CCD MI-6: Performance in a 6.7L MIcrotower

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Good to hear you are interested!

The biggest issue of course is the relation of cost to production quantity, so the recent poll showing 25 interested this year was great impetus to keep working pricing.

As far as the big shuriken 2, I really like that cooler, so I had to make sure it would fit, lol.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Sorry, I haven't made much progress lately. I have some other issues going on that need attention. BUT, I did get some prototype parts in!

They don't have a finish on them yet, just bare mill-finish aluminum. Here is 1 set of the main parts: chassis, divider, hdd bracket, cover and front plate (if you can make them out in all the clutter!):


Here is a view of the chassis, with the reconfigured 92mm fan mounting area slash hdd mounting area:


Then I put everything together and took some pics:







I'm busy with other things, but am trying to work through some prototype issues. The cover fab didn't really work out too well, as they couldn't do both of the deep bends very well, and some corner brazing/welding isn't too pretty yet. And working some part tolerancing.

Oh, I put some surrogate perforated panels in the cover. I'll be mixing and matching at least 2 sizes and getting temperatures to see which is best. In the pics above, the small hole pattern is on the GPU side, the larger pattern is everywhere else.

And after taking all these close ups, I took a pic of the Mi-6 by my son's ATX mid-tower gaming setup so we could all remind ourselves of the scale, lol.

Hopefully I'll get all the parts powder coated in the next week, then that should look much better.
 

Boil

[H]ard|Gawd
Joined
Sep 19, 2015
Messages
1,421
You know we are gonna want to see a build video, yeah…?!?

When you have time, of course…!

I like the black button screws with the 'silver' front; I think that will look sharp with everything else in black powder coat…

A polished front will be awesome for some laser etching, you may need to sell extra front panels…!
 

iFreilicht

[H]ard|Gawd
Joined
Sep 23, 2014
Messages
1,348
Very nice. Even if the welding at the rear of the cover doesn't look that good yet, it seems like a good approach. If done well, it should look very good. How are you securing the panels right now? Just double sided tape I suppose?
 

firas

2[H]4U
Joined
Oct 29, 2006
Messages
2,448
quick question, can we say that this is (in some way) a DAN-A4 with shorter GPU support (mainly the 970 mini/1070 mini) and bigger CPU cooler support and minus ~ 1 liter?

edit: what would be the next possible option to reduce the size? designing a similer case around a ~500W FlexATX PSU and then we're done?
 
Last edited:
D

Deleted member 222586

Guest
quick question, can we say that this is (in some way) a DAN-A4 with shorter GPU support (mainly the 970 mini/1070 mini) and bigger CPU cooler support and minus ~ 1 liter?

edit: what would be the next possible option to reduce the size? designing a similer case around a ~500W FlexATX PSU and then we're done?
This case will have MUCH, MUCH, MUCH better cooling capabilities than the DAN-A4 simply because of the intake fan. That fan alone changes everything.
 

Aircoookie

Limp Gawd
Joined
Feb 6, 2015
Messages
375
This case will have MUCH, MUCH, MUCH better cooling capabilities than the DAN-A4 simply because of the intake fan. That fan alone changes everything.
Nope. This case uses the same concept as the A4, each component can directly intake air from outside the case. The extra fan will change little about that. Better CPU cooling will be possible in the MI6 only because the heatsink can be 15mm bigger.
 

dondan

[H]ard|Gawd
Joined
Apr 15, 2013
Messages
1,742
prava Why do you think you have so mutch knowledge about thermal designs? Did you ever own a case with that sandwitch design? This one fan is only for cooling the SSDs it will not give any advantage on the CPU and GPU.

Furthermore it is not necessary, because GPU and CPU cooling will work on optimum. They can suck fresh air directly through the outside.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
quick question, can we say that this is (in some way) a DAN-A4 with shorter GPU support (mainly the 970 mini/1070 mini) and bigger CPU cooler support and minus ~ 1 liter?

edit: what would be the next possible option to reduce the size? designing a similer case around a ~500W FlexATX PSU and then we're done?
Firas,
The Mi-6 idea is like you said, a different tradeoff than the A4 makes. It isnt a liter smaller, more like .5 L. Hard to get any smaller than Dan's. The Mi-6 makes a tradeoff on height and width to get a smaller depth and smaller footprint on the desk.

As far as smaller, Hahutsy is working on that Hutsy XS with a smaller psu as you mention. His thread is on these boards as well.
 
