fan is disabled in the bios. The speeds I got were: 100% 1534RPM, 80% 1251, 60% 916, 40% 631, 30% 524, 20% 494, 10% 488RPM. SpeedFan reported the RPM from 100% to 45% (note 2), no tach detected below 45%.

The fan is sitting on top of the PC case, 3" from the rear. At 85% I can hear it, 80% and below I can't from my normal sitting position at the keyboard about a meter away. There's also a 220mm fan turning 394RPM that's slightly closer, but not in line of sight (or line of ear) that I can't hear. This is at 4PM next to the front of the house in July, so it's not real quiet but not noisy either. A very quiet office environment.

Yes, the spec sez 400RPM at the low end and what I got was 524. This is absolutely typical behavior for every variable-speed fan I've ever seen. The specified minimum speed is a rumor. I suggest you factor this into any PWM plans you have.

At 30% and 524, I leaned over as far as I could and lifted the fan right up to my ear. All I heard was fan noises, including blade noise. I couldn't hear any bearing noise or any other strange sound. I did not repeat this at all the available speeds - the position was awkward, for one thing, and this is just a quick report.

The 4-pin fan connector is well keyed. There's a connector on my mobo that at a glance seems to be the one. Uh uh, it's an audio connector. The keying prevents you (me) from goofing. Plugs right into the correct connector, no problem.

The wiring is unusual and deserves special mention. A-C has provided a way to drive one PWM signal to 2 or 3 fans via a cable daisy-chain. The basic fan cable from the fan to the 4-pin female is 400mm, ~15". From there, 3 wires (no tach) go 2.5" to a male 4-pin connector, and a ground and tach signal extend from that a ways to a usual 3-pin fan connector.

Here's what you cannot do: plug fan #1 in to the mobo PWM header and place the fan 400mm away. Then plug fan #2 in and place it 400mm from fan #1 and 800mm from the mobo PWM header. Repeat: this will not work!

Here's what you can do: daisy-chain 2 or 3 fans. Each fan must be within ~400mm of the mobo PWM header, but any two fans can be up to 800mm away from each other.

For what it's worth, I cut the extra 4-pin plug and its wiring off the one fan I'm running without any hesitation. I don't plan to daisy-chain PWM fans.
----------------------------

A word or two about the PWM specification, which is an Intel document. The latest version I have is Rev 1.3, Sep 2005 "4-Wire Pulse Width Modulation (PWM) Controlled Fans". When I google "PWM spec" V1.2 pops up. There may be an even newer version than 1.3; I dunno. [I posted this URL of 1.3 in an earlier thread].

The specification provides three profiles of what happens when you try to run the fan below 30%. The first is that the fan continues running at the 30% speed, and that's what the A-C 120mm does.

The specification calls for a 2-second start pulse at the 30% setting. When I saw that I thought "big-time BS!". I'm happy to report that A-C kicks the fan for 2 seconds at the 100% setting (note 1). When specification writers have their heads mislocated, the specification shouldn't be followed!

Note 1: this was detected with Quiet Fan enabled so the fan setting is 70 or 80% upon computer boot. The 2-second 100% start pulse can then be heard.

Note 2: understand that the computer RPM report (either in the hardware monitor or SpeedFan is based on a <5second sample of the tach signal. At low RPMs plus or minus one tach pulse is a biggie. The RPM figures I reported were measured with a tachometer but with no appreciable stabilization period - I was looking for general speed, not exact speeds. The way to get accurate RPM figures with SpeedFan is to use the chart facility and then eyeball-average the chart. Besides the expected +/- one tach pulse, counting the tach pulses is a low-priority interrupt function and the OS services a lot of other interrupts first, which means you get more than just one tach pulse count variability for any one sample period.
---------------------------

The maximum speed of the Cooler-Master and Scythe 120mm PWM fans is 1200RPM, slower than this A-C fan. Last Friday, the Scythe US site announced their PWM fans, but with no details. Ebay is "selling" the Scythe fans, meaning "accepting bids" with a closing date of July 24. I assume that means somebody placed an order with Scythe, and the scheduled delivery date is July 24 (but I'm definitely not an expert on Ebay!).

edit: added URL of PWM spec 1.3

Felger Carbon · Post by **Felger Carbon** » Tue Jul 03, 2007 6:54 pm

I clipped the +12V lead on the fan I just tested and put 4ea 1N4004 1A power diodes in series with the cut ends. This has the effect of dropping the voltage ~2.8V. I didn't even have to unplug the fan to do this, it was "computer-top invasive fan surgery"!

