NCASE M1: a crowdfunded Mini-ITX case (updates in first post)

To change the axis of debate, can we make it 5mm WIDER?

I still really really want to go dual H80i. :(

I could care less about the 20mm vertically.
 
You see what happens when you start giving in to special demands, before you know it the case has the size of the prodigy because everyone wants to fit their own special hardware in there.
 
To change the axis of debate, can we make it 5mm WIDER?

I still really really want to go dual H80i. :(

I could care less about the 20mm vertically.

Use these instead if the stock fans: http://www.frozencpu.com/products/7...peed_D12SM-12C.html?tl=g36c15s771&id=VsU22yFN

Probably not as much pressure as stock but it will fit without increasing the foot print of the case.
You see what happens when you start giving in to special demands, before you know it the case has the size of the prodigy because everyone wants to fit their own special hardware in there.

Let's exaggerate more please. Even if they added 50mm to all 3 dimensions it would still be smaller than the prodigy and fit more stuff.
 
Last edited:
However, since the size came into doubt nobody can know for sure which is the best compromise. My suspicion is that there would be significant, measurable improvements in all build types, but variable benefit dependent on setup:

  1. AIO water CPU only - probably least benefit since the rads are getting clean air anyway.
  2. water on CPU & GPU - better OCing potential and lower overall noise.
  3. standard / top down air cooling - substantial improvements to temps and noise.
  4. air cooling with short towers / NT01-PRO - very large improvement.
  5. 'silent' air cooled with single fan venting upwards (no side fans) - extremely large improvement.
Hard data to back up or debunk the speculation would be compelling

I would be interested in a real test, but i just dont think adding fans to the top will really help based on where the cpu would be at, and where the fans are located. Now obviously it depends on the setup. I don't think any water cooled people will notice. Top down people may notice. Tower cpu cooler people would want the cooler oriented front to back so its either getting cooler air from the back vent or exhausting out it, because the gpu is in the way on most itx boards so they probably wouldn'tnotice fans above either...

It wouldn't be hard for necere or wahaha to just mount fans on top and see what they do for temps.
 
I told myself I was done with watercooling... and sold all my stuff dirt cheap, but I just ordered a Swiftech H220 just for this case.. so please don't change anything :p
 
FWIW, I think adding 20mm to the height of the case is worth it if it means adding more ATX PSU options. Even if SFX is more appropriate to highly optimized SFF builds the M1 was designed for, ATX is more familiar to builders (like me) who don't have as much experience with the smallest SFF builds. While I personally am willing to go SFX if necessary to build a machine in this case, I think more ATX PSU options will help attract more people to the production campaign.

At any rate, I'm looking forward to more test results from the prototype, and however it turns out I'm planning on backing the production campaign. Thanks to Necere and W360 for turning this awesome case into a reality!
 
Use these instead if the stock fans: http://www.frozencpu.com/products/7...peed_D12SM-12C.html?tl=g36c15s771&id=VsU22yFN

Probably not as much pressure as stock but it will fit without increasing the foot print of the case.

Meh, H80i already comes with 2 fans. Plus, there's no point of buying 2 more if they're not going to perform as well.

I think the best consolation for me would be to have a full H80i on the left mount above the cpu (for GPU cooling), and another H80i with a single 120x38 on the right mount for CPU cooling. I think this setup maximizes the space for tubing and the single 38mm fan would probably perform just as well as dual 25mm, albeit with a high dB.

Either way, dual H80i is going to happen.
 
FWIW, I think adding 20mm to the height of the case is worth it if it means adding more ATX PSU options. Even if SFX is more appropriate to highly optimized SFF builds the M1 was designed for, ATX is more familiar to builders (like me) who don't have as much experience with the smallest SFF builds. While I personally am willing to go SFX if necessary to build a machine in this case, I think more ATX PSU options will help attract more people to the production campaign.
You do realise that SFX is just smaller than ATX, nothing else changes ? It doesn't work differently, it's just more compact.
 
I would argue that it's the direction and velocity of the airflow that makes the difference, not the pressure.

But you're using fans. So regardless of whether you're pulling more air through the case than you're pushing in (negative pressure situation) or vice versa, you're still forcing air (and the dust floating in that air) through the case. In either case, you're probably going to end up with the most dust at the point where the air is being pulled in, be that case vent holes or the inlet side of fans. If anyone could prove scientifically that one method ends up with significantly more dust in the case over time than the other, I'd be shocked.

