Will this build with passive cooled i7-4765T work?

New to PC silencing? Read & post your questions here. Dedicated to rosy_toes.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Will this build with passive cooled i7-4765T work?

Post by Vicotnik » Fri Dec 13, 2013 12:47 am

loty1825 wrote:And I wouldn't count on undervolting, because it can cause poor CPU stability. And I suppose Haswells that can be stable at lower voltage probably get labeled as S and T series.
No, no, no. It doesn't matter if you or Intel undervolts. There's nothing special about the S and T series except that Intel has factory underclocked them.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Fri Dec 13, 2013 4:12 am

I wasn't trying to say low TDP versions are any different by design, but could be different by production quality.

I'm not very familiar with Intel's CPU production process, but as far as I understand all CPU's with same architecture, number of cores, threads and cache are basically the same. But not all production runs are equally successful. Out of each production run some samples are tested and according to test results they get badged as particular model.

For example:
If i7 Haswell test samples can operate stable much above 3.9GHz, they will be badged as 4770K, because they have good overclocking potential, if they can operate just above 3.9GHz they will become 4771 or 4770S. If I remember it correctly the Haswells CPU voltage regulator is inside the CPU and by default it's auto-adjusting voltage that corresponds the frequency at which CPU is operating at the moment. Base frequency of 4770S is lower than the 4771, so 4771 will never have to operate at as low voltage as 4770S. So I guess Intel test at how low voltage the CPU is stable and those that are more stable get badged as 4770S.

And then it depends if one gets lucky and gets a good CPU from otherwise not so successful run, then it can be stable at low voltages, otherwise it might be not.

Please don't flame me if I got it wrong. But that's how I imagine it in lamers terms.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Will this build with passive cooled i7-4765T work?

Post by Vicotnik » Fri Dec 13, 2013 4:58 am

It's mostly demand and market segmentation that determines what Intel decides to do with any given CPU. Some cherry picking at the top perhaps, but it's not like Intel is in any rush. When competition is harder or they have problems with the yield for some reason then cherry picking is more common.

edit: Check out the discussed in this thread
viewtopic.php?f=28&t=66782

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Fri Dec 13, 2013 11:37 am

It's called cherry picking? :D A funny term for serious high-tech procedure.

If there's no cherry picking I guess manufacturing procedures are much more reliable than they used to be years ago. It also makes sense then, why only top end CPU model is unlocked for overclocking. If all the CPU models were unlocked, everyone would buy the cheapest one as it would overclock just as as good as the top one.

The bottom end of the thread you posted that link to is that low TDP CPU's are only worth buying if the mobo doesn't support underclocking and undervolting or if the CPU cooler isn't cabable of cooling higher TDP CPU's. And don't buy better CPU than you really need because it's just waste of money.

I'm considering buying Q87 mobo and Q87 chipset doesn't support overclocking and am also considering passive cooling of the CPU. So low TDP CPU might be worth buying for my build. To underclock I'd also have to buy a K CPU, but it lacks some features I might use.

Why Q87? Because of vPro. Why do or will I ever need it? I miss the WOL ability on my current mobo, so it's a must on my next mobo. And remote accessing shut-down system with vPro sounds much much more than just that, so I want to give it a try. I like to try out stuff so I'll be running virtual machines too. With VT-d it's possible to dedicate a piece of hardware to a VM. But K CPU's don't support VT-d. They also lacks TXT. TXT and VT-d are both necessary to run Hyper-V. I won't be running Hyper-V for sure, but some other VM software might also require them. It also lacks TSX-NI, which I don't really understand what it does, but has something to do with multi-threaded aplications accessing memory in better way or sth. like that. So I guess it can't suck to have this ability.

So will I need this stuff? I don't really know before I try it. But I don't want to find myself banging my head to the wall after some time because I didn't get hardware with those features, like it's now because I can't WOL so I have to leave my system powered on every time I leave home, even if I'm away for days. But on the other hand I may also end up banging my head, because I didn't take the K CPU and Z87 mobo.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Sat Dec 14, 2013 2:13 pm

I ordered the G.Skill F3-12800CL9D-8GBSR2 RAM kit today. There RAM isn't in stock, so I'll have to wait 2-10 days to get it. The RAM modules won't fit under CR-95C. I decided that system stability and performance mean more to me than passively cooled CPU. While I wait for the RAM to arrive I have time to think about other parts of the build.

Now I'm considering these parts in the build:

CPU: ?
We'll see when I test my current CPU with more RAM. But it won't be K version for sure. I've been testing VirtualBox in the last couple of days, and I like it very much for it's speed and straight-forward GUI and settings. I will definitely use it. But it requires VT-x to run 64-bit operating systems and with VT-d there's also a cool possibility to passthrough a hardware device to a virtual machine. And K versions don't support that.