The idea was to lower the fan RPMs for a given setting. This was more successful than I had anticipated:

100% dropped from 1534 to 718, 90% gave 582, 80% dropped from 1251 to 345, and 75% gave 231RPM! The fan would not run below 75%, although 70% took a long time to coast to a stop.

So I decided to try 2 diodes instead of 4: 100% 1158, 90% 1036, 80% 819, 70% 624, 60% 445, 50% 276, 45% 193RPM, 40% won't run. Hey, I like the concept!

FYI, a power diode drops roughly .7V per diode, but Schottky-barrier diodes (1N5817) drop .3V. This last set of numbers was with a 1.4V drop from a nominal 12V, but here's the choices with the right set of diodes: .3, .6, .7, .9, 1.0, 1.2, 1.3, 1.4, ...

The place to apply this mod is between the outer edge of the fan case and the edge of the fan motor hub, so super-glue or epoxy can lock the diodes in place. Mouser is where I buy the diodes; they're dirt cheap.

And you aren't limited to a fan that turns 1500 to 500RPM. 1.4V gives 1160 to 193RPM, for instance. I suspect the fan is pretty quiet at 193RPM!

Bluefront · Post by **Bluefront** » Wed Jul 04, 2007 2:37 am

I know that PWM fan control is bad-mouthed frequently on SPCR, and other forums........sometimes for good reason, sometimes not. I wonder how these AC fans would perform hooked to different boards? Are these new fans immune to PWM-induced noise (clicking).

The first time I hooked up a YL fan to an LIS2 controller (PWM).....I was shocked by the noise. I certainly hope this new breed of PWM-designed fans, is quiet no matter the particular board.....

Nice review....

Mr Evil · Post by **Mr Evil** » Wed Jul 04, 2007 5:25 am

Bluefront wrote:...Are these new fans immune to PWM-induced noise (clicking)...

4-wire PWM fans use a PWM frequency of ~25kHz, so there will be no clicking, but there may be a very high frequency whine, which hopefully will be above the range of most human's hearing.

Felger Carbon · Post by **Felger Carbon** » Thu Jul 05, 2007 9:39 am

For advanced and ambitious PWM users only!
Combining Quiet Fan with PWM for noise savings

(Quiet Fan is a bios facility offered by Asrock on their more recent PWM mobos. Other mobo makers may have similar bios features.)

If you're good at working with finicky electrical wiring you can replace the +12V on the PWM fan provided by the 4-pin mobo PWM connector and instead wire it to an analog fan controller. I did, but this isn't for amateurs.

First disable Quiet Fan in the bios. With 100% PWM programmed using SpeedFan, you can adjust the fan controller voltage for whatever RPM and CPU heat sink temperature you want at max load (BurnK7).

Then enable Quiet Fan, setting whatever temperature trip point you want and setting "Slow" as the fan speed selection. Now the PWM will be 70% below the trip point and 100% above.

If you set RPM100 to _, RPM 70 will be _:

1158 624 (16.1 dBA quieter than 1158)
1000 427 (22.2 dBA quieter than 1000)
900 292 (29.3 dBA quieter than 900)
800 178 (39.2 dBA quieter than 800)
(theoretical noise levels based on 60LOG10(rpm2/rpm1)

As you can see, the combination of Quiet Fan, PWM, and an analog fan controller can produce considerable noise savings! Especially when I consider the one HSF I'm currently using turns the 120mm fan at ~650RPM! Alas, the mobo it's on has no PWM...