I have to agree with you. I don't think the pressure matters diddly squat, it's the direction and where the air flows. I could see you getting different airflow and turbulence patterns with the air going one way vs the other, and that could make a difference. In terms of physics, positive pressure has more ABILITY to remove heat as there is more air (in terms of mass) which means more capability to remove heat. However the difference, as you mention, is almost nothing.

However as far as dust with positive vs negative pressure I think the main difference is that with a positive pressure setup the air is only coming into the case at one point, where the fan is. In that case you can go ahead and put a filter on the fan and stop the dust there. Where as with a negative setup air is coming in every nook and cranny, and you can't really filter it all very easily. Besides that, though, I doubt there is much difference in terms of dust.
 
I have to agree with you. I don't think the pressure matters diddly squat, it's the direction and where the air flows. I could see you getting different airflow and turbulence patterns with the air going one way vs the other, and that could make a difference. In terms of physics, positive pressure has more ABILITY to remove heat as there is more air (in terms of mass) which means more capability to remove heat. However the difference, as you mention, is almost nothing.

However as far as dust with positive vs negative pressure I think the main difference is that with a positive pressure setup the air is only coming into the case at one point, where the fan is. In that case you can go ahead and put a filter on the fan and stop the dust there. Where as with a negative setup air is coming in every nook and cranny, and you can't really filter it all very easily. Besides that, though, I doubt there is much difference in terms of dust.


It heavily depends on case design and fan locations. Its not a universal thing. In my oppinion this case would benefit most from a fans on bottom as intake and fans on the side above the cpu as intake, and all exhausted out the top and back...

This shows it can make a difference, but again it more depends on setup and fan location... LINK
 
This shows it can make a difference, but again it more depends on setup and fan location... LINK

Interesting link, but those tests were done in a large case (55 liters) with large empty spaces. The effect in a 12-14 liter case chock full of hardware is quite different.

For those saying "pressure doesn't matter, only airflow". How do you think the airflow happens? It's like wind: movement of air from a high pressure area to a low pressure area. In a case that's 50cm deep a single fan will only reach 30cm before the forward force is disrupted. But in a case that's only 25cm long a single fan will reach all corners.

Anyway, once we all get our cases, we can experiment and show actual results rather than speculating :)
 
If I understand it correctly, we should prefer high-pressure fans inside our very small cases, because otherwise it won't reach the obstructed areas. And we want positive pressure in our case so our case doesn't become a slow vacuum cleaner.

I once had a negative pressure case that had PILES of dust, literally cones, forming at some small openings. Just remember that fan filters block some % of air-flow, so if you use them, be sure to calculate and test the right setup.
 
I want this case so badly :p. I check this thread at least once a day for any updates. Anyone else addicted to this case?
 
If I understand it correctly, we should prefer high-pressure fans inside our very small cases, because otherwise it won't reach the obstructed areas. And we want positive pressure in our case so our case doesn't become a slow vacuum cleaner.

I once had a negative pressure case that had PILES of dust, literally cones, forming at some small openings. Just remember that fan filters block some % of air-flow, so if you use them, be sure to calculate and test the right setup.

Not really. High pressure fans are required for highly restrictive environments... like a heatsink or radiator.
 
You do realise that SFX is just smaller than ATX, nothing else changes ? It doesn't work differently, it's just more compact.

Certainly, but there are currently only two SFX units that are rated at enough power to support high-end hardware, both from Silverstone. While these units will provide enough power to run a CPU and high-end graphics card, they aren't above the "recommended" levels for most of that hardware; I worry that people who want more power headroom or are stubborn about brands will be turned off from the M1 because of the limited PSU options.
 
This will be the first part of the thermal and noise testing, covering a reference blower-style GPU and single 120mm closed-loop liquid cooler.


For stress testing I ran Furmark for 20 minutes, followed by 40 minutes more of Furmark plus Prime95. The CPU was left at stock clocks, GPU overclocked to varying degrees.

For testing under more typical usage scenarios, I played some extended sessions of Warframe and Mechwarrior Online. The former only put 20-30% load on the GPU, while the latter tended to push it harder, at between 80-90%.

