I've been busy for the last few weeks, so please forgive the slow reply.
Abula wrote:Yes, i already tested it with my H170 Pro4, but not all PWM fans are design the same way, meaning that some fan can stop under certain PWM like Noctua NF-P14R PWM, and others can't like Noctua NF-S12B PWM. If you do want to stop certain fans, chose carefully that its a fan that can be stopped.
That's a valuable tip, thanks. My current plan is to have just the single 120mm Noctua fan from the NH-L12 positioned at the top of the case and the NH-L12 installed with the fins aligned vertically.
Abula wrote:MSI has their own way they designed their bios, somewhat different to AsRock, specially on skylake. MSI can control PWM fans on CPU_FAN1 and CPU_FAN2, both are true PWM fan headers, but their range is not unlimited, they are binded to 12.5% increments, while asrock can vary 1%,-
That's a serious win for ASRock in my eyes. I really want to push things as far as I can (in terms of reducing noise) so having 1% increments is a big plus.
Abula wrote: -on MSI your bottom limit is 12.5%, with AsRock you can go to 1% as the minimum (i havent seen a fan below 12% so its not a big deal), now where it differes is that with MSI your next breakpoint (as minimum) is 12.5 + 12.5 = 25%, with AsRock is whatever you want (cant be lower than your before breakpoint, but can be 1% higher if you wish).
I'm not quite sure I understand what you mean by "breakpoint". Do you mean that you can have a fan curve that goes from high to low and back to high?
Abula wrote:Now another thing that differs is that MSI in the past their CHA_FAN1,2,3 headers were voltage controlled, meaning you could undervolt 3pin fans easliy, just the range is different, was from 40% to 100% (in 10% increments), at the end was very good, but in skylake i did one build with MSI H170 Gaming3, and while i could undervolt the 40% fine or 50%, 60%, 70%, 80%, 90% or 100%, i coudlnt make it work dynamically with the CPU temp, every time i enable the smart feature (to make it dynamic) it went to 12volts and i couldnt control it, my guess is that it becomes a PWM header, but i never really tested it. With AsRock the CHA_FAN1 is most cases is autodetect, like the CPU_FAN, meaning it can control PWM or 3pin, the rest, like CHA_FAN2 are PWM, but there are certain headers like in my H170 Pro4, CHA_FAN3 is a 3pin none controllable, you do need to check the manual to see what the mobo will have, all headers can be stopped on the AsRock, as i said before, will more depend on the fan you chose.
More great information, thanks. As I'm looking at m
ITX, I'm almost certainly limited to just the two fan headers but I will bare the above in mind when I'm looking more deeply into their specifications.
Abula wrote:On Asus + FanXpert, under windows, you can run the tuning so FanXpert establishes the min max of each fan, here it will override any restrictions the bios has on each header, fanXpert takes full control of it. Here fans can be stopped under the FanXpert profile of Silent, all fans can be stopped with the exception of the CPU_FAN (asRock mobos do allow this on bios). With Speed fan i cant tell you much, its a very versitle fan control software, but i never been able to recognize all my sensors, and i personally prefer bios over software, so recently i haven't tried, i know many here in SPCR do like an have made Speed fan work wonderful, maybe someone else can vouch in to tell their experience.
I think something may have got lost in translation so to speak. I actually use SpeedFan and have used Fan Xpert but I've yet to BIOS fan control on a modern mobo. I wanted to know how close modern, BIOS based control could get to SpeedFan. Non the less, thank you for the reply. It's a perfect example of what makes these forums so valuable for silence enthusiasts.
mikeb wrote:Just a heads up with the AsRock motherboards running linux. I am seeing constant excessive CPU usage(~12%) with the systemd journal(ie most ubuntu/redhat/debian distros) logging several thousand BIOS errors per second. This is with the latest AsRock BIOS and all updates applied. It appears to have been resolved for the matching ATX Boards BIOS months ago, but not yet in the MATX board I purchased, as they seem to get far less attention. You might want to carefully check the BIOS updates for the specific model before you purchase if you go with AsRock.
That said the AsRock BIOS is good, hopefully I don't need to return it. If I do I will be moving to an Asus.
That's something I'll keep an eye open for but I'm going to hazard a guess and say it's probably not as much of an issue with the current front runner motherboard I have my eyes on.
---------------------------------------------------------------------------
Just a little update.
The possible spec of my built has changed slightly. Initially I had my eye on the ASRock H170M-ITX/ac as it ticks all the boxes I described in my original post. It's also (oddly) cheaper than the non-ac version in the UK. There is one issue though... well two technically. The additional storage interface on the board is an mSATA/mPCIe slot that's located about half way between the PCie slot and the CPU socket. This would almost certainly cause issues with clearance between the mSATA drive, which would stick up vertically from the board, and the NH-L12. One possible solution would be the find a somewhat rare half height mSATA SSD. But such drives are relatively hard to find and of questionable quality.
So my attention has turned to the Fatal1ty Z170 Gaming-ITX/ac. It's complete overkill for a HTPC and there are cheaper option out there if clearance wasn't an issue but it'd end up costing me the same as the board above. The money it'd cost me to buy the half height mSATA drive would go the the mobo and I'd have the preferred option of using an SSD I already have thanks the the motherboard having 6x SATA ports. There's also another big plus with the Fatal1ty board. It has a HDMI 2.0 port. Which means it can theoretically output 4k @ 60hz (although I can't find any affordable CPUs that support it). It's not a feature I had on my list but it would represent a degree of future proofing... it could increase the lifetime of the system, so I think it makes sense to add HDMI 2.0 to the list of requirements.
Which all brings me to my current position. If I'm going to get the longest lifespan out of the HTPC, it makes a lot more sense to wait for the Kaby Lake CPUs to drop (fyi, Intel Developers Forum is the 18th-20th of August). They should support HDMI 2.0 along with HDCP 2.2 and a bunch of other stuff that should be a big plus for 4k HTPC users.