It's easier to just cut the red +12V wire on the rear of the fan and put some diodes in series. 4 1N4004 diodes = 710RPM @ 100%, 2 diodes = 1158RPM. Remember, 1N4004 ~.7V, 1N5817 ~.3V; keep the diode's white band closest to the fan motor hub.

For minimum fan noise, the RPM at 100% should just barely be able to cool the CPU under 100% load.

sun.moon · Post by **sun.moon** » Thu Jul 05, 2007 1:11 pm

I've had the Arctic Cooling 120 PWM fan in use for some time now and am quite satisfied with it. No clicking, no hissing, can't distinguish it over anything else in my system. I guess it has been available in Europe for a while longer. I have it on my P5B running under QFan. It always runs btw. 430 and 440 rpms according to PC Probe II.

Just my opinion/experience...

- sun.moon

jessekopelman · Post by **jessekopelman** » Thu Jul 05, 2007 2:38 pm

I consider the PWM daisy-chain to be a very useful feature of these fans. Many people only have a single PWM header -- CPU. At the same time, it seems useful to have coordinated fan speed variation based on CPU temp. Unless you are using a heatsink ducted directly to an exhaust fan, having the CPU fan speed up with temperature just dumps the heat into your case for other components to deal with. If the exhaust fan is also PWM controlled by CPU temp, you avoid this problem. Noisewise, you don't mind the exhaust fan ramping up because it is hidden by the CPU fan ramping up.

Felger Carbon · Post by **Felger Carbon** » Thu Jul 05, 2007 7:38 pm

sun.moon wrote:Just my opinion/experience...

So, how come you old-world types couldn't take the time to clue us grubby colonials in?

Ryan Norton · Post by **Ryan Norton** » Mon Jul 09, 2007 7:21 am

These are out of stock at Newegg right now, where they sell for US $9.99. Seen them anywhere else? I'd love to try one as the CPU fan on my Ninja + P5B, so I'm looking around.

OK, found some at NCIXus,

http://www.ncixus.com/products/25013/AF ... 20Cooling/

$6.38, not bad!

jessekopelman · Post by **jessekopelman** » Mon Jul 09, 2007 5:47 pm

Ryan Norton wrote: OK, found some at NCIXus

Ooh, they have the anti-vibration frame version too, which Newegg did not seem to carry. Note: This one is only usable as an exhaust fan due to its construction.

darren · Post by **darren** » Thu Jul 12, 2007 5:33 am

I've been using the Artic Fan 12 PWM for almost 4 months now without any issues. I replaced the Antec Tri-Cool exhaust fan in my Solo case with it and used the 4-pin SYS_FAN connector on my Gigabyte GA-965P-DQ6 . It throttles up and down nicely along with the CPU fan on the CPU connector. I can't compare it to other models for noise, but I can say I've had no problems with it in terms of clicking, vibration or anything else and it's been on my desk at ear level about 2 feet away.

atlr · Post by **atlr** » Fri Jul 13, 2007 7:34 am

jessekopelman wrote:Ooh, [NCIXus] has the anti-vibration frame version too, which Newegg did not seem to carry. Note: This one is only usable as an exhaust fan due to its construction.

Good to know.

anti-vibe model http://www.ncixus.com/products/24331/AC ... 20Cooling/

http://www.newegg.com/Product/Product.a ... 6835186015
http://www.ncixus.com/products/25013/AF ... 20Cooling/

mfr
http://www.arctic-cooling.com/fans1.php

P.S. NCIX ships from Canada and that journey adds about US$20 to a 4-fan order

Felger Carbon · Post by **Felger Carbon** » Fri Jul 13, 2007 9:15 am

FYI, Newegg has the 80 and 92mm A-C PWM fans in stock, but fresh 120mm stock isn't due until toward the end of the month. Meanwhile, Directron and PCToys carry the PWM line - I just received 2ea 120mms from PCToys yesterday (with some other stuff).

My Scythe PWM fans, ordered 8 days ago from FrozenCPU in NY, are due this afternoon. I'm keeping my fingers and toes crossed. The camel caravan from NY seems to move slowly in warm weather!

