Kimera Industries Cerberus: The 18L, mATX, USA-made enclosure

Allanitomwesh said:
Sounds good :) Nothing worse than a wobbly handle from a loose screw. It looks great in red btw
Click to expand...

I'm thinking of having lock washers as well to keep the bolts from accidentally backing out. I'll do lots of testing to ensure that the handle won't come off accidentally in normal use.

Yeah, I haven't seen a nice dark red case so I'm excited for that.

mrpawick said:
Will there be options for colors? Like Red, Black, Silver... and dare I say... Gunmetal?

I can't wait to throw my money at that case.
Click to expand...

I'd like to offer as many colors as possible. If I can meet the minimum quantity to make it worthwhile I'll offer that color.

ekuest said:
subbed because nice case.
Click to expand...

Thanks!
 
This certainly looks tasty. I'll be keeping an eye on this thread for sure. Might re-home my i7-920 rig into one of these and relocate it to my entertainment center.
 
Thanks! Unfortunately I've had to put the project on hold for a little while, some stuff came up and the money I set aside for this was needed for other things :(

I'm hoping it'll only a month or two of delay.
 
We'll be waiting patiently, there's nothing else on the market right now that hits the spot like this concept does. :)
 
I m liking this a lot. I ll definitely be watching to see what the pricing looks like if offered. My only comment is that making the 500watt psu fit without hassle would be an attractive feature.
 
chemdog said:
I m liking this a lot. I ll definitely be watching to see what the pricing looks like if offered. My only comment is that making the 500watt psu fit without hassle would be an attractive feature.
Click to expand...

The 500W is what I want to use for my own build so I'm trying to make sure there is enough room for it to fit without too much trouble.
 
I do have a little more feedback...Looking at my needs for a perfect gaming case, i d say its a slightly larger ncase with a handle and a easy, slightly roomier wiring setup with the 500watt psu which is looking very tight/unworkable in the ncase. One large 2 slot gpu, air cooled cpu and 16gb ram max...MicroATX seems unneccessary for the board for me, so on a case like this, i d probably use mini-Itx.

An idea could be to have a secondary, optional, lower set of mounts for Itx that skips the first rear card slot, leaving slots 2 and 3 for a single gpu setup, leaving plenty of space to use the 500 watt psu wiring, and a 3.5 drive on the bottom. Give everything a little more space for the itx user... 2 tabs that protrude down into the rear entry space could do it.

Any way it looks like the current setup should work fine as well. It will be interesting to see the prototype built up. Hopefully the 500w lands by the time you start building.

I like the steel instead of aluminum, would prefer a black power button over red/black, I also prefer the placement of the power button/usb etc on the ncase at the bottom.
 
chemdog said:
An idea could be to have a secondary, optional, lower set of mounts for Itx that skips the first rear card slot, leaving slots 2 and 3 for a single gpu setup, leaving plenty of space to use the 500 watt psu wiring, and a 3.5 drive on the bottom. Give everything a little more space for the itx user... 2 tabs that protrude down into the rear entry space could do it.
Click to expand...

That idea was discussed for the M1 but the problem is the rear IO cutout. The only feasible way to do that is to have a removable motherboard tray but that is a lot extra complexity.

It's a tight fit in the current design because of the angled power plug. With a custom-made low-profile plug it would be possible to shift the PSU bracket upwards, leaving more room for the 500W but that would require a fairly large number of orders to do.
 
Aibohphobia said:
With a custom-made low-profile plug it would be possible to shift the PSU bracket upwards, leaving more room for the 500W but that would require a fairly large number of orders to do.
Click to expand...

Do you have any idea of the MOQ for a low-profile plug?

If the MOQ wouldn't be met, would it still be possible to add a second set of screw holes in the psu bracket, allowing it to be shifted a bit upwards?

The reason I'm asking is that I wouldn't hesitate to modify a regular-profile plug on my own.

Also, any news? Stupid to ask really, since if you had news you'd posted them already, but anyway? :)
 
backfeed said:
Do you have any idea of the MOQ for a low-profile plug?