D

Deleted member 222586

Guest
Nope. This case uses the same concept as the A4, each component can directly intake air from outside the case. The extra fan will change little about that. Better CPU cooling will be possible in the MI6 only because the heatsink can be 15mm bigger.
Wrong. All the openings in the case will work as exhausts simply because you have active airflow inside the case. That extra fan changes everything. That fan is the only thing in the case that creates a positive pressure inside the case and will force all hot air outside the case. It is a game changer. You seem to think that just because a fan is "close to" an opening it will take all the air from that, and that is completely wrong. Air will take the easiest path, which means that unless the fan is exactly attached to the frame you will have recirculating air fucking up your temperatures. That can't be avoided.

With the A4 you only have to put a >250W gpu with non-blower cooling and test how hot the case becomes after a 2 hour gaming session. It is unavoidable because of the lack of active cooling.

prava Why do you think you have so mutch knowledge about thermal designs? Did you ever own a case with that sandwitch design? This one fan is only for cooling the SSDs it will not give any advantage on the CPU and GPU.

Furthermore it is not necessary, because GPU and CPU cooling will work on optimum. They can suck fresh air directly through the outside.
I understand that you want to defend your creation. I do. But you are delusional if you think that components "suck fresh air directly through the outside". That, to be honest, is a lie. You have tested with non-blower systems @ 200W total load, and with blower systems @ 250W, and for a very short period of time. You can call that "good testing" but it isn't. It simply isn't.

Air will take the easiest path, and that means that air will recirculate if you don't have any means to force it in or out. And with your case you simply don't, because you can't install any fans in the case. This means that the gpu will heat up the backside of the card, which will heat up the whole case, in time. You only have to go for a ~380W @ load configuration and play for several hours. The case alone will be hot to the touch.

Heck it happened with my M1 and I did have an intake fan (but not in an optimal position, mind you). With the A4 it is an unavoidable solution.

Magic does not exist. High-powered systems require decent cooling solutions. And that translates into active cooling. IE forced airflow. This way you can guarantee that you are forcing new air get into the case. Otherwise you are talking about magic.

PS: heck why do you think Necere dropped his vertical slim case? Because his thorough testing found exactly what I'm saying: that the case becomes and oven and sensible parts such as the HDD's (and ssd's) suffer simply because there is no active cooling to prevent it.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Prava,
As Dan mentioned, the CPU and GPU cooling designs follow the A4 method: cool, crisp, minty fresh and delicious air delivered directly to the coolers. No recirculation of warm case air back to the coolers.

Since the coolers exhaust their air into the case, I would reasonably expect that the internals could get to 50-60C under load with high-end CPU and GPU. I added the flexibility for a 92mm bottom case fan to inject some cool air into the case. This is mainly to address concerns about SSD temperature throttling. I'm not a ssd temperature throttling expert yet, but have run across several articles about it (and I did stay at a Holiday Inn Express last night). As I think most of us already know, the throttling is not going to be a concern for light to moderately-taxed ssd's, -well maybe if the ssd is a M.2 stuck under the MB with no airflow or heatsinking. The A4 has the divider to heatsink an M.2 ssd to, I don't, since I have a cutout for access.

I wanted the Mi-6 to have maximum flexibility for the user to do high-end video editing, mild overclocking or whatever, and minimize throttling concerns, regardless of where the ssd(s) is mounted.

There is a good article by Puget Systems here: Samsung 950 Pro M.2 Throttling Analysis

Here is a chart from the article, showing temp and throttling occurring with the GPUs idle at around 65C drive temperature.



And a couple more pertinent articles:

Samsung 950 Pro M.2 Additional Cooling Testing
Samsung SSD 950 PRO 512GB M.2 NVMe PCIe SSD Review - Page 3 of 11 - Legit Reviews
 

iFreilicht

[H]ard|Gawd
Joined
Sep 23, 2014
Messages
1,348
That fan is the only thing in the case that creates a positive pressure inside the case and will force all hot air outside the case.
What about the CPU and GPU fan? They create positive pressure as well.

But you are delusional if you think that components "suck fresh air directly through the outside". That, to be honest, is a lie.
So what you're saying is, a fan placed directly next to a vent won't necessarily intake air from that vent? Where else would this air come from? As you say, air takes the path of least resistance, and that is through the vent that the fan sitting directly in front of.

I get what you're saying though. These vents aren't directional, so if they are much larger than the intake fan, they can be a passive exhaust as well. But I don't see how this differs from an open bench setup. On an open bench, an open air GPU will likely recycle some air as well, and that is what dan is referring to by saying the cooling works "optimally".

Or in other words, if you put the components in the A4-SFX or the MI-6, how does that setup differ from having those same components in the same spatial configuration without any metal in between? In both cases, you have the CPU and GPU fans take in air from the sides, which will then disperse to the top and sides.

The issue with the LRPC was indeed that the HDDs were heating up too much, because there was no active airflow around them, but I don't see how this is relevant to this discussion. The GPU and CPU are getting air as directly as possible, from the outside of the case and have their own cooling fans.