I wonder if the CPU stayed frozen during NY's recent heat wave?

jessekopelman · Post by **jessekopelman** » Fri Jul 13, 2007 10:12 am

I guess after waiting months and months (they've been widely available elsewhere since March) for these A-C fans to come to the US, the floodgates have opened.

S.SubZero · Post by **S.SubZero** » Fri Jul 13, 2007 8:47 pm

I ordered two of these from Directron.com just now, $8.99 each and in stock, without NCIX's insane shipping ($17 for two $7 fans?!). I didn't see the anti-vibration ones on their site tho.

http://www.directron.com/af12025pwm.html

Ryan Norton · Post by **Ryan Norton** » Sun Jul 15, 2007 6:24 am

Yeah, I got all pumped about how cheap these were at NCIX then I saw the shipping cost. Not happening!

Felger Carbon · Post by **Felger Carbon** » Sun Jul 15, 2007 7:39 am

Ryan Norton wrote:Yeah, I got all pumped about how cheap these were at NCIX then I saw the shipping cost. Not happening!

NCIX is in Canada. If you're in the US of A, parts have to be shipped across an international border, thus the high shipping cost. It's not as if NCIX is pocketing that money.

Sometimes you'll want to order across a border to get something that's not otherwise available. Never order across the border for what you can get here, if you're a Yank.

BTW, this accounts for most of Ebay's insane shipping costs. A lot of these guys are in the far east where stuff is available first. But they have to ship internationally to get stuff to you across the Pacific and over yes, the doggone border.

Redzo · Post by **Redzo** » Sun Jul 15, 2007 8:22 am

What about vibration ? I have had several AC fans and 3 out of 4 vibrated heavily (on full tilt though) but 2 of them were vibrating even when running @6-7V...
I know that they come with anti-vibration frame but mine were vibrating so hard that rubber holders that keep it attached to frame broke off.
That was my last contact with AC fans

Felger Carbon · Post by **Felger Carbon** » Sun Jul 15, 2007 3:13 pm

Redzo wrote:I know that they come with anti-vibration frame...

IMHO, that "anti-vibration frame" is a pro-vibration frame!

I bought a couple of those fans when they first came out. Threw them away in a hurry.

kaange · Post by **kaange** » Mon Jul 16, 2007 5:40 pm

darren wrote:I've been using the Artic Fan 12 PWM for almost 4 months now without any issues. I replaced the Antec Tri-Cool exhaust fan in my Solo case with it and used the 4-pin SYS_FAN connector on my Gigabyte GA-965P-DQ6 . It throttles up and down nicely along with the CPU fan on the CPU connector. I can't compare it to other models for noise, but I can say I've had no problems with it in terms of clicking, vibration or anything else and it's been on my desk at ear level about 2 feet away.

Pretty much the same experience (and time frame) as me except my m/b is a P965-DS3.
It seems no noisier at the same rpm as my 120mm Nexus, Noctua or Scythe S-Flex fans.

When I had my CPU heavily undervolted, the fan was sometimes reported as running below 200 rpm but that my have been misreporting as the system was not stable when running at those voltages. Now it runs as stock between 400rpm and 1500rpm which is OK for me - I need the full speed when heavily overclock for video processing.

S.SubZero · Post by **S.SubZero** » Wed Jul 18, 2007 5:46 am

How complex is the cabling? I've been trying to get a diagram of just how this all looks but Google isn't really helping.

I have two of these on order. They will be replacing both case fans in my Lian-Li PC7B. One will be in the front, one will be in the back. My Intel DP965LT has the 4-pin header near the CPU fan which I'm using for my Zalman 9500AT, but it also has a *second* 4-pin PWM header in the lower rear section of the board in-between the PCI/PCIe slots. I'm planning to use that header for the fans and I'm just trying to envision how the cabling will be.

kaange · Post by **kaange** » Wed Jul 18, 2007 6:12 pm

Each of the AC PWM fans has a female 4 pin connector on a short lead off the fan so you just plug the 2nd fan's normal connector into this.