If the MOQ wouldn't be met, would it still be possible to add a second set of screw holes in the psu bracket, allowing it to be shifted a bit upwards?

The reason I'm asking is that I wouldn't hesitate to modify a regular-profile plug on my own.
Click to expand...

I suspect between 1-2 thousand orders would be needed to justify the expense of custom plugs.

With the current bracket design there isn't an easy way to shift it upwards, my plan was if I couldn't get enough orders for custom plugs at first was to offer a new bracket alongside the custom plug if I got enough orders later.

backfeed said:
Also, any news? Stupid to ask really, since if you had news you'd posted them already, but anyway? :)
Click to expand...

Nothing to report for now. Will probably be another month before I have the funds in place to get started and I'm still waiting on the Rampage V Gene to be announced. I'm hoping they don't go with a vertical M.2 slot like on the X99 Deluxe because that could be a problem.
 
Last edited:
Aibohphobia said:
With the current bracket design there isn't an easy way to shift it upwards, my plan was if I couldn't get enough orders for custom plugs at first was to offer a new bracket alongside the custom plug if I got enough orders later.
Click to expand...

Ok, that sounds like a good idea!

Aibohphobia said:
Nothing to report for now. Will probably be another month before I have the funds in place to get started and I'm still waiting on the Rampage V Gene to be announced. I'm hoping they don't go with a vertical M.2 slot like on the X99 Deluxe because that could be a problem.
Click to expand...

Ok. Will be waiting patiently.

While there are several case projects running right now that are of interest to me, yours is the only one that wouldn't require any hardware compromises, the only one that allows good ventilation without having to resort to tiny fans, and also the only one sensible enough to use steel. :cool:
 
I ve been thinking the solution for the tight clearance between 500w psu and full sized gpu could be solved( for me at least), by using a m-ATX Gene VII mobo, and putting single card in the lower card slot. As i read the spec sheet there are 2 of the PCI express 3.0 x16 slots.

I m interested to see some reviews on the 500w psu when it arrives. Hopefully they make a premium unit.
 
chemdog said:
I ve been thinking the solution for the tight clearance between 500w psu and full sized gpu could be solved( for me at least), by using a m-ATX Gene VII mobo, and putting single card in the lower card slot. As i read the spec sheet there are 2 of the PCI express 3.0 x16 slots.
Click to expand...

On the boards that allow it putting the video card in the lower slot will be better since it'll be closer to the intake vents on the bottom of the case.
 
I shot off emails to both the sheet metal fabricator I want to use and a company that makes custom molded cords yesterday to get the ball rolling. Waiting to hear back from them.

The custom power cord will almost certainly require tooling so I'm just trying to find out how feasible it will be at the quantities I'm expecting.

Edit: I'm also waiting on a NZXT X41 140mm water cooler I ordered. I had planned to use the Noctua C12P but the socket on the EVGA X99 Micro is so close to the top PCIe slot that no top-down cooler will fit. The X41 is the only water cooler that gets SPCR's approval so I need to make sure it fits and is indeed quiet enough. I don't think any changes need to be made to accomodate it but I want to measure it to be sure.
 
Last edited:
Aibohphobia said:
Edit: I'm also waiting on a NZXT X41 140mm water cooler I ordered.
Click to expand...

Very good aio with good results, did you choose this over the x40 because of thickness or performance?

Also when installing just remember the tubes are long, so take that into consideration when installing.

I've been following this for a while but I said I'll comment to add to the 'interest' to keeps your hopes up. :)
 
smithydan said:
Very good aio with good results, did you choose this over the x40 because of thickness or performance?

Also when installing just remember the tubes are long, so take that into consideration when installing.

I've been following this for a while but I said I'll comment to add to the 'interest' to keeps your hopes up. :)
Click to expand...

Didn't even know there was a X40 :p

It's the only AIO cooler that gets the thumbs up from SPCR other than the X61 which won't fit so that's what I ordered.