Mobo: Asus CS-B
As there won't be a K CPU, over-(or rather under)clocking abilties of Z87 boards won't be of any use. Z87 boards also don't support VT-d. So I'll stick with Q87 being the most feature rich business chipset. I checked it again and in fact Asus CS-B is on QVL of the G.Skill RAM I ordered. I skimmed through manuals of several other boards that are also on the QVL from different manufacturers, but only Asus makes detailed enough manuals, that specify what DRAM voltage range setting it supports and it also lists many tweaks that other boards manuals don't. So it might be most feature rich mobo, or it's just that other manufacturers manuals suck. MSI boards are supposed to have best fan control, but the only thing they say about DRAM Voltage setting is that if it's set to Auto then it will be set automatically and if set to Manual it can be set manually. :lol: So I can't know out of that if I'll be able set the voltage 1.25V. And it's the same with other manufacturer manuals I read. The only second thoughts I have about the CS-B is Asus poor BIOS fan control I read about in some other discussion on these forums. But I guess it could be solved with voltage regulators and fans with very low RPM?

RAM: G.Skill F3-12800CL9D-8GBSR2 1.25V 2x4GB kit
Already ordered it. Nothing to choose anymore.

SSD: Samsung 840 PRO series 256GB 25/S600
Instead PRO series instead of EVO. Expensive, but it's the fastest and that 3000TB test showed it's also durable. Eventhough EVO would probably be fine, it would still bother me to have it in my system after seeing those test results.

HDD: WD WD30EZRX 3TB 64M/S600 GreenPower
Unchanged.

DVD-RW drives: Plextor PX-891SA, Optiarc AD-7241S and LG GH22NP20
I received the Plextor today. :D So I have all three drives now. After some testing I'll decide wether I'll keep just two or all of them.

Sound card: ASUS Xonar Essence STX or 60€ Sound Card with ASIO support and external DAC on SPDIF.
Best option would be external DAC, but I'd need at least cheapest (about 60€) sound card with ASIO support anyway. I can get Xonar Essence STX for about 180€. I did some research on Google and I think it's impossible to get an external DAC that is more decent than DAC built into the Xonar card. For 120€ it's impossible to get a decent DAC at all and I can't afford to add another 200 - 300€ just for DAC. So I think I'll choose the Xonar. And maybe buy an external DAC sometime later.

PSU: ENERMAX EPM500AWT Platimax 500W or FRACTAL DESIGN Newton R3 600W Modular or CORSAIR RM450
It's not that platinum efficiency is neccessary, but I'd rather have it, even if it costs a bit more. And I'd also like the PSU to have modular cabling. I know the Enermax and Fractal design are overkill, but are pretty much only choice I have left (besides Seasonic made) with platinum efficiency, and the Fractal Design is semi-passive. Both cost between 130-140€. The RM450 costs 95€. Any opinions about the Enermax and Fractal Design maybe?

Case: Fractal Design Define XL R2 Titanium Grey
Everyone on these forums seem to recommend the FD R4 model. But since R4 doesn't have enough 5.25" bays I'd choose it's bigger brother.

CPU cooler: ?
Not sure. Probably some tower with 140mm FAN that has very low RPM at 40% PWM signal (the lowest Asus BIOS PWM signal)? But I'd like it not to cover any of the RAM slots, so I can add another RAM kit to the mobo later if needed. Any suggestions?

Case fans: ?
Optional. Will get them later if it shows I need additional cooling.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Wed Jan 08, 2014 3:23 pm

Still waiting for the RAM to arrive, local vendor didn't receive it yet. But today he lent me 2x4GB Crucial Ballistix kit untill my RAM arives. So I could finally test how Audition utilizes CPU with more RAM. And it's no difference. I also analyzed CPU and hard drive usage with xperf. CPU usage is the same with 4GB or 8GB RAM. And hard drive usage is very low, almost none with 8GB RAM, but Audition doesn't finish it's tasks any faster. So neither RAM nor the hard drive seem to be the bottlleneck in my current system. It seems it's the way that Audition works with threads. There's main Audition thread that seems it uses single thread of a single core to the full and there's also other threads that do run simultaneosly on other CPU threads, but they use very little CPU time. So I guess higher single thread performance will give much better performance gain than more cores. I've also found out that Haswell mobos have much higher DPC latency compared to the Ivy Bridge, and that Xonar STX card is native PCI with bridge to PCIe. Xonar ST also has better jitter clock. So I'll buy the ST and I want chipset with native PCI support.