Temps and fan speeds were monitored with GPU-Z, Speedfan, and Realtemp.


Configuration 1a

  • ECS H61H2-MV Mini-DTX motherboard
  • i7 3770K
  • Corsair H60i
  • Reference HD7970
  • ASUS Xonar DX
  • SilverStone ST45SF-G
  • 2x 3.5" HDDs
  • 1x 2.5" SSD
  • Slim Optical


This first configuration is pretty low airflow, with only the single 120mm fan on the Corsair liquid cooler. The GPU and PSU fans are left to fend for themselves, with no case fans to help them out. Lian Li fan filters were installed both on the H60i's fan and at the two bottom fan mounts (despite no case fans present at the bottom). The MDTX board puts the GPU one slot down and allows for an additional card to be installed (in this case I used an ASUS Xonar DX).


Config 1a thermals

Furmark
7970@1100/1500
21C ambient
88C GPU after 20min, 55% fan

Furmark + Prime95
7970@1125/1575
21C ambient
88-90C GPU after 40min, 57% fan
CPU max 67/71/63/67C



Even with the overclock on the GPU, the system remained stable for the hour+ of stress testing and exhibited no unusual behavior.

The GPU temps were fairly high, although it seems the reference card's fan profile is set to keep it under 90C or so, as the highest fan speed I saw was only 60% (which is nevertheless quite loud).

CPU temps seemed fine as well (still ~40C to TJ max). They could've been lower using a more aggressive fan setting (see below), but I didn't see the need.

Under gaming conditions it varied a bit, but 70-80C on the GPU @ 40-45% fan speed was fairly typical

In terms of the case itself, the areas that got the warmest were the right side panel, the top vents in the vicinity of the PSU, and the rear/top right corner of the case. Nothing got hot to the touch, but the entire right side of the case was fairly warm after an hour of stress testing. The air coming out of the top (and the vents themselves) over the motherboard didn't get that warm, comparatively, so it's not clear that a top fan would be helpful (in this configuration, at least). My conclusion is that a more useful addition may be to ventilate the right side panel, both to allow some of the hot air trapped behind the motherboard tray to get out, and possibly to allow the PSU to be flipped to draw in outside air.



Config 1a noise

This will be as much an impression of the particular components used in the test as it will be of the case itself.

At idle, the system was fairly quiet in this configuration. The H60i's fan was plugged into the ECS motherboard, which kept it running at ~1k RPM. The 7970's fan ran at 20%, which is quiet to my ears. I found the SilverStone ST45SF-G likewise quiet at idle. I found the most dominant source of noise to be the (still fairly quiet) whir of the H60i's pump.

Under load, the 7970's blower ramps up and dominates everything. It is loud, no doubt about it. Under gaming conditions it averaged 40-45%, which is tolerable if headphones are worn while gaming (which I do). Under stress testing, the GPU's blower went up to 60%, at which point headphones are no longer adequate. Mercifully, the sound is mainly airflow, rather than tonal, which makes it (slightly) more bearable. I suspect though that the reference 7970 is just a loud card, and will perform similarly in any case.

I had the ECS board set to the quiet fan setting, so even after 45 minutes of Prime95 it never saw fit to raise the H60i's fan above 1000 RPM. Consequently, it wasn't a significant source of noise. I also ran only Prime95, with no load on the GPU, and the system noise did not increase appreciably over idle.

The ST45SF-G, as mentioned, was quiet at idle. Under load, whatever noise it was making was masked by the 7970's blower. However, on stopping Furmark/Prime95, the PSU's fan took longer to spin down than the GPU does and I was able to hear some of the tonal noise people complain about.

The case, while not doing much to block noise on account of all the vents, didn't seem to amplify it, either. The panels fit tightly and I didn't notice any vibration there. I tried 5400 RPM 3.5" drives mounted in both the fan cage and at the bottom, and didn't find them to be a significant source of noise. On the other hand, I also tried an old 7200RPM drive I had and it resulted in a deep, resonant sound that was further amplified by the desk it was sitting on. This might just be a fluke with that particular drive, or might be because it's older, I'm not sure. In any case, the 5400 RPM drive was quiet. I also noticed a similar deep resonant sound with the slim optical drive (a Sony model) - but only when it spun up to a high speed. At lower speeds (e.g., DVD playback), the noise level was reasonable (though not as quiet as a standalone player).