The long tubing will actually come in handy in this case with the hinged rad bracket, should be able to swing it all the way out to get easy access to the internals :)
 
Got the X41 in and installed today and I'm really liking it so far:

2014-09-26_Overclocking-test-setup_small.jpg


larger version


For those that haven't been keeping up with the SX600-G thread this is a test rig I built once I saw the TDP/power consumption figures of the new Nvidia cards. I just had to try and see if the SX600-G could power a Haswell-E system with SLI GTX 980s to prove that a high-end dual-card rig powered off SFX using a mATX board was feasible.

Parts arrive
Assembled
Test setup and single card results
SLI!
More results
Spreadsheet of collected data

TLDR: 6-core Haswell-E, SLI GTX 980s, 600W SFX, 512W AC draw (~460W DC)

And now we have a killer app for this case :D


Anyway, at stock clocks at full load the CPU was sitting at ~50°C with the pump and fan at full speed. It got hotter once I turned the fan and pump down (forgot to note how much hotter though) and the noise levels seem pretty good. I'll be testing it more tomorrow.
 
Following your tests with interest!

The fan in the upper left corner of the photo, would that be a replacement candidate for the SX600-G stock fan?
 
Yes, it's the Sanyo Denki 109P0805M702 :)

I need to do more testing with it but it seems like a slight improvement over the stock fan.

It doesn't have the chatter but it does have the distinct ball bearing noise instead. At least it's a smooth, constant noise so I think it's more tolerable than chatter but I'm looking forward to when Wisk gets those FDB versions of the stock fan to see the results.

Even if the the FDB fan turns out to be a great replacement, they seem hard to get a hold of so realistically those looking for a low noise SFX PSU may be better of with the SX500-LG.
 
Is it just some random Japanese keyboard in your test rig photos? Doesn't look like either the Realforce or the HHKB.

Aibohphobia said:
I have a powder coat color brochure on order, I'd like to get the exterior panels in a nice, deep red to match the ROG boards :D
Click to expand...

Now that you have the EVGA X99 mATX board with two 980s...are you still getting a Rampage V Gene for this case in red? :p
 
smithydan said:
We'll even better.
Click to expand...

I tested coiling the tubing around and holding the rad about where it would be in my design and it will work with a single fan. I don't think there's enough clearance for push-pull but a single high static-pressure fan should be sufficient.

Anyway, I tried simulating opening the hinged rad bracket and there was plenty of slack in the tubing to do so. With the pump at full speed temps are good though it's a bit noisy. Temps rise with the pump dropped down to 9V but it's much quieter than any other AIO I've heard. It's not silent and I'd still prefer to use my Noctua NC-C12P cooler but I could live with it.

I haven't tested every CLC there is but so far I think the X41/X31 will be my recommended water cooler for this case.

veryrarium said:
Is it just some random Japanese keyboard in your test rig photos? Doesn't look like either the Realforce or the HHKB.
Click to expand...

My boss at the computer shop I work for ordered a big box of Dell keyboards for a great price but when they arrived he realized why they were so cheap :p

I think most of them got thrown away but there's still a few left floating around the shop as spare keyboards.

veryrarium said:
Now that you have the EVGA X99 mATX board with two 980s...are you still getting a Rampage V Gene for this case in red? :p
Click to expand...

I've been very disappointed with the EVGA board. I'm really counting on Asus to come to their senses and release a Rampage V Gene because this case won't be complete without it :D

Here's a mockup I made of a hypothetical R5G:

Rampage-V-Gene_mockup-900x859.jpg


I made a thread if you want to show your support for an Asus X99 mATX here: http://hardforum.com/showthread.php?t=1835457
 
Last edited:
All existing mATX X99 motherboards I can find seem a bit stupid.

It would make sense to have x16/x16/x8 bandwidth for the three pci-e slots, version 3.0, using all possible 40 lanes. It seems none of them do it this way, but instead go x16/x8/x4. Even worse, I think I even saw the x4 being pci-e 2.0 on at least one of the models. WTF?

