How do I verify if PWM fans are actually using PWM mode (PWM on/off)

[AbsoluteJoe]

I have an X79 Extreme4-M and want to determine if the CPU and CASE PWM fans are actually using PWM mode.
Neither the BIOS nor AXTU Hardware Monitor reports the actual PWM fan status. The term "Automatic" is not
enough to verify that the fans PWM feature is actually being used. The actual current PWM status (on/off)
needs to be displayed. I don't really want to scope the fan leads to determine if it's PWMing or not.

As others have noted, more work needs to be done in the fan speed and thermal controls area. Ideally, each fans
speed controls should be user customizable for both temperature (sensors individually selectable) and speed range.

[parsec]

Interesting question. What makes you think PWM is not working, the rather sluggish and clumsy way the fan control software works on these boards? Do you think it is just using voltage control on the four pin CPU fan header, and Chassis 1?

If you connect a three pin fan to a true PWM fan header, it should run at full speed. If it doesn't and the fan speed varies, then IMO PWM operation is not working or missing.

I have never seen a board that had a PWM status indicator or selection on a PWM CPU fan header. I believe that is an AMD thing, you won't find that on Intel CPU boards.

I agree the fan control software on these boards (I have a Z77 board) is primitive, non-intuitive, and really the worst I have seen in a long time. It's not active until Windows starts, or am I missing something? What is the point of the "Target Temperature" setting?

[CRRacer]

I have the Z77 Extreme4 and noticed that in the EUFI I can set the fans to full on, or auto. When I select auto it gives me two more options, target temp and fan speed. Really? This is auto? Abviously I want the lowest CPU temp and regardless of setting that, if I don't adjust the target fan speed, the fan remains full on (mostly because the target fan speed was set to 9 (EUFI) or 10 (in the ASrock tuning utility).

Playing around with the tuning utility, I set the target temp to the lowest possible and I set the fan speed to 5, at this speed I can't really hear the fan. I am aiming for quiet here. My question is, in doing this, if the target temp is reached or exceeded, does it automatically increase the fan speed??? I cannot find this answer and was curious if anyone knows?

[parsec]

CRRacer, I agree, the fan control software settings are confusing and are not explained anywhere that I can see.

I use a PWM controlled fan on my CPU cooler, used with my ASR Z77 EX 4 board. If I set the CPU fan speed to 5, it does seem to increase and decrease the fans speed with the CPU temperature. I think the CPU fan speed setting number will be the lowest speed that fan will run at, whatever speed a particular fan runs at, at that setting/number. The "Target Temperature" setting doesn't make any difference that I have noticed so far, but I've only had this board for about two weeks.

The fan control software provided with another manufactures boards I have lets you choose between several fan speed profiles, or you can create your own. It works fine, the CPU fan speed changes with the CPU temperature, and is active most of the time, not just once Windows starts, as with the ASR software.

The ASR fan control software is the only thing I don't like about this board. Otherwise its been a great board, and in some ways better than others I have used. You would think that fan speed control is not a difficult thing to accomplish, but it is a weak point of other manufactures boards I have seen, and in some CPU coolers too.

[CRRacer]

OK, kinda makes sense now that you mention it. I have the CM Hyper 212+ for my cpu cooler and when I set the fan speed to 5 in the software I noticed the fan speed at 1500, so I set the software to 1, fan speed still showed 1500, perhaps that is as low at the rpm is for this fan.

Either way, I set it at lunch and while the cpu was inactive and showing 38c, the fan speed never increased, but lowering the fan from full on to setting 5, my cpu temp didn't increase either. Will have to stress it a little tonight and see what happens.

But I agree, I really like this board. I got it kind of a backwards way, bought a z77 sabertooth that had a single pin in the socket that wasn't quite right. Have no idea if it was bad or not, but RMA'd the board with newegg and after reading the RMA stories and bent pins I was preparing for the denial of replacement so I bought this board before I even shipped the other one back. This board has been flawless, and even better is newegg replaced the board and the new one is perfect. Bad thing is now I ordered another i7 2600k to put in the asus board.....

[parsec]

The fan used with a CM Hyper 212 is also sold separately as the CM Blade Master fan. It's a very good 120mm PWM fan, rated at 600 - 2000 RPM. I have a Hyper 212, and I bought a Blade Master fan, and the fans are identical. I'm using this fan with a Scythe Mugen 3 CPU cooler on this board, with an i5-3570K.

In the software, with the Fan Target Speed (TS) set to 1, and the Target Temp (TT) set to 45C (lowest setting), I get a fan speed of ~1170 RPM (don't forget to click the Apply button at the bottom of the screen, nothing changes if you don't click Apply, and then it takes a few seconds.)

With TS set to 1, and TT at max (65C), the fan speed did not change at all, still at ~1170 RPM.

With TS set to 10, and TT at minimum (45C), the fan speed is ~2000 RPM.

With TS set to 5, and TT at min, fan speed is ~1630 RPM.

With the CPU at idle or low load, such as just writing this post, I can set the TS to 1 (1170 RPM) and the CPU temp is ~34C.

I ran several short Intel Burn Tests (IBT) with TS set to 1 and TT at min. The fan speed did speed up to ~1700 RPM, but it was slow to respond to the CPU temp change. I could clearly hear the differences in fan speed, and it took over five seconds for the fan to change speed at all. My ASUS P67 board and it's fan control software increases the CPU fan speed instantly when the CPU temp goes up during stress testing.