In the worst case, it could even be that the fan in the bottom is pushing the air out through the vents next to the GPU and CPU, increasing the flow of hot air towards the intake of those components and thus the amount of recycled air. The solution to this, which would work in both cases, is to tape off the intake vents where they could act as an exhaust, only leaving out the fans themselves. This would work equally well for both cases.

And please, try to formulate your arguments a little less aggressive. Accusing other users of purposeful lying and calling them delusional won't help them understand your point better.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
One sentence I forgot to put in my last post:

I have not tested my case, so my discussion is based on behavior of other cases. I just got full up prototypes this week, and will be getting in MB, CPU, GPU, PSU etc in the near future for testing.​
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
What about the CPU and GPU fan? They create positive pressure as well.



So what you're saying is, a fan placed directly next to a vent won't necessarily intake air from that vent? Where else would this air come from? As you say, air takes the path of least resistance, and that is through the vent that the fan sitting directly in front of.

I get what you're saying though. These vents aren't directional, so if they are much larger than the intake fan, they can be a passive exhaust as well. But I don't see how this differs from an open bench setup. On an open bench, an open air GPU will likely recycle some air as well, and that is what dan is referring to by saying the cooling works "optimally".

Or in other words, if you put the components in the A4-SFX or the MI-6, how does that setup differ from having those same components in the same spatial configuration without any metal in between? In both cases, you have the CPU and GPU fans take in air from the sides, which will then disperse to the top and sides.

The issue with the LRPC was indeed that the HDDs were heating up too much, because there was no active airflow around them, but I don't see how this is relevant to this discussion. The GPU and CPU are getting air as directly as possible, from the outside of the case and have their own cooling fans.

In the worst case, it could even be that the fan in the bottom is pushing the air out through the vents next to the GPU and CPU, increasing the flow of hot air towards the intake of those components and thus the amount of recycled air. The solution to this, which would work in both cases, is to tape off the intake vents where they could act as an exhaust, only leaving out the fans themselves. This would work equally well for both cases.

And please, try to formulate your arguments a little less aggressive. Accusing other users of purposeful lying and calling them delusional won't help them understand your point better.
iFreilicht,
I'm also thinking about the fans* case fan forcing air out of the side vents, but I think Dan's A4 works pretty well. With the variation in CPU and GPU fan locations from mfgr to mfgr, the side vents have to be oversized to cover all configurations. So the "leakage" of hot air from the side vents is inevitable. I added rear venting to the Mi-6 case because the top exhaust vents are somewhat limited compared to the A4, and I didn't to force a bunch of air to exit the side vents right next to the intakes.

I do hope that blocking unused side vent area won't be necessary, I don't think it will.

EDITED to correct fans text.
 
Last edited:

dondan

[H]ard|Gawd
Joined
Apr 15, 2013
Messages
1,742
prava: This is wrong currently I have a EVGA GTX 980 Ti FTW in it. This is a 250W blower style card. If you don't believe i can make a picture. The card is under 82°C after 2 hours gaming. Yes the case gets hot because aluminium takes heat very good. But the cooling principle works as expected. Hot air moving through the topside and no FAN is needed. Also you are wrong with airflow, I made a smoke flow test. The cards take nearly 90% of the air from the outside. The components are generating positive preasure so the air moved out through the topside.

I know you are a type of a guy that I can't convince so believe what you want.

Firewolfy Sorry using your thread for this. prava we can move PM or in my thread.
 
Last edited:
D

Deleted member 222586

Guest
Prava,
As Dan mentioned, the CPU and GPU cooling designs follow the A4 method: cool, crisp, minty fresh and delicious air delivered directly to the coolers. No recirculation of warm case air back to the coolers.

Since the coolers exhaust their air into the case, I would reasonably expect that the internals could get to 50-60C under load with high-end CPU and GPU. I added the flexibility for a 92mm bottom case fan to inject some cool air into the case. This is mainly to address concerns about SSD temperature throttling. I'm not a ssd temperature throttling expert yet, but have run across several articles about it (and I did stay at a Holiday Inn Express last night). As I think most of us already know, the throttling is not going to be a concern for light to moderately-taxed ssd's, -well maybe if the ssd is a M.2 stuck under the MB with no airflow or heatsinking. The A4 has the divider to heatsink an M.2 ssd to, I don't, since I have a cutout for access.

I wanted the Mi-6 to have maximum flexibility for the user to do high-end video editing, mild overclocking or whatever, and minimize throttling concerns, regardless of where the ssd(s) is mounted.