Sure, x16/x8/x4 = the max 28 lanes the 5820K supports, but what if you have a 5930K or 5960X?

It seems all (?) motherboard manufacturers assume mATX = budget-only in this case.
 
The EVGA does do x16 / x16 / x8 with the 40-lane CPUs. The extra $200 over the 5820K is just a waste for what I want to do though since I'll only need x4 for the middle slot, at least until we get PCIe SSDs so ridiculously fast that they need x8 :D

The ASRock boards I don't understand, they have all these lanes available but they cripple the bottom slot at x4 :confused:
 
Edited the OP with a status update. TL;DR: still working on getting prototype made, will hopefully get it done before end of year :p

Went to a LAN party last weekend and decided to take my testbench setup with me so I ended up throwing it in a HP case that was laying around:

Lan_temp_case.jpg


Lan_temp_case_harness.jpg


Getting everything assembled in that cramped HP case really made me appreciate various features of my design, things like the hinged rad bracket will make cramming such a high-end build in a small case much easier :D

I found a local buyer for my 5820K so I ordered a 5930K that should be arriving tomorrow so I can test SLI temps with a empty slot between the cards.
 
Last edited:
I agree with you about the hinged bracket, I also happened have a mATX case made by HP with an inverted motherboard layout and the side vent holes exactly in the same location as with your case, and due to no air intake on the front panel I attached a fan on the side vent just like how you put your rad there, but the fan cable that hangs between the side panel and the mobo was really annoying whenever I opened/closed the panel. This idea of a side fan/rad mount bracket that's in NCASE M1 and in yours should be more popular in the future.

In-Win has been selling BK623 and BM series that have a side mounted front 5.25 inch bay holder that extends all the way to the rear end. I once thought about getting a second hand BK623 and modding that bracket to mount 240mm rad (slightly sticks out front of the chassis but barely fits inside the thick plastic front panel) but found that this case was made of 0.5mm thick steel and thought it could end up with unbearable vibration so I gave up on the idea.

That's too bad those manufacturers only accept SolidWorks or other read CAD drawings. I hope your project will come through as I really want this case.
 
Last edited:
If a 5930K will ultimately be the only solution to the motherboard problems when running SLI, then so be it. :)

Too bad about the freelancer, and it seems manufacturers generally prefer SolidWorks, yes. I think it was the same with the NCASE M1 project (where I think Wahaha360 did the conversion work, not sure).

With luck, maybe somebody on this forum is capable of helping you?
 
Got it working, had to reinstall the Nvidia drivers though, wouldn't recognize the bottom card at first. I have to give EVGA credit, I called their tech support thinking that it was yet another failure of their board but I was on hold for about 15 seconds, got someone in America, and that was what he recommended and it fixed the issue :)

2014-10-09_5930K-test_1.jpg


Larger version


Setup

  • Intel i7-5930K
  • NZXT X41 140mm AIO cooler
  • EVGA X99 Micro
  • Crucial 2133 2x4GB DDR4
  • 2 x EVGA GTX 980 w/ reference cooler
  • Silverstone SX600-G


Methodology

FurMark at 1920x1080 and 2x MSAA ran for 15 minutes then measurements recorded. GPU temps and RPMs taken from HWMonitor.

Did three runs with different configurations:

  • Test 1: Cards in top two slots, no space between, no fan
  • Test 2: Second card in bottom slot, middle slot empty, no fan
  • Test 3: Second card in bottom slot, middle slot empty, Noctua P12 @ 900 RPM in front of cards

The Test 3 configuration is the recommended setup for running SLI in the case.


Results

2014-10-09_5930K-results.png


Very happy with the results :)

Having a empty slot between the cards and and a fan in front to bring in fresh air dropped the temperature difference from 9°C to 1°C AND dropped the top card's fan RPM from 1500 RPM over the bottom card to just 200 RPM over :D

It's a shame with the current boards that it requires the more expensive 5930K to get this configuration working but it's nice to see the addition of the fifth slot undeniably justified :p


veryrarium said:
I agree with you about the hinged bracket, I also happened have a mATX case made by HP with an inverted motherboard layout and the side vent holes exactly in the same location as with your case, and due to no air intake on the front panel I attached a fan on the side vent just like how you put your rad there, but the fan cable that hangs between the side panel and the mobo was really annoying whenever I opened/closed the panel. This idea of a side fan/rad mount bracket that's in NCASE M1 and in yours should be more popular in the future.
Click to expand...