Felger Carbon · Post by **Felger Carbon** » Wed Jul 18, 2007 6:48 pm

kaange wrote:Each of the AC PWM fans has a female 4 pin connector on a short lead off the fan...

Male connector. The normal fan-lead connector is female.

Rocky2001 · Post by **Rocky2001** » Sat Jul 21, 2007 10:40 am

I bought this fan as my only back exhaust fan for my case. Asus P5B motherboard has only ONE PWM 4 pin plug which is used for the CPU fan.

The instructions are to plug the PWM case fan to the PWM 4 pin on my motherboard and then have my CPU cooler piggy back off this fan. Then there is a fan RPM sensor plug that I'm supposed to plug into one of the "chassis" plug.

When I boot up to my BIOS, it shows my CPU rpm at 600rpm which really is the exhaust fan and my chassis rpm at 1200 rpm which really is my CPU fan. Both which are now switched around. Is this normal? The instructions say to ignore the RPM alerts in the bios. Is this normal for the RPM signals to be switched around? I'm almost positive I didn't install this wrong (it's pretty full proof).

Is this normal? I guess I need to disable all the fan monitor in my bios so I will not get the boot error saying my CPU fan is low (when really it's not)

Felger Carbon · Post by **Felger Carbon** » Sat Jul 21, 2007 11:09 am

Rocky2001 wrote:When I boot up to my BIOS, it shows my CPU rpm at 600rpm which really is the exhaust fan and my chassis rpm at 1200 rpm which really is my CPU fan. Both which are now switched around. Is this normal?

Boot your bios. Stick your finger in one of the two fans. See which RPM drops to zero, which will answer your question, or at least give you something to think about. Now go put a band-aid on your finger.

Rocky2001 · Post by **Rocky2001** » Sat Jul 21, 2007 2:44 pm

Felger Carbon wrote:
Rocky2001 wrote:When I boot up to my BIOS, it shows my CPU rpm at 600rpm which really is the exhaust fan and my chassis rpm at 1200 rpm which really is my CPU fan. Both which are now switched around. Is this normal?
Boot your bios. Stick your finger in one of the two fans. See which RPM drops to zero, which will answer your question, or at least give you something to think about. Now go put a band-aid on your finger.

NO. The bios is reading is definately swtiched around. I put my finger on the chassis FAN (600rpm) and it thinks it's my CPU fan.

Has anyone istalled one of these in their system?

jessekopelman · Post by **jessekopelman** » Sat Jul 21, 2007 8:53 pm

Rocky2001, did you try the obvious and just plug the fans in the opposite from recommended order (i.e. CPU fan first, case fan as piggyback)?

Rocky2001 · Post by **Rocky2001** » Sun Jul 22, 2007 1:22 am

jessekopelman wrote:Rocky2001, did you try the obvious and just plug the fans in the opposite from recommended order (i.e. CPU fan first, case fan as piggyback)?

That would work but there is no piggy back option for the CPU fan.

I'm wondering if the wires that came with my fan are properly put together.

Felger Carbon · Post by **Felger Carbon** » Sun Jul 22, 2007 5:31 am

Rocky2001 wrote:That would work but there is no piggy back option for the CPU fan.

It sounds as if you have only one A-C PWM fan. I have not seen any of these fans - 80mm, 92mm, 120mm - that did not have the "piggy back option". What is the second fan? Does it have 4 wires?

skiddy · Post by **skiddy** » Sun Jul 22, 2007 6:44 am

[quote="Rocky2001"] Then there is a fan RPM sensor plug that I'm supposed to plug into one of the "chassis" plug.
quote]

That wire (should be yellow) needs to go into the 4-pin mobo connector for the cpu, if you want it to read out as cpu rpm. If you plug it into the 'rear fan' plug, software will read it as the rear fan.

The Arctic-Cooling PWM fans are in and they work!

The Arctic-Cooling PWM fans are in and they work!

The first PWM fan mods!

Help with this fan PLEASE...

Re: Help with this fan PLEASE...

Re: Help with this fan PLEASE...

Re: Help with this fan PLEASE...