There is a good article by Puget Systems here: Samsung 950 Pro M.2 Throttling Analysis

Here is a chart from the article, showing temp and throttling occurring with the GPUs idle at around 65C drive temperature.


And a couple more pertinent articles:

Samsung 950 Pro M.2 Additional Cooling Testing
Samsung SSD 950 PRO 512GB M.2 NVMe PCIe SSD Review - Page 3 of 11 - Legit Reviews
The first sentence you mention I couldn't agree less. A blower-style cooler does create negative pressure beucase, well... it exhausts out of the case. We can assume that all the air that it exhausts is fresh air getting into the case. That, so far, is correct. But what about the cpu? The cpu doesn't intake (because the fan isn't attached to the panels of the case, and thus air will take the easiest path) nor doesn't exhaust, so how exactly does air circulate around it? It will be a mixture of convection + internal airflow... and that won't ever yield good results. Yes, on a 200W system sure thing, no problem at all... but go try it on a 400W system. You will realise, then, that the case itself gets hot to the touch, and that fucks up the temperatures of every single component of the case. Simply put, the case gets hot because you don't have active cooling into it. It is simply a matter of time, because it will happen.

I can tell you that my M1 (much bigger than the cases in question) with a reference GTX760 had absolutely 0 problems with temperatures. Sure, it simply was impossible for a problem to show up. It was a sub-250W system with a blower-style gpu and a Corsair H50 (a system that is self-exhausting since it is self-intaking... meaning that it creates an airflow current that regenerates air inside the case). But once I changed that card for a non-reference Sapphire 290X Tri-X... everything changed. We were talking about a 400W system with a non-reference gpu cooler. Sure, my gpu got very decent temperatures... but after around 2 hours of gaming, the case was very warm to the touch, and all the HDD and SSD temperatures spiked up to 50ºC (a bad temperature, unacceptable).

Magic doesn't exist. The same air that goes in will go out... but air, by itself, will not go in nor out unless forced to. Yes, convection is a factor... but is close to meaningless given the minor temperature differencials we are talking about. You need to factor that some components will get no airflow at all... meaning that if the stuff around them gets heated up, they will get heated up too without a chance to dissipate the heat. This, so far, has happened to me in a Silverstone FT03, Ncase M1 and Lian Li V350.

What about the CPU and GPU fan? They create positive pressure as well.

So what you're saying is, a fan placed directly next to a vent won't necessarily intake air from that vent? Where else would this air come from? As you say, air takes the path of least resistance, and that is through the vent that the fan sitting directly in front of.

I get what you're saying though. These vents aren't directional, so if they are much larger than the intake fan, they can be a passive exhaust as well. But I don't see how this differs from an open bench setup. On an open bench, an open air GPU will likely recycle some air as well, and that is what dan is referring to by saying the cooling works "optimally".

Or in other words, if you put the components in the A4-SFX or the MI-6, how does that setup differ from having those same components in the same spatial configuration without any metal in between? In both cases, you have the CPU and GPU fans take in air from the sides, which will then disperse to the top and sides.

The issue with the LRPC was indeed that the HDDs were heating up too much, because there was no active airflow around them, but I don't see how this is relevant to this discussion. The GPU and CPU are getting air as directly as possible, from the outside of the case and have their own cooling fans.

In the worst case, it could even be that the fan in the bottom is pushing the air out through the vents next to the GPU and CPU, increasing the flow of hot air towards the intake of those components and thus the amount of recycled air. The solution to this, which would work in both cases, is to tape off the intake vents where they could act as an exhaust, only leaving out the fans themselves. This would work equally well for both cases.
What I'm saying is that if you put a fan directly attached to a panel or with some separation in between it will suck a varying quantity of air in. If you had no panel at all, or the venting wholes were very big then we could talk about how meaningful it is when comparing open-air vs inside a case... but honestly holes are never very big, and they make a huge factor. You only have to calculate the opening area the panel in front of the fan has compared to the area in between... and then consider that the panel has restriction which the space in between doesn't. I bet that with 1cm (cpu fan to panel) separation 80% of the air comes from inside the case and only 20% comes from outside (so long as you don't have any other fans in that eco-system).