I still want CPU with vt-d capability so there's 2 CPU's left to choose from - i5-3570 or i7-3770:
Image
Those performance increase numbers are performance increase compared to my current i3-2100T CPU according to cpu-world.com benchmarks. Looks like in theory I would still get better single-threaded performance by about 3% if I choose the i7. At a glance spending 50%+ more for 3% performance increase seems not worth it. But I intend to keep the system for a few years and with 3% faster CPU I can probably save 100 hours or so of time. If I look at it from this perspective it seems worth it. But will I really see the 3% performance gain in reality.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Will this build with passive cooled i7-4765T work?

Post by MikeC » Wed Jan 08, 2014 3:55 pm

loty1825 wrote:But will I really see the 3% performance gain in reality.
uh.... no.

We have gone through a ton of hardware in the lab over the years and one thing is perfectly clear: Performances differences less than ~10% are trivial unless you're somehow comparing two systems side by side. It's not easy to "feel" such a difference in the vast majority of cases. Even when the difference is much higher, unless you're in constant production/high load, it's just not significant. I'd probably never make a buying decision on a 3% difference.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Will this build with passive cooled i7-4765T work?

Post by CA_Steve » Wed Jan 08, 2014 5:40 pm

Also, native PCI support went away with Sandy Bridge (ix-2xxx). So, just stay away from PCI cards and go with PCIe.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Wed Jan 08, 2014 11:49 pm

CA_Steve wrote:Also, native PCI support went away with Sandy Bridge (ix-2xxx).
According to this table on Wikipedia business chipsets 6 and 7 series do have PCI, consumer chipsets don't.

Intel also shows legacy PCI support on this diagram.
Image

Is legacy PCI native or does legacy mean that there's a bridge to PCIe?

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Will this build with passive cooled i7-4765T work?

Post by CA_Steve » Thu Jan 09, 2014 8:23 am

I stand corrected. Looks like it's native on those PCH's.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Thu Jan 09, 2014 12:49 pm

So I'll go with an Ivy Bridge i5 and a Q77 mobo.

Besides i5 3570, I can also get 3470 at the local vendor. 3570 is about 5% faster according to cpuworld benchmarks and costs 10% more. So in theory it isn't worth it, but in reality that's just about 15€ (20$) more. So I probably should go with the 3570, right?

For the mobo there isn't much choice, as Q77 mobo's are running out of stock all across the globe. Reading this forums I got an impression that Intel, MSI and perhaps also Asrock have good BIOS fan control. Gigabyte's supposed to be not that good and Asus the worst.

MSI: I don't think MSI has ever even made a Q77 desktop board. Or at least I can't find it on their website or any online shop selling it. It did make some industrial Q77 though, but it's not on my RAM's QVL.

Asrock: Has one model - Q77M vPro, but it seems it's out of stock everywhere I looked.

Gigabyte: Also has just one model also - GA-Q77M-D2H. I read about problems people had with low voltage RAM running on auto settings and it has no options in BIOS to set RAM voltage manually. And supposedly not that great fan control. I don't think I'll go for this one.

Asus: Only lists a single model on their website - P8Q77-M. It has manual DRAM voltage control and it can be set even lower than 1.25V. It seems to have great BIOS with many options. But the fan control sucks. It comes with Asus FanExpert (not 2 though). I read somewhere that FanExpert just writes settings to some on-board chip on Windows start-up and then it can be closed, because the on-board chip controls the fans, not the software. If that's is indeed so it might be not that bad afterall. What I also don't like about this board is not all the caps are solid state.

Intel: Has two boards DQ77MK and DQ77CP. They are similar boards. The DQ77MK costs a bit more, has two NIC's, one more SATA port and PCIe slot and one less PCI port. Both would suit my needs, but DQ77CP seems harder to get. Both have options to manually adjust RAM voltage from 1.2V to 1.8V. What I also like very much about Intel boards is the most thorough techincal documentation of all the board makers. But I also don't like that on both boards not all caps are solid.

Right now I'd choose one of the Intel boards or the Asus board, if fans are really controlled by an on-board chip.

Any comments more than welcome. Thanx.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Fri Jan 10, 2014 9:43 am

I found out that the Asus P8Q77-M case fans can't be controlled by Q-Fan 2 and fan xpert. So I'll acquire Intel DQ77MK. All the fan headers on the Intel board are PWM. So I'll need a heatsink with PWM fan(s) and PWM case fans.

I'm thinking of getting a heatsink that performs well with dual 140mm fans as well as with a single fan. This would give me some versatility because I could add/remove second fan if desired. The heatsink also shouldn't cover any RAM slots. Considering the price and performance HR-02 macho seemed to be best choice at first, but it's stock fan isn't that good, so it would add some extra cost to replace the fan. And it doesn't come with the clip for the second fan, and separate fan clips seem very hard to get (and no I don't want to use zip ties instead). So I'm thinking of buying Noctua NH-U14S. It's pricey but it has good stock fan. Taken that into account it's not much more expensive than the HR-02. And it also comes with clip for the second fan. For the second fan I intend to buy another NF-A15 PWM. As for the case fans probably another pair of NF-A15 PWMs. Some Scythe 140mm PWM models might cost less, but they're harder to find where I live than Noctuas.