The IB CPU run hots, compared to my SB i7-2600K. The 3570K hit 70C+ in ~15 seconds during the stress test, while the 2600K will reach 60C after a minute, even at a higher clock rate.

So the automatic fan speed control of the PWM CPU fan does work, but it's response is slow. You must click Apply every time you change any setting, or it won't take effect.

[CRRacer]

Yep, I was hitting apply.

I played with it a bit this morning and saw the problem. I'd apply, then look at fan speeds and see no change. It takes 5 to 10 seconds for the new fan RPM to register with the software, I wasn't giving it that 5 to 10 seconds....

Set to high fan speed I get roughly ~2000 RPM, at the '5' setting, I get ~1500 RPM and at '1' setting I get ~1000 RPM.

Noise wise, I didn't really notice much db difference between 1-5 unless I got really close to the side vent on the case, so I just leave it at 5 and of course the TT was already set to lowest point, (45c I believe). Of course I bypassed all this by just setting it up in the EUFI.

It's good to know the fan model should I choose to buy a second fan for push/pull, I just wonder if there is much advantage here. It's already directly inline with my rear case fan (Cooler Master CM 692 II advanced USB 3.0) which sits about 3" away.

Typical idle temps for me are 38C, but this PC is in my bedroom and it's a little warmer upstairs than it is downstairs where my Core 2 Duo E8400 clocked to 4.07 GHz sits. Surprisingly the Core 2 idles around 28C. I have the i7 clocked to 4.2, have tried it as high as 4.6 with no issues, but I didn't see a need to go that high since my WEI rating for the CPU is 7.7 at 4.0, and 7.8 at 4.2, 4.4 and 4.6. I haven't tried 4.8 in the auto clock (it's in red (kinda scarey! the color, not the thought of 4.8...), to see if it'll hit 7.9.

Curious to compare this to the Sabertooth when all my pieces get here tomorrow to finish that build. It'll be cooled by the Antec Kuhler H2O 620 (hey, I'm a sucker for a newegg bargain on something I'm interested in...)

[parsec]

Yes, this CM fan is pretty quiet up to ~1500 RPM. Actually, any good fan is pretty quiet up to that speed, but beyond that noise really increases. Almost all extra quiet 120mm fans do not go above ~1500 RPM. I also like Scythe 120mm PWM fans, they have a new model that IMO is better than the CM fan, but I'm using the ones I have on other PCs.

I've used push-pull on air coolers and water cooler radiators. The difference is not large in my experience, 2 or 3 degrees C at most. That does depend on the cooler too. If it's very large, then another fan can help. A good test is with the CPU at idle, set the CPU fan at full speed. If the CPU temp drops quickly and by 5C or more, then another fan would be an improvement. If not, then a second fan won't help much.

I have an E8400 too. They run very cool compared to newer CPUs, that does not surprise me.

You'll probably like the SaberTooth board, I have a SaberTooth X58 and SaberTooth P67. The P67 has been a great board, except being unable to wake from Sleep, which is an ongoing issue with some Intel 6-series chipset boards. You'll love the fan control software (Thermal Radar in AI Suite II) compared to the ASR fan control, night and day difference. I predict it's POST time will be longer than the ASR board, the POST and boot time on the ASR is the fastest I've seen.

But check the CPU temp reading in Thermal Radar by comparing it to a third party software. On my board, they changed the UEFI to send a different CPU temp reading to the software that is ~10C lower than the true temp. After a UEFI update, my CPU temps were 10C lower, not core temp but overall CPU temp. I was getting CPU temps of 15C or less at idle during the winter, or even single digit CPU temps, which is impossible. Also, if you run another temp monitoring program as the same time as Thermal Radar, the data get corrupted in Thermal Radar, and the readings are insane. They may have fixed this in the newer version of the software on your SaberT, but just watch out for this.

You'll also be seeing temps from sensors all over the board (the P67 board has 11) that may surprise you. A single sensor "mother board" temperature does not begin to tell the whole story.

No board is perfect, there always seems to be at least one thing that is weird,

[CRRacer]

You're right, I do like the AI Suite and all the sensors!

My experience though is the Asus board gets in to windows faster. I don't know if it is getting through the post quicker or not, just know that after post, windows loads quicker. Could be because I had Win 7 HP on the Asus and Win 7 Pro on the ASRock though. ASrock just does get the windows logo together before it hits the desktop. Asus I get 4 colored dots about halfway through to there distination before I get the desktop. Also Asus has the Samsung 830 SSD and the ASRock has the Crucial M4 SSD. Haven't compared speeds yet, but preliminary looks like the Samsung is the faster drive, least ways from the scores it gets in passmark.

[parsec]

It probably is the difference between the home and pro versions of Windows, pro is larger is size and more things load and start. The SSDs are a little different too, the Samsung can be faster than the M4 in real world performance, rather than benchmarks.

Compared to my ASUS P67 board, the ASR Z77 board posts and boots faster. Might be the Z77 chipset and its software (Intel ME, etc). If you are using the IRST driver, Z77 uses the 11... version, which can be run on P67, but who knows what else is going on.

Does the Z77 SaberT use the Marvell chipset for the extra SATA III ports, or the ASMedia, like the ASR? IMO, the Marvell is slower to POST.