On an open bench non-blower gpu's won't recycle air, or not to a big factor. Consider how most non-blower cooling fins are aligned (they are perpendicular to the motherboard), which means that most air will go up (considering the motherboard is horizontal) whereas the gpu's suck air from the side. Of course, it all changes if you stack several gpu's. But for a single gpu ecosystem temps are miles ahead better in open-air rather than in not-so-good towers. I know, because I have experimented a lot with said systems. My GTX295 dual-pcb and, afterwards, my Gigabyte GTX480 SOC were temperature-destroyers in non-optimal cases... whereas they shined on my Banchetto (open-air case). I'm not making things up here, I'm simply stating what I have experimenting through the years on dozens of different gpu's and different cases. You can control the fan profiles of your cpu and gpu... but you can't control the temperatures of your drives unless there is some sort of active cooling around them. And if you have a case that doesn't have any sort of forced cooling in it, residual heat will raise the temperature of the overall chassis which, in turn, will heat up the rest of the stuff you have in the case. Just imagine you have an air-con in your room. It takes away 350W. What happens if you have 300W of heating elements inside? Well, it will simply cool the room until everything gets to an equilibrium. What happens if you produce 365W of heat? Well... the room temperature will raise. Slowly, but it will. In a computer there are many heating elements. Sure, a gpu is the worst one of all, but many parts of the gpu get hot and not every single bit is cooled by the cooler.
A computer system is exactly the same... with a caveat: if you have fans in it you can more or less control the path the air takes and force air in our out. But if you don't... well, you better cross your fingers and hope that turbulence, convection, and the random fans your components have do a good job.

In your design the fan on the bottom will be a positive thing in any case. Why? Well, it simply creates an airflow current into the case and, by itself, will make the air recycle from time to time so that the both the gpu and cpu get fresh air all the time. It will not interfere with anything simply because you have the cpu and gpu at the same height (on different compartments) and the fan blow directly on top of them. Air from one compartment won't go into the other, but it will force both compartments to have fresh air all the time.

And please, try to formulate your arguments a little less aggressive. Accusing other users of purposeful lying and calling them delusional won't help them understand your point better.
You are right about this bit, and I apologize to you.



One sentence I forgot to put in my last post:

I have not tested my case, so my discussion is based on behavior of other cases. I just got full up prototypes this week, and will be getting in MB, CPU, GPU, PSU etc in the near future for testing.​
I hope you test your case with and without a fan. Consider that for a thorough temperature testing (although in your case you are limited by the gpu length) you require several hours of continual load-testing. Simply play with your computer for several hours and monitor the temperatures and fan speeds while you do it. I mention fan speeds because if the fan reduces the noise of your components then it is actively improving their cooling capabilites.

In any case, I'm looking forward for your results. I'm all in for new stuff in the community SFF crowd, if anything so that the industry players can stop being lazy and start releasing decent products.

prava: This is wrong currently I have a EVGA GTX 980 Ti FTW in it. This is a 250W blower style card. If you don't believe i can make a picture. The card is under 82°C after 2 hours gaming. Yes the case gets hot because aluminium takes heat very good. But the cooling principle works as expected. Hot air moving through the topside and no FAN is needed.

I know you are a type of a guy that I can't convince so believe what you want.
The case gets hot because you don't have any means of exhausting the heat out of it. Also, you should know that the temperature of the gpu is completely meaningless because temperature depends on the noise the gpu is outputting. In any case, tell me, how good are drive temperatures after 2 hours of gaming? How much worse would they become if you were to put a non-blower card in the system?

This is not about believing or not believing, this is about extensively testing or not. For you it might be okay to force your drives to work at up to (or even more than) 50ºC, for me it simply isn't. I'm okay with that. But I'm not in users providing information that isn't factually tested, specially those that have interests because they are selling a product.

PS: you should know that many NVIDIA cards throttle at more than 83ºC. So, your 82ºC aren't exactly good...because the card won't simply go above that mark.
 
Last edited by a moderator:

Necere

2[H]4U
Joined
Jan 3, 2003
Messages
2,735
Let me chime in here and say that using the GPU|CPU fan to intake air directly from outside the case can work; the degree to which they recycle air depends on how close the fans are to the side and how much restriction they have to overcome from tight vent holes and/or filters. The latter is where I think you'll run into problems with direct intake for GPUs (and to a lesser extent low profile CPU coolers and SFX PSUs), since they use slim fans exclusively, and those don't provide a lot of static pressure to overcome restriction. None of the case projects here on the forum using the direct intake style layouts are using dust filters, so the GPU/CPU/PSU fans won't have to work as hard. Likewise, the more open the vents (as on e.g., the NFC S4 mini), the less trouble you'll have with heat build up. For me, not designing for dust filters is not an option, and therefore direct airflow designs (like the LRPC was) aren't viable.

The other issue, and this I tend to agree with Prava on, is heat build up. When you're relying solely on passive exhaust via case overpressure (as, frankly, the M1 was designed, although at least it allows for the case fans to be set to exhaust), you can get areas of stagnant air or air vortices that isn't doing its job of moving heat out of the case. Components that lack their own active cooling - drives, motherboard chipsets, memory, VRMs - can run hot in this situation. Lower TDPs and blower GPUs can mitigate the issue, but of course this comes with its own trade offs (lower performance, noise).

