Well, I agree that the Asus bios fancontrol is a step in the right direction. My only real objection to it (apart from the inability to control 3-pin fans on the cpu fanheader and having to buy a pwm fan) is the fact you can't set the chassis fans to less than 40%. I don't understand why such an artificial limitation had to be put in place. The same goes for Intel, why would you limit a header to 6V? My system is very cool during idle/light loads, there is no need for cpu or casefans to spin at all. A second nuisance is that the 2 chassis fan headers cannot be controlled independently, but I'm willing to overlook that.

Now I'll admit that fex. 1500 x 0.40 = 600rpm is fairly silent. But it's more a matter of principle and that's also where Speedfan comes in. If I want to completely turn off my fans I should be able to do so. If I want fan 2 to spin faster but not fans 1 & 3 when HDD 4 reaches a certain temperature I should be able to do so. Speedfan offers this flexibilty and I also like how it conveniently shows temps and voltages. I have found Speedfan automatic control to be very reliable. And there is an option to set fans to 100% on program exit.

Anyways, I noticed there's a new beta version of Speedfan, 4.43. From the changelog:
"Beta 6 improves fan control for IT8712F and adds full support for IT8721F (that was often mistakenly identified as an IT8718F). IT8721F support for fan control has been fully verified and it works great. This beta starts to add support for Fintek F71809 and Fintek F71859. Please contact me if you have any of the two. Several other fixes and improvements are there."

It would be interesting to see if this version works on the Asus boards. Since I'm wondering if the issue has to do with Speedfan or with the Asus boards themselves. (like on some P5 boards where Speedfan wouldn't work after updating the bios).

BIOS control also won't allow you to adjust chassis fan speed as a function of discrete GPU temperature, for example. Would be pretty neat, but I can see why it's not going to happen.

Hey... what's going to happen to the Asus P8H67 board now??

It has been pulled from retailers. Newegg has it deactivated, and so does Microcenter. This must be in light of the chipset flaw revealed by Intel, correct? So what now? Will it come back?? Will it get re-manufactured with the flaw fixed and then released for sale again? I'm going to be super pissed if I have to look for a motherboard and waste time reading reviews all over again ......

AFAIK, all the current model boards will come back unchanged -- except for that flawed chip.

From reading the fan, this review and searching the Internet, it looks like connecting 3-pin fans to the chassis fan header should be okay and work with Q-Fan. The board might complain at POST (at having no CPU fan) but that warning should be defeatable. I'm thinking of the Pro version in hope that the 4+2 VRMs might use a little less power and make it closer to the Intel.

Well, that's an option in some situations. But ofcourse you sacrifice a fanheader, the amount of which can't exactly be called plentiful in the first place. Point is, if you want speedcontrol on more than 1 fan, you'll need at least 1 pwm one.

From looking at the specs, when it comes to fancontrol it seems your best of with MSI or Asrock. Many of their mATX boards have 2 or even 3 fanheaders. Some of the Asrock P67 ATX boards even have 2 cpu fan-headers for push/pull configuration (although 1 is 3-pin, so you still need an adapter for Silver Arrow fex.). Considering Asus and Asrock are basically same company it makes me wonder even more why the 'budget' brand can get it right but the main brand is so lacking. Just 2 fanheaders on Asus P67 mobo's...yoohoo it's the 21st century, we have cases with 4 or more fans and dual push/pull fans on our cpu coolers, aftermarket graphic cards coolers with fans that would like to slow down sometime as well. And most importantly, we're sick of the noise from our pc's. But hey, you can overclock your system using your Bluetooth cellphone! Gigabyte is even worse, many of their mobo's have just one controllable header, which is frankly unacceptable (and a big pity because their boards are superb in every other aspect).

In terms of idle power consumption, is Sandybridge really that much better?

The Intel DH67BL is 17W at idle, but so is the Intel DH57JG mini-ITX board.