And for PSU I'll give Fractal Design Newton R3 600W a try. It's overkill but it's the lowest power R3 model. I chose it over RM450 for it's better power efficiency. And if I make all the correct settings in BIOS and it still whines, I can return it under warranty.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Will this build with passive cooled i7-4765T work?

Post by Vicotnik » Fri Jan 10, 2014 3:23 pm

I wouldn't bother with a second fan on the CPU cooler. Only makes sense if you overclock or use one of the very high end CPUs, maybe not even then.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Fri Jan 10, 2014 5:11 pm

loty1825 wrote:I'm thinking of getting a heatsink that performs well with dual 140mm fans as well as with a single fan. This would give me some versatility because I could add/remove second fan if desired. The heatsink also shouldn't cover any RAM slots. Considering the price and performance HR-02 macho seemed to be best choice at first, but it's stock fan isn't that good, so it would add some extra cost to replace the fan. And it doesn't come with the clip for the second fan, and separate fan clips seem very hard to get (and no I don't want to use zip ties instead). So I'm thinking of buying Noctua NH-U14S. It's pricey but it has good stock fan. Taken that into account it's not much more expensive than the HR-02. And it also comes with clip for the second fan. For the second fan I intend to buy another NF-A15 PWM. As for the case fans probably another pair of NF-A15 PWMs. Some Scythe 140mm PWM models might cost less, but they're harder to find where I live than Noctuas.
I woudlnt go dual, one fan is more than enough for this cooler, doesnt get as recommended as the Mugen4 or HR02 for the price, but its a great cooler as it is.

Now on the second NF-A15 PWM be careful, as the one included on the cooler is not the same as the retail, the fan is more like the NF-A14 PWM retail, it goes higher on rpms all accross the board, to make the NF-A15 PWM included on the U14S similar to the retail version you have to use the ULNA supplied with the fan (or maybe with the cooler also), this will make it very similar in range and scaling up or down. Personally i dont think you need it, there will be gains, but imo not worth the extra $20.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Fri Jan 10, 2014 10:50 pm

I don't want Mugen4 because it covers a RAM slot. HR-02's stock fan isn't that great. If I replace the stock fan with NF-A15 it would cost 47,18€ for the cooler + 23,01€ for the fan, so 70,19€ total. NH-U14S costs 71,55€. So for almost same price I get second fan clip, ULN adapter and Y-splitter cable that HR-02 doesn't come with.

I don't expect the CPU to produce so much heat, that a single fan cooler wouldn't be able to properly cool it down. But since all the PWM fans in the system will be throttled according to the CPU temperature, I was thinking that keeping the CPU temperature a few degrees lower would make case fans spin at lower speed. Besides if I'm spending 1500€+ for the system, another 20€ to test dual fan CPU cooling won't kill me. And if it turns out I don't need a second fan, I'll be more than happy to find some use for it in some other system.

Any opinions about NF-A15 as case fans?

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Will this build with passive cooled i7-4765T work?

Post by quest_for_silence » Fri Jan 10, 2014 11:12 pm

loty1825 wrote:But since all the PWM fans in the system will be throttled according to the CPU temperature, I was thinking that keeping the CPU temperature a few degrees lower would make case fans spin at lower speed.


(I would be glad to be wrong, but) it won't be the case, or so I guess.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Fri Jan 10, 2014 11:16 pm

loty1825 wrote:Any opinions about NF-A15 as case fans?
On Define R4 or XL R2 the frontal fans are 140 or 120, but the NF-A15 are not standard 140, it should not fit on the front without some modding (and even then im not sure), the correct fans if you want noctua and PWM are the NF-A14 PWM, they are standard 140mm fans on standard frame.
loty1825 wrote:I don't want Mugen4 because it covers a RAM slot. HR-02's stock fan isn't that great. If I replace the stock fan with NF-A15 it would cost 47,18€ for the cooler + 23,01€ for the fan, so 70,19€ total. NH-U14S costs 71,55€. So for almost same price I get second fan clip, ULN adapter and Y-splitter cable that HR-02 doesn't come with.
From SPCR Noctua NH-U14S Slim 140mm Tower Cooler
While going from one to two of our reference Noctua 140 mm fans doesn't net the NH-U14S much improvement
If you want go with 2, but personally i wouldnt. On the other hand the HR02 does benefit more from a dual fan setup, still i think for the CPU you are going with U14S is a good choice, but i would go single and with the fan as it is without ULNA, and good mobo that can drop it down.