Yes, I hate having stuff attached to the side panels. I ran into this several times during the LAN showing off my rig, I couldn't just easily pop the panel off, I had to carefully remove it and then make sure there was enough room beside it to set the panel with rad attached down so it didn't stress the tubing.


backfeed said:
With luck, maybe somebody on this forum is capable of helping you?
Click to expand...

Sure, if anyone with SolidWorks and sheet metal design experience is reading this and wants to help, send me a PM. I'm willing to pay.
 
Aibohphobia said:
Very happy with the results :)
Click to expand...

Yeah, that looks really good!

Could in fact get even better inside your future case, where the fan output will be forced into a tunnel of sorts. At least the back plates of the graphics cards should become cooler, which may indirectly improve gpu temperature (and hence fan speed). You could test it by constructing a simple cardboard tunnel with the dimensions of your case. :)
 
Yes, it'll probably be a bit better with the case panels to direct the airflow. I won't bother mocking it up for now since the results even from the open air bench were what I was looking for.

What I'm really interested in is trying to run the reference cards semi-fanless. I'm just waiting for a working version of NVFlash so I can override the minimum fan speed.

I figure if they can make 40W TDP cards with single slot fanless heatsinks that the 980s with their massive heatsinks should be able to idle at 10-15W without the fan running :p
 
I suspect that the reference cooler is awful at passive duty though, and that most of the heated air will find its way inside the case instead of going out (e.g. through the seemingly stupid gap in the side of the shroud, close to the external connector end of the card). Also, when positioned vertically, the top card would probably get quite a lot hotter. I'd think even the slightest breeze would be much better, running the fan at the lowest possible rpm.

But will be interesting to see. :)
 
It's only 10W and some airflow should make its way into the card from the front 120mm fan. Won't really know until I try.
 
There's finally a working version of NVFlash out so I modified a EVGA SC ACX rom back to stock power/clock settings so I could just get the semi-passive fan profile and it works :D

Idles at 44-47°C and thermogram looks good:

2014-10-23_semi-fanless-test-idle.jpg


The ACX curve is a little too easy on the fan speed at high load though, the cards hits it's thermal limit running FurMark where on the stock rom it would hit power limit first.

Not a big deal for me since I'll probably lower the power limit below even stock settings so I can run off the SX500-LG.

I also did notice some whine under certain load conditions :( At least it doesn't whine all the time or under every load.


I talked to the local machine shop last week but they weren't well equipped for what I want to do. They do plenty of precision sheet metal but it's mostly small stuff, they don't even have a CNC press brake. So I got in touch with Protocase to see what they could offer. Got the initial quote today for $8000 :eek:

Turns out the quote was so high because of all the vent holes, I'm assuming they're going to cut them all with a laser which would be very time consuming. I'm going to get a revised quote with some of the vent holes omitted (like on the back side panel), the bottom vents replaced with a large cutout and mesh, and use slots for the top and side panel.

Hopefully I'll get the revised quote Monday and if it's something I can afford we may be looking at a prototype in my hands by first week of November :)
 
Sorry I'm not familiar with this process at all, but was that $8000 quote just for a single prototype unit?

I was wondering if the right side panel ("right" when looking at the case from the front) really needed those vents, I'm fine with those omitted as you did to get a revised quote.

Speaking of the vents, I'm looking at your sketchup file (sorry, for the first time) and noticed that the rectangular area that the vent holes form on the the left side panel and the more-or-less rectangular area that is cut out in the hinged bracket don't align and the former rectangle is pushed back relative to the latter rectangle. Is that done on purpose?