Hmm good question. I am interested too. But the boards aren't exactly comparable. The DH57JG is mini-ITX, compared to the substantially more components and slots featured on both of these m-ATX boards in this review. That will probably make some difference of a few Watts or more. Question is how much of a difference.

Yes I've been searching for a review of a Sandy Bridge mini-ITX board that compares it to a Clarkdale mini-ITX board... but haven't found one yet

There's also a big difference between mini-ITX boards: of the Clarkdales, the Intel and ECS are very efficient, the Zotac less so, and the Gigabyte somewhere in between. The Zotac and Gigabyte have more components to power, like USB3, wireless etc.

I'm still a bit confused on fan control (SpeedFan vs BIOS; PWM vs DC; etc). Fans that don't go below 40% is disappointing to me; I don't want extra noise.

I want one of these boards for HTPC playback, and nothing more than that (I use my desktop for heavy encoding).
I want the quietest board, as I'm not going to fret over 5 watt advantages (electricity is very cheap for me). And I want to know how to get there with what kind of fans.
I also need low temperature board so I don't end up running noisy fans.
Bootup...don't care. My HTPC is on full time to receive shows, or if I feel like conserving some watts I just put it to sleep.

Not having a VGA port is very important to me. I use this port to convert to composite video for my older RPTV.

If someone can put me on a better path for a new HTPC, please inform me.

Jake

P.S. Good review, and thank you. Only suggestion I have is to display a single page for printing.

I have the DH67GD, but am experiencing the same thing---neither Speedfan 4.42 nor 4.43 appear to actually change the speed of the fans. I am using PWM fans.

It might be that it just doesn't work with that motherboard. Take a look at something like this:
http://www.performance-pcs.com/catalog/ ... s_id=25981

The SPCR article suggested that it does work. On page 6, near the bottom, "[The Intel DH67BL] is fully supported by SpeedFan, which can adjust the speed of all of the board's three fan headers, though one of them can cut the fan's power to a minimum of only approximately 6V."

Granted, I have the "GD", and the article talks about the "BL", but surely they have the same fan controllers. Also, corduroygt (quoted above) has the BL version and has the same problem.

Perhaps it's a sample variance issue? E.g., maybe Intel uses whichever fan controller is the cheapest at the time of manufacture.

Computers... you can't live with them and you can't live without them.

I have have had the same kind of headaches with other boards. Attempting to control fans through the headers has always generated me nothing but headaches. I don't personally like having to do all that fiddling around.

I like to pick good fans and then just run them at a constant voltage. I find it easier to experiment with fixed reproducible settings using straight fixed voltages.

I just purchased an Intel DH67GD and was similarly disappointed to see that speedfan didn't control fan speed out of the box.
(one of my primary reasons for choosing this board was that it had been confirmed to work with speedfan).
After several hours of trying to fix this I found out that once you reflash the bios to the most recent version, a fan control setting appears (in the bios). If you change this setting from auto fan speed to manual fan speed then suddenly speedfan controls the speed perfectly.

I got a DH67GD (a DH67BL + FireWire + DisplayPort) in the end and the fan control is great with the latest (v110) BIOS, which is much newer than the one in the SPCR review. The controls allow setting of:
- target temperatures for CPU, PCH (chipset), VRM and RAM) along with damping characteristics for the ramping)
- minimum and maximum duty cycle for each fan header
- they also run the fans at 100% briefly on power on/resume from hibernate/sleep.

It looks a lot like this one:
http://benchmarkreviews.com/images/revi ... ios_11.jpg
at
http://benchmarkreviews.com/index.php?o ... mitstart=5

They also allow what seems to be re-assignment of the fan headers for different purposes (e.g intake chassis fan header used for CPU fan). I tried this (the CPU fan is 4-pin PWM only and I wanted to use 2 Noctua DC fans for the CPU) and although it changes the text in the BIOS (and Intel Desktop Utilities), it doesn't appear link the non-CPU fan header to the CPU temperature.

