Project thief: Haswell Mini-ITX watercooled build

Why did you get the EVGA GTX 670 and not the GTX 760 ? The latter is about 20-30% (370$ vs 250$) cheaper depending on where you live and about as fast.

The socket issue on the Asus H87 is frustrating, they clearly don't want anyone to use anything except the boxed cooler on that motherboard...
 
Why did you get the EVGA GTX 670 and not the GTX 760 ? The latter is about 20-30% (370$ vs 250$) cheaper depending on where you live and about as fast.

The socket issue on the Asus H87 is frustrating, they clearly don't want anyone to use anything except the boxed cooler on that motherboard...

Actually, I've had the 670 for about a year, I just had to throw the box in the mix. If I was spending money now, I'd go with 7xx something for sure. :)
 
Ah OK that explains. Any idea when you'll be getting the Asrock motherboard ?
 
I hope that answers your question. One problem with the Asrock was that I had to turn off Deep Sleep and S5/S6 CPU power modes to fix BSOD 124 under heavy load. Overclocking options are excellent. BIOS update over the internet from *within* BIOS is a new experience for me. :)
 
The problem is, SG08 case has a 180mm fan, and 120mm mounting holes underneath it. The H90 has 140mm fan. Time to do some modding! I initially marked it out wrong, but corrected it later.

014.jpg


This used to be the left door from Lian-Li PC-61B case, black anodized aluminum about 1mm thick. It's been sitting in my garage for about 10 years, and I finally found good use for it.

015.jpg
 
FWIW, almost all of the Asus mITX motherboards and one I have from Asrock (and some of the Gigabyte models) have the same issues with the Corsair-style watercooler mounts. You're not sacrificing integrity by trimming the backplate to fit since you'll still have four points of contact. On one install, I just used nylon nuts to secure the WB since I wasn't planning on removing it very often.
 
FWIW, almost all of the Asus mITX motherboards and one I have from Asrock (and some of the Gigabyte models) have the same issues with the Corsair-style watercooler mounts. You're not sacrificing integrity by trimming the backplate to fit since you'll still have four points of contact. On one install, I just used nylon nuts to secure the WB since I wasn't planning on removing it very often.

Indeed, I could have used nylon washers and nuts to mount the waterblock, and forego the bracket. Or I could have shaved the bracket. But my days of ghetto modding are over, and I take exception to the companies designing products for OEM parts only.
 
Nice build, clean instalation of the H90 , congratulations

Out of curiosity what temps are you getting?

I wonder if you can overclock the 4770k on this board, I was under the impresión that you can increase the max turbo clocks a couple of clicks up when using a unlocked cpu,if you are not overclocking have you tried undervolting ?

keep the good work up

Joe
 
Nice build, clean instalation of the H90 , congratulations

Thank you. :)

Out of curiosity what temps are you getting?

I'm getting 30C idle and about 60-62C load when folding (prime95). It's a bit less when gaming. The fan and pump speeds are set in BIOS right now, starting at 30% and ramping up to 80% at 65C, 100% at 70C. I haven't done any extensive checking yet, haven't had much time.

The CPU clocks down to 800MHz when it's fully idle, and ramps up to 3.9GHz when it's loaded. I tried changing the multiplier per-core in the BIOS, but I had to troubleshoot some issues with sleep (computer was waking up randomly - it turned out to be "Wake on pattern match" option in network adapter, if you can believe it) and some BSODs (BSOD 124, turned out to be S5/S6 sleep mode, I turned those off), and haven't messed with multipliers again.

I wonder if you can overclock the 4770k on this board, I was under the impresión that you can increase the max turbo clocks a couple of clicks up when using a unlocked cpu,if you are not overclocking have you tried undervolting ?

Asrock motherboard comes with A-Tuning software, which will do a simple overclock test. It easily goes up to 4.2GHz, with no sacrifice in quietness. I think if I spent more time on it, I could get it up to 4.3 or 4.4 easy. The overclocking options in BIOS and in A-Tuning software are extensive and easy to use. So far I'm very impressed with this board.
 
Last edited:
So, having read some thermal paste reviews, I've discovered a few new things for myself.

First off, it's not "thermal paste" anymore, it's TIM - thermal interface material.
Secondly, my venerable Arctic Silver 5 is no longer the king of the road - the crown is split between Arctic MX-4 and Tuniq TX-4 (ominously similar names).

Ordered Arctic MX-4 from newegg (it was on sale, to boot), and applied it last night.