The back-to-back, direct intake style layout was something I came up with myself years ago, long before the M1 project (or any of these other projects) started. But based on my experience since then, it's not something I'll be pursuing. Any designs I do in the future will have forced case airflow with dedicated chassis fans. Unfortunately, this does mean that they can't be nearly as small, but that's the price for proper cooling.
 
  • Like
Reactions: K888D
like this

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Case testing should be interesting. These are my thoughts for hardware for initial testing:

MB: Asus Z170 of some sort
CPU: i5-4690K make that i5-6600K
GPU: 1060 or 1070 itx
PSU: Corsair SF600
Cooler: Big Shuriken 2, and a couple others of the same general size for comparison.
Bottom fan: Noctua 92mm thin
HDD, SSD: tbd

The 1060 and 1070 are great, but they don't put out as much heat as others, so I might change that.

Thoughts are appreciated.
 
Last edited:
D

Deleted member 222586

Guest
Let me chime in here and say that using the GPU|CPU fan to intake air directly from outside the case can work; the degree to which they recycle air depends on how close the fans are to the side and how much restriction they have to overcome from tight vent holes and/or filters. The latter is where I think you'll run into problems with direct intake for GPUs (and to a lesser extent low profile CPU coolers and SFX PSUs), since they use slim fans exclusively, and those don't provide a lot of static pressure to overcome restriction. None of the case projects here on the forum using the direct intake style layouts are using dust filters, so the GPU/CPU/PSU fans won't have to work as hard. Likewise, the more open the vents (as on e.g., the NFC S4 mini), the less trouble you'll have with heat build up. For me, not designing for dust filters is not an option, and therefore direct airflow designs (like the LRPC was) aren't viable.

The other issue, and this I tend to agree with Prava on, is heat build up. When you're relying solely on passive exhaust via case overpressure (as, frankly, the M1 was designed, although at least it allows for the case fans to be set to exhaust), you can get areas of stagnant air or air vortices that isn't doing its job of moving heat out of the case. Components that lack their own active cooling - drives, motherboard chipsets, memory, VRMs - can run hot in this situation. Lower TDPs and blower GPUs can mitigate the issue, but of course this comes with its own trade offs (lower performance, noise).

The back-to-back, direct intake style layout was something I came up with myself years ago, long before the M1 project (or any of these other projects) started. But based on my experience since then, it's not something I'll be pursuing. Any designs I do in the future will have forced case airflow with dedicated chassis fans. Unfortunately, this does mean that they can't be nearly as small, but that's the price for proper cooling.
I agree that it can and it will work, but unless the fan is exactly in contact not all air will come from the outside, and thus you will recirculate hot air inside the case.

Regarding back-to-back layouts... there is a very interesting one that allows for SLI on MATX @ <12L with front intake fans. Flex ATX are much much better than when you designed the M1 and allow for very interesting layouts. Though something tells me you are already working on something similar :)

(TEASER) - Project Orthrus - Smallest SLI mATX case (<11.9L)

It is certainly much bigger than this one (more than double if i'm not mistaken) but allows for more flexibility. Not sure If you have seen it. But imo, active cooling is the right call for builds > 200W.



Case testing should be interesting. These are my thoughts for hardware for initial testing:

MB: Asus Z170 of some sort
CPU: i5-4690K
GPU: 1060 or 1070 itx
PSU: Corsair SF600
Cooler: Big Shuriken 2, and a couple others of the same general size.
Bottom fan: Noctua 92mm thin
HDD, SSD: tbd

The 1060 and 1070 are great, but they don't put out as much heat as others, so I might change that.

Thoughts are appreciated.
In your design you have a length-constrain for the gpu. So, the most powerful (power consumption wise) gpu you could probably ever fit in would be something along the lines of a Fury Nano. All things combined thermals should be ok in your design becase a) you have active cooling and b) you can't put >200W as they simply won't fit in there. Still, thorough testing never hurts, but I wouldn't bother with modern hardware since any card that uses lots of power will be a perfect example. And I don't think blower-style exists in reduced lengths so...
 

Necere

2[H]4U
Joined
Jan 3, 2003
Messages
2,735
Regarding back-to-back layouts... there is a very interesting one that allows for SLI on MATX @ <12L with front intake fans. Flex ATX are much much better than when you designed the M1 and allow for very interesting layouts. Though something tells me you are already working on something similar :)

(TEASER) - Project Orthrus - Smallest SLI mATX case (<11.9L)

It is certainly much bigger than this one (more than double if i'm not mistaken) but allows for more flexibility. Not sure If you have seen it. But imo, active cooling is the right call for builds > 200W.
Yes I've seen it (I did comment in the thread). It has dedicated chassis fans, which is an improvement, but because of how narrow it is it's still reliant on side panel ventilation for the CPU and GPU intakes. Your suggestion of solid side panels wouldn't work (well) as-is, since there's virtually no room for the CPU cooler intake, and very little for the GPU intakes. I consider a 10mm gap to be about the minimum between a fan and a solid panel for adequate intake airflow, so in a back-to-back design like the Orthrus it'd need at least an extra 20mm added to the width. Even so, without any exhaust fans I don't think the design would do well with open cooler cards.