With a Noctua NH-C14 and the 2 Noctua fans on the minimum speed on the intake header (~ 670 rpm), the cooling was good enough not to need BIOS control. Except that I wanted them slower/quieter than this. Fixed with a FanMate would have been an option except for the high starting voltage on the Noctua fans.

So I now have 2 Scythe PWM 120 fans on the CPU fan header. These go down to 200 rpm or so (if desired) with the BIOS fan control. The DC case fans seem to be controllable down to 5% duty cycle with the BIOS control.

Here the specification and results (case, PSU and optical re-used from viewtopic.php?t=31528):

Case: P180 (no filters, front intake grills and door grills modified for more air, top vent blocked)
Case fans: 2 x Nexus 120mm (upper chamber intake / lower chamber middle – split off inlet fan motherboard header at 5% minimum duty cycle), Scythe SFF21E (upper chamber back outlet – on motherboard outlet fan header at 15% minimum duty cycle)
PSU: Nexus Value 430W
Motherboard: Intel DH67GD
CPU: Intel Core i5-2500K
CPU heatsink – Noctua NH-C14
CPU heatsink fan – 2 x Scythe SY1225SL12LM-P (Split off motherboard CPU fan header at 30% minimum duty cycle)
SSD: Crucial M4 128GB
Optical: Some DVD-RW drive
Other: USB webcam, Bluetooth dongle, card reader

Idle (Windows 7)
Power: 29W (from APC PowerChute), 32W (power meter)

CPU – 32C
PCH (chipset)– 63C
Memory – 33C
VR (VRMs) – 31C

CPU fans– 409 rpm
Inlet fans – 421 rpm
Outlet fans – 401 rpm

Prime95 (in-place large FFTs)

Power: 98W (APC PowerChute), 110W (power meter)

CPU – 57C
PCH (chipset)– 69C
Memory – 42C
VR (VRMs) – 45C

CPU fans – 409 rpm
Inlet fans – 421 rpm
Outlet fans – 401 rpm

So:
- the fans don't increase in speed at all.
- it can continously use the single-threaded 3.7 GHz Turbo Boost (never drops with to the 3.4 GHz long-term boost - due to the efficient heat sink?).
- It's quick!

Thanks to SPCR and the forum members for being such a great resource.

Greetings. I am curious to know how peripheral card fitment is on the Asus board, particularly if running a modern dual-slot GPU in the PCI-E x16 slot and a soundcard in the PCI slot. Is there reasonable clearance or not? I see the Intel board appears to have more clearance between the PCI-E and PCI slots.

Here are some of the BIOS (or UEFI) fan/temp control screens from the latest DH67BL BIOS revision (0122).

Fan Control & Real Time Monitoring


Front Fan


Front Fan (Manual)


Processor Temperature


As roberlin pointed out, the front/rear fan headers must have the Control Mode set to Manual in the BIOS to allow SpeedFan to control them, and as Nigel said, the CPU fan header doesn't provide speed control for 3-pin fans. I don't have a 4-pin fan so I don't know whether SpeedFan can control the CPU fan header speed when such a fan is connected.

There seems to be a bug in the 0122 BIOS that prevents adjusting the Minimum Duty Cycle setting for the front fan header. It's stuck at 30 in the BIOS, though SpeedFan can still control it provided the Control Mode is set to Manual. Anyone else observed this?

The BIOS reading of 45-48C for the CPU temp is not consistent with the numbers in Windows hardware monitor utilities (including Intel's own) which report 25-30C idle in my case.

Just a heads up for fellow DH67BL owners - some silly duffer at Intel mixed up the front and rear fan header settings in the 0156 BIOS update, so that now the front header BIOS setting controls the rear header, and vice versa

EDIT: wow, just noticed from the previous post date it's been a year since I bought mine for my new Sandy Bridge desktop. Feels like I only just put the thing together last week/month - time sure flies as you get older.