Config 1a pics

















Configuration 1b

This is a slight change from the above configuration - I added a 120mm fan to the fan bracket (which required removing one 3.5" HDD from the cage). I wanted to see how much of a difference the extra airflow would make. The fan used is a Nexus model running at 1000 RPM.

To sum it up - it didn't make much difference, either in terms of noise or temperatures. I will say I *felt* a little better about it though, considering the airflow from the added fan is hitting the 3.5" drive, the back of the video card, and providing a bit more air to the PSU. I don't know what temps the PSU is seeing and I wasn't monitoring HDD temps for these tests, so any actual improvement is speculation on my part.

Config 1b pics












Configuration 2

  • ASUS P8Z77i-Deluxe Mini-ITX motherboard
  • i7 3770K
  • Corsair H60i
  • Reference HD7970
  • SilverStone ST45SF-G
  • 1x 3.5" HDD
  • 1x 2.5" SSD
  • Slim Optical
  • 2x Nexus 92mm fans @ 1500 RPm, bottom front and rear mounted
  • 1x Nexus 120mm fan @ 1000 RPM, side bracket

Similar to the previous setup, the ECS MDTX motherboard was changed out for the ASUS, which puts the GPU a slot higher and allows for bottom-mounted fans. I wanted to see how much difference the extra airflow would make, and in particular, how well it works to have a 92mm basically force-feeding outside air to the GPU's intake.

I installed a 3.5" drive on the bottom, but left the side mounted double bracket empty.

It's worth mentioning that the rear 92mm fan can only be mounted inside with the H60i ports oriented at the top or bottom..


Config 2 thermals

Furmark
7970@1100/1500
21.7C ambient
88C GPU after 20min, 53% fan

Furmark + Prime95
7970@1125/1575
22.5C ambient
88-92C GPU after 40min, 55%-61% fan
CPU max 58/64/55/57C



Basically, the added airflow didn't make much difference here. The fan ran a couple percent lower, which makes essentially no difference in terms of noise. The CPU ran cooler, but that's due to the ASUS motherboard running the H60i fan faster than the ECS board did.

It seems to me as long as the GPU has access to a nearby source of cooler air, it doesn't much matter how fast that air is moving towards it - at least in the case of a blower-style cooler. It'll get the air it needs regardless.

I did monitor HDD temps and GPU VRM temps for this test (during gaming anyway) using HWInfo. The bottom 3.5 HDD hit temps of around 42C, while the hottest GPU VRM sensor reading was about 12C less than the GPU core temp (which seems low, but I'm not sure).


Config 2 noise

This setup was a bit noisier at idle, on account of the 1500 RPM 92mm fans, but otherwise similar to the previous setup.


Config 2 pics









Misc pics






Still to come: Arctic Cooling Accelero Xtreme testing, CPU+GPU watercooling.
 
Last edited:
Thanks for the update

Interesting that the additional fans didn't do much for the cooling in the "stock" setup.

Can't wait for the next round of testing updates.
 
Last edited:
Thanks for the porn. I have the Powercolor 7870 blower style and it's loud under load. I heard that the Nvidia 600 series blower are less noisy.
I woinder if one or two fans at the bottom will improve GPU temps.
 
Thanks for posting Necere!

Thanks for the porn. I have the Powercolor 7870 blower style and it's loud under load. I heard that the Nvidia 600 series blower are less noisy.
I woinder if one or two fans at the bottom will improve GPU temps.

Config 2 he had a bottom fan right under the blower fan

It seems to me as long as the GPU has access to a nearby source of cooler air, it doesn't much matter how fast that air is moving towards it - at least in the case of a blower-style cooler. It'll get the air it needs regardless.
 
Do you expect that we'll be able to fit a 120mm single rad in the bottom front of the case? Assuming you're using a SFX PSU and a short GPU like the GTX 660 and 670? Or will there be clearance issues with the motherboard?
 
Do you expect that we'll be able to fit a 120mm single rad in the bottom front of the case? Assuming you're using a SFX PSU and a short GPU like the GTX 660 and 670? Or will there be clearance issues with the motherboard?
If you:
  • watercool the GPU
  • use a slim rad (<30mm)
  • use a slim fan (12mm)
  • drill your own mounting holes
Then yes, I think it should fit.