Aside from the airflow concerns, an issue with back-to-back designs in general is the reliance on very long flexible PCIe risers, which can be unreliable and/or cost prohibitive. We see that with the A4, where the 3M riser is probably adding 50% to the cost of the case. The Orthrus doubles down on that, with two risers adding $160 to the cost. Even with a 1k+ volume discount, this isn't going to allow for pricing that is in any way viable for more than an extremely niche product.

Likewise, the reliance on 1U or flex PSUs is problematic, since the former are typically designed for servers and the latter for lower power systems, and in both cases selection appropriate to performance consumer builds is very limited. It's likely that you'd need to include the PSU like e.g. the EVGA Hadron, which adds considerably to the price of the case and introduces its own concerns with warranty support and regulatory compliance. For a one- or two-man show it's an added layer of complexity and cost that can end up sinking the whole operation.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Well, I have the Li-Heat 300mm PCIE 16x 3.0 shielded riser and will be testing it to confirm Li-Heat's tests. It is less flexible than the 3M, but fits in the case and the price looks very attractive at $20-30 depending on qty.

Of course the proof is in the pudding, but I'm confident based on Li-Heat and other user tests.
 

vipz

Gawd
Joined
Apr 11, 2005
Messages
818
Have you considered adding ~30mm to the height of the case to allow a 120mm filtered intake at the top? You could maybe reduce case depth and still fit a good amount of 2.5" drives in the bottom. You could flip the fan and PSU locations if you want it intaking from the bottom.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Have you considered adding ~30mm to the height of the case to allow a 120mm filtered intake at the top? You could maybe reduce case depth and still fit a good amount of 2.5" drives in the bottom. You could flip the fan and PSU locations if you want it intaking from the bottom.
Vipz,
Interesting concept. So throw a big 120 on top to pull air in, or push air out. That won't help case depth though, since that is driven by the MB 170mm plus 20mm for the SSD/HDD mount on one side and the GPU length of up to 7.4" plus airflow clearance on the other.
 

vipz

Gawd
Joined
Apr 11, 2005
Messages
818
I was hoping you could drop the SSD/HDD mount next to the mobo and reduce supported GPU length. Worth trying :D
 

Boil

[H]ard|Gawd
Joined
Sep 19, 2015
Messages
1,421
Whoa…!

Got rather ugly in here for a minute…!

Getting stoked all over again on this chassis…!

This Corsair SF600 SFX PSU is longing for home…!

No pressure…!

;^p

(…okay, maybe a wee bit o' pressure…)
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Since we had one meaty, heated discussion, how about another? Hmmm, something important to the development of this case, yet different than throttling SSDs (or each other).

How about this?



WHAT NICKNAME SHOULD THE MI-6 HAVE?

I know, I know, it's been bugging you too, especially with all the cases already out there with short letter and/or number designations. Well now is the time to give me your thoughts, or 2 cents worth. (Note 2 cents is the minimum. 1 cent submissions will be returned postage due).

My thoughts are a name that goes along with the intent of the case, small and light, yet can be stuffed with performance hardware. Some ideas I have been kicking around:

Pitbull -- Small yet tough, and will eat your lunch and everyone else's, but does have an earned bad reputation.
Tazmanian Devil -- Fun name, but have you seen the actual critter? Pretty ugly.
Badger -- Pretty good name. Not really exotic like a TD, but might be fitting.
Honey Badger -- African badger that some say is even tougher than the American or European badgers. Sounds good if you are familiar with it, otherwise it sounds like a nice little honey-eating badger.
Wolverine -- Bigger than a badger. For those that don't know, it's a rare north american predator that seemingly always has a bad attitude and ready to fight.
My Ex -- Kinda self-explanatory to most, but maybe not that good of a selling point...
Ninja -- dunno.

And the names considered and discarded as really bad ideas, ugly varmints or just not setting the right tone for a fun powerful computer case are: Mongoose, ferret, meerkat, hyena, guinea pig, fox, stinkin squirrel (sorry I always include stinkin when I see the squirrels in the backyard), komodo dragon, peanut, acorn.

Peeps with good ideas get a free link to the thread posts (-at exactly the same time everyone else sees it), so you will be sure not to be left out.
 