First I used the spread method (which still gives me nightmares - think of all the air bubbles!), then replied with the line method as described in the AS5 manual.

My temperatures dropped good 3C degrees, as compared to AS5. Idle dropped to 31 to 28, and load when prime95'ing dropped from 60 to 57. This is with fan and pump in very silent mode.

Core #1 is consistently showing slightly higher temperatures - 28C idle, 57C load. Other cores are lower by 1 or 2 degrees. Core #3 seems to do best. I guess this is due to waterblock design, since the temperatures stayed this way after I reapplied the MX-4. The spread method and line method both yielded the same temperatures, so use whichever one strikes your fancy.

Bottom line: if you have a spare $7, go get yourself a tube of Arctic MX-4.
 
I like it!

I plan doing something similar for my new HTPC build.

I can't help but wonder - though - why the fan blows in rather than sucks out. It would seem counter to the natural flow of things, since hot air tends to want to rise...
 
Zarathustra[H];1040305965 said:
I can't help but wonder - though - why the fan blows in rather than sucks out. It would seem counter to the natural flow of things, since hot air tends to want to rise...

Excellent question, and I'm surprised no one asked it yet. :)

There are two reasons, actually:

1. Noise. Fan blowing into the radiator makes quite a bit more noise as compared to fan sucking out of the radiator. This is the so called plenum effect in play, since there is a small space between the radiator and the fan itself.

2. Dust. If the fan is sucking air out of the case, it creates the negative pressure in the case, and all the dust gets sucked in. If the fan is blowing air into the case, it does so through the filter, and creates positive pressure in the case, so all the dust is blown out.
 
Excellent question, and I'm surprised no one asked it yet. :)

There are two reasons, actually:

1. Noise. Fan blowing into the radiator makes quite a bit more noise as compared to fan sucking out of the radiator. This is the so called plenum effect in play, since there is a small space between the radiator and the fan itself.

2. Dust. If the fan is sucking air out of the case, it creates the negative pressure in the case, and all the dust gets sucked in. If the fan is blowing air into the case, it does so through the filter, and creates positive pressure in the case, so all the dust is blown out.

Interesting. In spite of all this I wonder what the performance effects would be by blowing up.
 
Zarathustra[H];1040307618 said:
Interesting. In spite of all this I wonder what the performance effects would be by blowing up.

- Noise would be a little higher (I can't put the fan on top of the radiator due to fitment issues, so fan would be blowing into the radiator).
- CPU temperature would be a little higher, since it's using a warmer case air to cool the radiator
- There would be more dust in the case. A lot more dust, trust me on that one.

Of course, YMMV.
 
- Noise would be a little higher (I can't put the fan on top of the radiator due to fitment issues, so fan would be blowing into the radiator).
- CPU temperature would be a little higher, since it's using a warmer case air to cool the radiator
- There would be more dust in the case. A lot more dust, trust me on that one.

Of course, YMMV.

The dust could probably be mitigated by taping ones own dist filters over the intake holes in the chassis.

The noise is an interesting thought, one which I was unaware of. So blowing into a radiator causes more noise than sucking out of one, at the same RPMs, right? But what if you could adjust the fan to lower RPM's as it is no longer fighting the natural flow of hot air upwards?

I am inclined to trust Silverstone's design of this case blowing in, rather than sucking out - as they tend to know what they are doing, but at the same time, they are the inventors of the Raven design with the 90 degree rotated motherboard to take advantage of rising heat flow...
 
Zarathustra[H];1040307716 said:
The dust could probably be mitigated by taping ones own dist filters over the intake holes in the chassis.

you can't cover all the holes in the case, dust will get in. look at any case with negative case pressure, and you'll see lots of it everywhere. :)

The noise is an interesting thought, one which I was unaware of. So blowing into a radiator causes more noise than sucking out of one, at the same RPMs, right? But what if you could adjust the fan to lower RPM's as it is no longer fighting the natural flow of hot air upwards?

hot air moving upwards isn't that much of an issue in a tiny case like that. the fan is at like 900 rpm now, going any lower will have a detrimental effect on the cooling.

I am inclined to trust Silverstone's design of this case blowing in, rather than sucking out - as they tend to know what they are doing, but at the same time, they are the inventors of the Raven design with the 90 degree rotated motherboard to take advantage of rising heat flow...

try it both ways (hehe - i mean the fan blowing in and out of the radiator). i preferred to keep the filter on top of the case functional. like i said, YMMV.
 
Back
Top