Btw all haswell CPUs underclock to 800mhz if you allow it on the power management, and this makes them idle very low, in my case they idle at 30C, not much difference you will get with muliple fans at very low rpms, its mostly the CPU that doesnt generate much heat at this state, where haswell becomes a demon is in load... where i can barely sustain good temperature under stress, but your choice of CPU wont go as high since its factory nerfed so you should be fine.

Btw i still have picture of the my Gskill Sniper 1.25V, its not that tall, here it is with HR02 behind,

Image

Here is a couple of pics of the HR02 with NF-A15 PWM so you get some rough idea into what to expect (remember the HR02 is asymmetrical and its flush to the CPU socket).

Image

Image

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Will this build with passive cooled i7-4765T work?

Post by quest_for_silence » Sat Jan 11, 2014 12:40 am

Abula wrote:Btw all haswell CPUs underclock to 800mhz


Isn't he on Ivy, eventually?

Anyway, you know, a dual fan setup won't give him any considerable/useful temperature advantage, either at idle or load, and it won't give such less noise as he think, period: any other reasoning about, any other choice, it's just to have fun.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Sat Jan 11, 2014 6:10 am

quest_for_silence wrote:
Abula wrote:Btw all haswell CPUs underclock to 800mhz


Isn't he on Ivy, eventually?
My bad, because of the cpu on the title of the thread i mixed it.

Pappnaas
Posts: 726
Joined: Mon May 16, 2011 11:23 am
Location: Germany

Re: Will this build with passive cooled i7-4765T work?

Post by Pappnaas » Sat Jan 11, 2014 6:57 am

Regarding CPU options:

Have you compared the new Xeon E3-12XX V3 series to your planned i5/i7 modell? Could be worth a look.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Will this build with passive cooled i7-4765T work?

Post by CA_Steve » Sat Jan 11, 2014 8:19 am

Pappnaas wrote:Regarding CPU options:

Have you compared the new Xeon E3-12XX V3 series to your planned i5/i7 modell? Could be worth a look.
He's going with IVB for the lower DPC latency.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Sun Jan 12, 2014 6:24 am

Abula wrote:On Define R4 or XL R2 the frontal fans are 140 or 120, but the NF-A15 are not standard 140, it should not fit on the front without some modding (and even then im not sure), the correct fans if you want noctua and PWM are the NF-A14 PWM, they are standard 140mm fans on standard frame.
NF-A15 is 140mmx150mm with 120mm mounting holes. It looks like XL R2 has mounting holes for 120mm. Are you sure NF-A15 won't fit? It looks like it fits at the back, but at the front it looks very tight to fit two NF-A15s, but I guess a single fan might fit?

ImageImage
from SPCR review wrote:The A15 has a smoother acoustic profile than the A14 but both sound fairly good to the ears.
They're the same price, so I'd rather choose A15 if they fit.
Abula wrote:Here is a couple of pics of the HR02 with NF-A15 PWM so you get some rough idea into what to expect (remember the HR02 is asymmetrical and its flush to the CPU socket).

Image

Image
Is this the stock HR-02 clips you are using? I thought NF-A15 might not fit with the stock clips because it's 25mm thick, the HR-02 stock fan is 26.5mm thick. If I go with single fan and NF-A15 fits Thermalright clips should I also consider Thermalright Archon SB-E with NF-A15 fan?

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Sun Jan 12, 2014 11:42 am

loty1825 wrote:NF-A15 is 140mmx150mm with 120mm mounting holes. It looks like XL R2 has mounting holes for 120mm. Are you sure NF-A15 won't fit? It looks like it fits at the back, but at the front it looks very tight to fit two NF-A15s, but I guess a single fan might fit?
The NF-A15 is a 140mm fan on a custom frame with 120mm holes, the holes will match a 120mm fan mounting, but the outside is wider, so the question that i cant answer you is if the fans will fit on the plastic R4 frame, the NF-A14 bearly fits, the fan cage is made exactly for 140s, if the frame is longer might not fit.
loty1825 wrote:
from SPCR review wrote:The A15 has a smoother acoustic profile than the A14 but both sound fairly good to the ears.
They're the same price, so I'd rather choose A15 if they fit.
If the sonic signature is what you are after, then forget noctuas, all their 140s were worst than the SPCR Noctua reference 140mm fan, there are even at least 3 better fans, Antec True Quiet 140, Antec two cool 140 and Phantek 140, that said the reason for me to persue noctua was their superb PWM design and their ultra long range of control, i can drop them to 200rpms or ramp them to 1500rpms depending on the conditions of my CPU, but the fans are loud at high rpms, i can notice them above 650rpms easily, so i have restricted via bios all this settings at the cost of higher temps, but this are choices that i decided on advance, specially the motherboard control of the fans. For me chosing 140mm noctuas should be because of their design, not their tonal quality.