Boil

[H]ard|Gawd
Joined
Sep 19, 2015
Messages
1,421
Badgers…? We don't need no stinking badgers…!

 
Last edited:

iFreilicht

[H]ard|Gawd
Joined
Sep 23, 2014
Messages
1,348
Well, I have the Li-Heat 300mm PCIE 16x 3.0 shielded riser and will be testing it to confirm Li-Heat's tests. It is less flexible than the 3M, but fits in the case and the price looks very attractive at $20-30 depending on qty.

Of course the proof is in the pudding, but I'm confident based on Li-Heat and other user tests.
Even if your riser works (as mine does right now), be aware that LianLi pulled the PW-PCI-E and PW-PCIE38 from market, which were produced by LiHeat and now bundles the PW-PCIE30-1 and PW-PCIE38-1 with their riser-dependant cases. They have the same retail price of about 64€ and their cases didn't change in price either, so I suspect this riser - which looks quite similar to the 3M one - is much more reliable at a similar price point. Now we'd only need to find out who makes it.
 

klatox

Weaksauce
Joined
Feb 13, 2013
Messages
125
I got a nickname for ya. With all the talk of heat recently, and the look of your case, I immediately thought of this. Probably not a selling point though! :p Regardless, I will be putting a stencil of a piece of toast on mine! Can't wait to see it in black.

THE TOASTER.


 
Last edited:

Boil

[H]ard|Gawd
Joined
Sep 19, 2015
Messages
1,421
I got a nickname for ya. With all the talk of heat recently, and the look of your case, I immediately thought of this. Probably not a selling point though! :p Regardless, I will be putting a stencil of a piece of toast on mine! Can't wait to see it in black.

THE TOASTER.

That made me think of the Amiga 2000 & the NewTek Video Toaster…
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
I got a nickname for ya. With all the talk of heat recently, and the look of your case, I immediately thought of this. Probably not a selling point though! :p Regardless, I will be putting a stencil of a piece of toast on mine! Can't wait to see it in black.

THE TOASTER.

LOL! Hilarious tie in to the heat discussion.

Man, now why did I invite you to the party?!
 
Joined
Sep 2, 2015
Messages
27
The case gets hot because you don't have any means of exhausting the heat out of it. Also, you should know that the temperature of the gpu is completely meaningless because temperature depends on the noise the gpu is outputting. In any case, tell me, how good are drive temperatures after 2 hours of gaming? How much worse would they become if you were to put a non-blower card in the system?

This is not about believing or not believing, this is about extensively testing or not. For you it might be okay to force your drives to work at up to (or even more than) 50ºC, for me it simply isn't. I'm okay with that. But I'm not in users providing information that isn't factually tested, specially those that have interests because they are selling a product.

PS: you should know that many NVIDIA cards throttle at more than 83ºC. So, your 82ºC aren't exactly good...because the card won't simply go above that mark.
he's right about this part, mobo and drive temps skyrockets when there isn't some way to exhaust the hot air, and after some time GPU and CPU will get hot too, my ssd will have to run at over 55oC when gaming in my 10L cube and I can't do anything about it
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
he's right about this part, mobo and drive temps skyrockets when there isn't some way to exhaust the hot air, and after some time GPU and CPU will get hot too, my ssd will have to run at over 55oC when gaming in my 10L cube and I can't do anything about it
What cube do you have? Lian Li?

Putting the 92mm fan in the bottom of the Mi-6 blowing right at SSDs should help keep the hot case air away.

I am looking at increasing the exhaust vent area, but testing will drive that.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
Even if your riser works (as mine does right now), be aware that LianLi pulled the PW-PCI-E and PW-PCIE38 from market, which were produced by LiHeat and now bundles the PW-PCIE30-1 and PW-PCIE38-1 with their riser-dependant cases. They have the same retail price of about 64€ and their cases didn't change in price either, so I suspect this riser - which looks quite similar to the 3M one - is much more reliable at a similar price point. Now we'd only need to find out who makes it.
Hmmm, MNPCTECH is still selling the "old" Li-Heat style riser here:
PCI-E Express Gen 3.0 Riser Card Extension Cable Black

One thing I did notice is that the riser I got from Li-Heat is different than the one listed on Lian-Li (and mnpctech.com), as mine has strain-reliefs where the flat cable is soldered to the boards. Maybe that ws response to a Lian Li issue?

Here is a bit of info on the 3 different risers:

I will be testing with the latest Li-Heat version that has the strain reliefs.
 

Firewolfy

Gawd
Joined
Sep 11, 2015
Messages
579
We get any more "good" ideas and I'll just stick with letters and numbers, like the "Mi-6.TO/RoomHeaterV.07"
 
Top