Btw, I would not expect to get a dual rad in the bottom. I tried it with a Swiftech MCR rad and it kinda sorta almost fits, but the problem is the front I/O cables are in the way and need to be bent at fairly extreme angles. That rad also has a unique shape to the endtank that allows a little more space for the cables than most other rads would - I think most rads are going to be out of the question. The Swiftech rad won't work though because it's too thick (~35mm) to be used even with slim fans and a watercooled video card.
 
Last edited:
Necere, are you planning on doing an air only thermals test with the Noctua NH-C14 you originally showed us fitting in?
 
I'd say low. mDTX never really took off, an ITX is only starting to gain traction with the big manufacturers.
 
I think new mini-DTX motherboards are a slim chance for 2013. Since most M1 buyers would use this for a better soundcard, many motherboard manufacturers (except Asus) would like you to buy a motherboard with better onboard audio, which generates more profit to them directly. AsRock has just revealed a new "high-end" onboard solution and there are other manufacturers that already have to technology to do this.

I see it as more likely we'll see improved onboard solutions than a new mini-DTX motherboard.
 
well how are y'all planning to integrate a discrete sound card into your systems?

You got multiple options :
1) no discrete sound card
2) USB sound card
3) get a board with mPCIe, buy a mPCIe->PCIe x1 ribbon cable, attach your sound card to the PCIe connector at the end of the ribbon cable, mount it to the 3rd slot.
 
well how are y'all planning to integrate a discrete sound card into your systems?

I believe there was some talk of using a mini-pcie to pcie x1 adapter cable in order to integrate a sound card.

I don't think it would be easy, admittedly I only scanned the shopping engines and stores briefly, but I couldn't see any adaptor cables that I could visualise routing off the motherboard, around the graphics card and underneath the where the card would be (assuming it would be mounted on the 3rd expansion slot of the case.)

A lesser problem might be mounting the card securely to the case without a firm pcie slot.
 
around the graphics card and underneath the where the card would be (assuming it would be mounted on the 3rd expansion slot of the case.)

Around ? Why would you route around the card ? Bellow is where most will route that ribbon cable. You got place in front and behind the PCI-E slot.
 
Around ? Why would you route around the card ? Bellow is where most will route that ribbon cable. You got place in front and behind the PCI-E slot.
around, below, past, whatever, even off the opposite edge of the mobo and underneath - whichever way works!

I was trying to emphasise the need to plan it carefully, those ribbon cables won't twist easily over a short distance and it needs to end up with the pcie socket in the right spot for the card. I wouldn't like to just assume that if the cable is trapped between the gfx card and mobo it will adjust enough to meet the sound card connector, on the other hand it may work out perfectly if the mini-pcie is located conveniently.

At least it's possible to choose left or right side exit with some of the adaptors, and there is always the option of soldering and making a custom cable.
 
The ECS board seams to have a very cool layout, im just not sure how severely I will be limited with it. From what I can tell - only usb 2, no wifi or bluetooth card, support up to ivy bridge, no cpu overclocking and a cheap feel to the board.
 
Looks great!!! Cannot wait to see the full water cooled build.

Thermals look fine, and I figured they would be very similar. I would say typical temps for a sff.
 
Necere, are you planning on doing an air only thermals test with the Noctua NH-C14 you originally showed us fitting in?
Yes. Doing that now in fact: NH-C14 + Accelero Xtreme on the 7970.

How do you like the ECS mobo?
Well, it's more of a value model, with the features to match. It's only $40 up on Newegg. In its favor, it 1.) has the CPU socket located away from the PCIe slot, 2.) has some overclocking options (though I didn't investigate them), 3.) has an extra PCIe slot, 4.) didn't give me any problems.

The things I didn't like about it: 1.) no USB 3.0; 2.) no power LED header; 3.) bad location for the HD audio header (right under the video card, which smashes the wires; to be fair, the ASUS board also suffers from this); 4.) extra PCIe slot is only x1.

What are the chances of new mini-DTX boards being released?
Quite likely. I mean, we'll see new MDTX boards continue to trickle out. Will we ever see high-end MDTX boards is the real question.
 
Back
Top