Personlly im very intergued by two 140mm PWM fans that i will try out in the next months, since SPCR gave such good review to the standard phantek 140, i want to try their new true PWM version, Phanteks 140mm Case/Radiator Cooling Fan (PH-F140XP_BK), with the QSA adapter according to phantek their PWM range of operation lowers to 300-900rpms, which to me is acceptable, although i still need to check how and what % each is to work with, some fans are design very weird on PWM. The second fan that i want to test is Noiseblocker NB-BlackSilentPro PK-PS 140mm x 25mm Ultra Quiet PWM Fan, i have very good experience with a PK1, really liked how it sounded, but from past Noiseblocker PWM fans (PLPS), they design their PWM to reach till 0%, so to drop them as low as they are rated, you need a motherboard that can get close to 0%, this for me is not an option on MSI, as the restriction is 12.5%, but again all this is speculation i dont own the fan, and this is why i want to test it, really wish SPCR would consider both fans on a future 140mm roundup, for me its all about PWM fans from now on, this is why i didn't go with the true quiet, but will see in the next build upon testing the fans.
loty1825 wrote:Is this the stock HR-02 clips you are using? I thought NF-A15 might not fit with the stock clips because it's 25mm thick, the HR-02 stock fan is 26.5mm thick. If I go with single fan and NF-A15 fits Thermalright clips should I also consider Thermalright Archon SB-E with NF-A15 fan?
Yes, those are the stock or what the original (not the macho) came with, the NF-A15 and TY147 are extremely similar, i would say its a design copy from noctua with some other changes, but where ever the TY147 fits the NF-A15 should fit. I also did a build for a friend on MACHO version for you to see, the clips now included mount on middle of the frame (not like mine over them), but i honestly don't think there will be an issue but up to you to crosscheck, ill leave you a pic so you get a better idea.

Image

Just one last thing that im wondering about your previous post,
loty1825 wrote:I found out that the Asus P8Q77-M case fans can't be controlled by Q-Fan 2 and fan xpert. So I'll acquire Intel DQ77MK. All the fan headers on the Intel board are PWM. So I'll need a heatsink with PWM fan(s) and PWM case fans.
I dont understand the Q-FAN 2 remark, can you elaborate into what you found out that made you go into intel? If you want the quietest setup, with the best signature of fans atm on 140s, i would go with Antec True Quiet 140 (3 is what i would chose), and whatever you like on CPU cooler, the R4 and XL R2 will fit any cooler, so whatever you prefer, from HR02 Macho (if you want a NF-A15 PWM change), a Scythe Mugen 4 (like CA_Steve build) or Noctua NH-U14S (that already comes with NF-A15PWM), or whatever else you like, but for this to work, you will need FanXpert and let it drop the all the case fans to their minium, that said all this fans will drop to really low rpms in this setup, like 250rpms or so, so with that minium airflow you will get high temps, but all of this is tweakable on the graphs to your liking, and even if FanXpert doesn't work out for you, you can use the included fan controller on the R4 or XL R2 to undervolt the Antec True quiet, they also have a switch that you can use to undervolt them... a lot of options, personally i still prefer fanxpert2 and let it be dynamically controlled, but i don't understand why the motherboard cant, according to your post.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Sun Jan 12, 2014 2:38 pm

It's not sonic signature I'm after specifically, but rather I want the fans to be virtually inaudible at say 0.5m distance when CPU is at 25% load. If that's possible then sonic signature doesn't matter, because I won't hear them. If it isn't then sonic signature matters, because I want the fans to have as little annoying tones as possible.

My mobo will be Intel DQ77MK, with all three fan headers PWM controlled. I definitely do want the CPU fan to be PWM controlled. I could ran the case fans undervolted at constant speed, but I guess it would be better if the case fans were also PWM controlled. Antec 140 fans aren't PWM controlled. I know a Phanteks 140 fan received very good review on SPCR, but that model also isn't PWM. So I've been looking into Phanteks 140 fans. And there's two PWM models PH-F140XP and PH-F140HP which looks similar to the one reviewed. But according to specs they can drop to about 600RPM, which is higher than Noctuas. That's why I've been looking more into Noctuas. The review of Noctua fans also says both A15 and A14 are worse than P14. I already checked if NF-P14 PWM are still available and it seems they aren't.

I know how will the fans drop depends on the mobo, so it might drop Noctua lower than Phanteks or not. I didn't find it mentioned anywhere in the documentation what is the lowest %. The mobo doesn't have options to set the curve manually with a graph, like some mobos do. But it has options in BIOS to set the CPU heatsink is better than the stock Intel, so it spins CPU fan at lower speed. It's possible to set 3 levels of fan loudness for all three fans. I'm not 100% sure but I think it's also possible to set temperature above which the fan will spin at 100%. And it's also possible to set temperature below which the fans don't run at all, so they stop spinning when they're not needed. The mobo uses NCT6776F to monitor temperaures and control fans. NCT6776F supports SMART FAN I, SMART FAN IV and Close-Loop Fan Control RPM mode if that tells anything. But even documentation for NCT6776F doesn't say anything specifically about PWM curves.