Other thoughts/questions I have, all regarding stability: (I know, I should have followed your progress more closely in real time and asked questions back then about little details so it could have given you feedback before sending your design to make a prototype)

1) I like the idea of an inner bracket on the left side for fan/rad mounting as I said before, but as yours is a hinged bracket, I'm wondering if a bracket that's screwed on the one end (the front end) but supported on the other end by a hinge could cause the mounted fan/rad to vibrate (if not wobble) more than we may imagine at the design stage.

2) Also wondering the stability of the long strip behind the motherboard extending from front to back near the top, as indicated in the following image in red:
2RmuoPF.png

As I looked closely at your sketchup file I don't think this strip is part of a single large steel plate for the entire right side with a big rectangular hole cut out in the middle. So I'm a bit afraid it might bend easily depending on the pressure applied on the standoffs when a motherboard is tightly screwed to them.

3) There were some questions earlier in the thread regarding how the handle can sustain the weight of the whole rig without causing any bending and you gave an answer I suppose, but I'm not so sure if the handle really can...um, handle the weight after looking at your file closely. The handle is screwed on to the fan bracket at two spots across the big center circle cutout, and each of the screw points are located in the middle of the mere 21mm-wide bridge-like portions where the entire weight of the rig including a PSU, a CPU cooler, a GPU, etc is exerted on. And the fan bracket itself is screwed to the top frame at its four protruding corners. Those tiny protrusions must be more vulnerable than just the corners of a plain square when a big weight is exerted, at least my intuition says so.

Once the prototype comes into shape and you build your test setup these thoughts/concerns of mine will be cleared, I guess, but at the moment I'm slightly concerned.
 
Yes, $8000 was for a single unit. Laser cutting is priced by distance cut and number of piercings. Each time the laser needs to stop one cut and start a new one takes some time for the laser to burn though (pierce) the sheet. So having lots of little holes makes for both lots of total distance and lots of individual piercings.

I assumed they had a CNC turret punch like is used for the vent holes on the M1 but apparently Protocase was planning to use a laser for all the vents and hence the crazy price.

The left panel (right? I'm not sure what the proper convention is) doesn't need to be vented, I just did that to make the exterior symmetric.

The vent hole pattern on the panel is positioned to roughly line up with the holes on the top panel. There isn't anything at the front end of the rad bracket that needs venting.


1 The hinges I plan to use are high quality precision made stuff so I don't anticipate any issues there but that's what prototypes are for. If it turns out to be wobbly I'll drop it and go for screws on both ends.

2 It's not modeled but the plan is for the top and bottom of that piece to have a flange bent over 180 to stiffen the rail.

3 That's why the bracket is bent on each end. Sheet metal relies on perpendicular bends for strength, a flat piece of metal isn't that strong but with the careful placement of flanges it can support plenty of weight. But again, I'm using the prototype to test stuff like this.
 
Aibohphobia said:
Yes, $8000 was for a single unit. Laser cutting is priced by distance cut and number of piercings. Each time the laser needs to stop one cut and start a new one takes some time for the laser to burn though (pierce) the sheet. So having lots of little holes makes for both lots of total distance and lots of individual piercings.

I assumed they had a CNC turret punch like is used for the vent holes on the M1 but apparently Protocase was planning to use a laser for all the vents and hence the crazy price.
Click to expand...
You'd probably want to ask them to make sure, but from their website it doesn't seem like they have an NCT machine. If they did, you'd want to find out what punches they already have tooled and adapt your design accordingly. This may not make sense with a low-volume manufacturer, though, since those punches have a limited lifespan and need to be constantly maintained and replaced. This is what lets us produce the M1 relatively inexpensively - we're essentially piggybacking on LL's existing production pipeline, so the cost is much less than if we had to tool up for a single case.
 
Thanks for chiming in Necere, I know you think ATX PSU is the way to go but I hope I've adequately addressed your airflow and filtration concerns from way back on page 2 :)

I'll have to explicitly confirm with them Monday if they have a NCT or not but it doesn't appear so.
 
You must log in or register to reply here.
Back
Top