Noiseblocker PK-PS has good RPM range according to specs, and some Noiseblocker fans received a very good review, so it might be a great fan.
Abula wrote:Yes, those are the stock or what the original (not the macho) came with
I don't quite understand. If it's not macho the original they came with, what was the original? And I'm not sure NF-A15 could be mounted with the clip on the second pic, because it doesn't have that notch for the clip on the side.
Abula wrote:I dont understand the Q-FAN 2 remark, can you elaborate into what you found out that made you go into intel?
To be honest I didn't dig into details of Q-FAN 2.

From Asus website about Q-Fan 2:
ASUS Q-Fan 2 technology intelligently adjusts both CPU fan and chassis fan speeds according to system loading to ensure quiet, cool and efficient operation.

I guess it's the name Asus uses for fan control on it's mobos.

And from the mobo manual:
Only the 4-pin CPU fan supports the ASUS FanXpert and Q-Fan 2 feature.
Seems obvious that the only fan that's controlled is the CPU fan.
If you want the quietest setup, with the best signature of fans atm on 140s, i would go with Antec True Quiet 140 (3 is what i would chose), and whatever you like on CPU cooler, the R4 and XL R2 will fit any cooler, so whatever you prefer, from HR02 Macho (if you want a NF-A15 PWM change), a Scythe Mugen 4 (like CA_Steve build) or Noctua NH-U14S (that already comes with NF-A15PWM), or whatever else you like, but for this to work, you will need FanXpert and let it drop the all the case fans to their minium, that said all this fans will drop to really low rpms in this setup, like 250rpms or so, so with that minium airflow you will get high temps, but all of this is tweakable on the graphs to your liking, and even if FanXpert doesn't work out for you, you can use the included fan controller on the R4 or XL R2 to undervolt the Antec True quiet, they also have a switch that you can use to undervolt them... a lot of options, personally i still prefer fanxpert2 and let it be dynamically controlled, but i don't understand why the motherboard cant, according to your post.
I've considered using 3-pin case fans and undervolting them with the included case controller. But the Intel mobo, eventhough there's no graph, has quite some options to set up fan control, and it even stops the fans below manually set temperature threshold, so I can't see how undervolted fans spinning at constant ramped-up speed will give better results than fans that don't spin when not needed and ramped up when needed.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Will this build with passive cooled i7-4765T work?

Post by quest_for_silence » Sun Jan 12, 2014 4:11 pm

loty1825 wrote:It's not sonic signature I'm after specifically, but rather I want the fans to be virtually inaudible at say 0.5m distance when CPU is at 25% load. If that's possible then sonic signature doesn't matter, because I won't hear them. If it isn't then sonic signature matters, because I want the fans to have as little annoying tones as possible.


I beg your pardon, but I think you don't know not even what it is a "sonic signature".

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Sun Jan 12, 2014 8:07 pm

loty1825 wrote:I know how will the fans drop depends on the mobo, so it might drop Noctua lower than Phanteks or not. I didn't find it mentioned anywhere in the documentation what is the lowest %. The mobo doesn't have options to set the curve manually with a graph, like some mobos do. But it has options in BIOS to set the CPU heatsink is better than the stock Intel, so it spins CPU fan at lower speed. It's possible to set 3 levels of fan loudness for all three fans. I'm not 100% sure but I think it's also possible to set temperature above which the fan will spin at 100%. And it's also possible to set temperature below which the fans don't run at all, so they stop spinning when they're not needed. The mobo uses NCT6776F to monitor temperaures and control fans. NCT6776F supports SMART FAN I, SMART FAN IV and Close-Loop Fan Control RPM mode if that tells anything. But even documentation for NCT6776F doesn't say anything specifically about PWM curves.
You can draw the minimums for each on the following fanXpert2 tests on the both noctuas you are looking,

Image

Image

Intel bios fan control is fine, i have work with 2 mobos from sandy bridge and it can drop down very low Noctua and Scythe fans (check signature for Download station and HPTC), but its kinda tricky like % of PWM are not the same in all headers, but dont worry at the end is just playing with numbers and rebooting and checking the rpms you end up, but no where as good as MSI (and lately seem Asrock have the best bios PWM fan control) is very nice on pure bios giving you the minimum and max and they do behave as the % pwm as it should.
loty1825 wrote:I don't quite understand. If it's not macho the original they came with, what was the original? And I'm not sure NF-A15 could be mounted with the clip on the second pic, because it doesn't have that notch for the clip on the side.
The HR02 was relesased like that HR02, it was nickel plated (as you can see on my pictures) and had like two or three more fins (dont quote me on this i never really counted them), a cpu preasure point knob, etc, all of this were taken out when the Macho version was relesaed to save money and have a more competitive price.
loty1825 wrote:To be honest I didn't dig into details of Q-FAN 2.

From Asus website about Q-Fan 2:
Quote:
ASUS Q-Fan 2 technology intelligently adjusts both CPU fan and chassis fan speeds according to system loading to ensure quiet, cool and efficient operation.

I guess it's the name Asus uses for fan control on it's mobos.

And from the mobo manual:
Quote:
Only the 4-pin CPU fan supports the ASUS FanXpert and Q-Fan 2 feature.
Seems obvious that the only fan that's controlled is the CPU fan.
To what i remember Q-FAN is just a way to switch voltage control (advance) to pwm control (auto), this is only available on the CPU header, the rest even though they have Q-FAN you cant make them PWM as they are fake 4pin headers, or at least thats the way asus design most of their motherboards. Now on mine i can control any 3pin fan on CHA_FAN header, and can control a 4pin PWM fan on those header aswell but undervolting it not via PWM control.

I didn't stay on Asus mostly out of 2 things, the way they measure CPU temp doesn't match anything else, and their 4pin fake chassis headers, but aside form that i do like a lot their FanXpert2 software.

Either way i think you should be fine with Intel mobo, not trying to sell you Asus, just you will have to do some tweaking and playing around with it until you get it the way you like it.

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Sun Jan 12, 2014 9:38 pm

Abula wrote:You can draw the minimums for each on the following fanXpert2 tests on the both noctuas you are looking
What mobo were those fans on?
Abula wrote:but no where as good as MSI (and lately seem Asrock have the best bios PWM fan control)
I considered MSI And Asrock boards, but MSI doesn't make Q77 boards, Asrock has one model, but it's out of stock everywhere I look, and being an old model I suppose it's not being made anymore, so I won't be able to acquire it.
Abula wrote:To what i remember Q-FAN is just a way to switch voltage control (advance) to pwm control (auto), this is only available on the CPU header, the rest even though they have Q-FAN you cant make them PWM as they are fake 4pin headers, or at least thats the way asus design most of their motherboards.
The P8Q77-M has 3-pin fan case fan headers, not the fake 4-pin type. I took a look at the BIOS settings in the manual and there are some chassis fan speed controls: Chassis Upper Temperature 40°C to 90°C and Chassis Fan Max Duty cycle 60% to 100%. So it does have some control over case fans, but that's it. The manual doesn't have any instructions on using FanXpert. And from the quote I posted before I still understand that case fans can't be controlled with the FanXpert. And there realy isn't much I can set in BIOS, I think Intel BIOS control is better. Even if Asus could control the fans with FanXpert, it's just FanXpert, not FanXpert 2. I don't know how much they differ, does the FanXpert have all the possibilities of the FanXpert2?
Abula wrote:I didn't stay on Asus mostly out of 2 things, the way they measure CPU temp doesn't match anything else
You mean the CPU temperature readings are way off?

loty1825
Posts: 29
Joined: Sat Dec 07, 2013 7:03 am

Re: Will this build with passive cooled i7-4765T work?

Post by loty1825 » Sun Jan 12, 2014 9:40 pm

quest_for_silence wrote:I beg your pardon, but I think you don't know not even what it is a "sonic signature".
Maybe I really don't, please enlighten me.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Will this build with passive cooled i7-4765T work?

Post by Abula » Sun Jan 12, 2014 9:55 pm

loty1825 wrote:
Abula wrote:You can draw the minimums for each on the following fanXpert2 tests on the both noctuas you are looking
What mobo were those fans on?
Asus Maximus VI GENE, my old haswell build that just became a fan testing PC =).
loty1825 wrote:
Abula wrote:I didn't stay on Asus mostly out of 2 things, the way they measure CPU temp doesn't match anything else
You mean the CPU temperature readings are way off?
Yes and no, Asus has very weird way of measuring CPU temps, its not core temps, its not package or any other that other probes have (some say thats the same of AIDA64 but in my test i didnt match either). Asus in the marketing scheme, they say they use a lot of sensors to calculate a more precise temperature reading of the CPU, the problem is that on idle its lower than core temps and on load there is a breaking point where it start to sky rocket, its like its exponential.... idk really its just very weird the way they do it, weather its more accurate idk.

Go with intel, personally i have had very good experience with them, just not as tweakable as other manufacturers, but they are very reliable and work very well, their Bios fan control is pretty basic but it works fine once you play with it.

Good luck with your build.

Post Reply