Was there ever an explicit justification for changes in ATX case design?

[BonzaiDuck]

The question I posed is too general. So let me be more specific.

I'm currently "resurrecting" and modding a CoolerMaster Stacker 830 midtower purchased around 2008.

These ATX cases have changed. The newer designs -- several of them -- put the PSU at the bottom of the case. The older case designs like that of my Stacker follow the traditional approach of putting the PSU in the top-rear of the case.

What were the advantages touted for putting the PSU in the bottom? Was this done to provide more options for water-cooling and radiators installed in the case top-panel? Was there ever anything written up about it?

 

[YBS1]

I always assumed it was done for two reasons. 1) Made the case less top heavy. 2) Made cable management much better, back then modular PSUs were pretty rare.

 

[JeffMD]

When PSUs started becoming so efficient that they no longer run with fans unless under load, they stopped becoming a top mounted exhaust fan for your case. Also as CPUs and GPUs get bigger, heat becomes a bigger issue. So allowing the user to mount his own cooling system at the top made more sense.

Cooler clearance is another, larger coolers simply weren't being given enough clearance from the PSU.

And then there is aesthetics, In a windowed door it is easier to blot out the PSU in the bottom and leave the top and expansion bays open for lighting accessorizing.

 

[aigomorla]

PSU's also got more and more massive..

Having a 1kw+ hang on the top via 4 screws could cause the case to warp at the top if it experienced shock vs having it lay down on the case floor.

In 2008 you'll be hard pressed to find a PSU as long as per say today's corsair 1kw+ psu's

 

[Newbian]

Not to mention letting psu's use fresh air instead of the heated air from inside cases extends their life.

 

[BonzaiDuck]

I always assumed it was done for two reasons. 1) Made the case less top heavy. 2) Made cable management much better, back then modular PSUs were pretty rare.

I never noticed an advantage either way for managing cables.

In the recent past, I had to proceed thoughtfully for connecting an eight-pin CPU/motherboard plug. For my purposes, I had to scour my parts-locker to find left-over modular cables from the same manufacturer of a previous PSU. But the cable bundled with the PSU was just too short to avoid routing it in ways satisfying neither me or others like me.

I can't disagree with any of the arguments put forward by all the respondents to this thread so far.

Since 2008, every case I've purchased was a CoolerMaster. I have the Stacker 830, and three HAF 932's. The HAFs are designed for the PSU in the bottom, pulling air in from case bottom and exhausting it. I've installed PSUs both ways in these HAFs: pulling in outside air for the PSU, and using inside case air (with the PSU fan facing upward).

The Stacker is a "resurrection" project. It hasn't been used for almost two years, and was decommissioned temporarily when I replaced the system my brother uses upstairs.

I'd post pictures, but haven't taken any yet. My great "mod" is maybe 2/3 finished. It completely separates the air intake cooling the storage devices and the CPU from the intake that cools motherboard components. I've put a Lexan duct over the motherboard, from which the CoolerMaster "CrossFlow" fan exhausts air to the case right-side-panel.

Any "excess pressure" from the two sets of intake fans ultimately forces air through the only outlet available: the PSU. So I expect the PSU to stay cool as a cucumber, or as cool as you could make it any other way with any other case.

Truth is -- I do like having the PSU in the case-bottom over the top-mounted location for it. I suspect I could even rebuild the Stacker to get the PSU in the bottom, but obviously, the prospects include a lot of work. And for what? I've spent a month fiddling with this case, thinking about what I was planning to do, taking measurements, cutting Lexan and art-board. If the PSU is in the case top-rear, then that's where it is, and that's where it's gonna be.

 

[piasabird]

Heat rises. If you put the power supply in the bottom, heat will not rise up to it and make it run harder.

There are obvious advantages and disadvantages to different placement of a power supply like putting it in the front of an ITX case.

 

[BonzaiDuck]

Heat rises. If you put the power supply in the bottom, heat will not rise up to it and make it run harder.

There are obvious advantages and disadvantages to different placement of a power supply like putting it in the front of an ITX case.

That was always my take on top-mount versus bottom mount. But increasing temperatures inside the case are not a major concern for me.

I plan every air-cooled build for an imbalance between intake (higher) CFM and exhaust (slightly less) CFM, and for maximum CFM "exhaust potential."

This latest quasi-semi-whatever-project was just something I've only done maybe twice before -- ducted cooling. Now ASUS "Sabertooth" line has given me some ideas that don't require the intricacy of Sabertooth. The motherboard portion of the exhaust has the lowest CFM potential. The CPU will have ducted air from two front-mounted 140mm Akasa Vipers -- after initially passing over every single storage device in the case. The case exhaust (functioning as the "pull" side of the CPU-fan(z)) has high CFM potential. The Akasas are rated at 110CFM with probably modest static pressure. The iPPC-3000 120mm exhaust fan's spec is just about that: 110CFM, higher static pressure. The Akasas are about double the CFM throughput of the exhaust, and they're all connected through a duct with the CPU cooler in the middle.

That still leaves the two side-mounted 140mm fans feeding the motherboard duct intake, the rest of the case for what space is left in it, and the PSU at the top. They are currently either "Shark" or "Sharkoon" fans, and will be replaced with another set of Akasas, or I might then choose to reduce the number of fans in the case, while continuing to segregate intake-to-exhaust for motherboard versus drives and CPU -- with all surplus going to the remaining case space and out the PSU.

I think I've finished the basic motherboard plate, or "part 1." "Part 2" seals off the storage->CPU intake and exhaust from the rest of the case. I might post a picture later, after I gather the resolve to grab my camera and just "do it."

Since airflow through the duct is restricted deliberately, there should be plenty airflow going through the PSU at the top. I'm sealing off every little aperture, and that's the only place for it to go.

 

[bigboxes]

I've had bottom mount PSU cases since at least 2006, maybe 2005. That's where they've been ever since.

 

[guachi]

Lian-Li explicitly mentions case stability as a positive reason for bottom mount PSUs.

Also, ducted cooling is pretty neat. I saw some Compaq(?) machines in 2000 that had ducted cooling. No fan was present on the CPU. Rather it was a large case fan ducted to the CPU that provided the air flow. It certainly made things quieter.

 

[Sheep221]

I use only top mounted PSU cases. Using PSU as back exhaust.

 

[dlerious]

Lian-Li explicitly mentions case stability as a positive reason for bottom mount PSUs.

Also, ducted cooling is pretty neat. I saw some Compaq(?) machines in 2000 that had ducted cooling. No fan was present on the CPU. Rather it was a large case fan ducted to the CPU that provided the air flow. It certainly made things quieter.

Speaking of Lian Li, I have a tower case , PC-7 something, that has top mount where I could mount at the very top (with fan under) or flip and put PSU under the fan.

Using Caselabs cases now.

 

[BonzaiDuck]

Speaking of Lian Li, I have a tower case , PC-7 something, that has top mount where I could mount at the very top (with fan under) or flip and put PSU under the fan.

Using Caselabs cases now.

Cases designed to provide such an option are exceptional. Once, I found an old IBM PC case from a Windows 97 system. It had a "cage" for the PSU which "unlocked," "unhinged," and simply "popped out." So I cut the cage and surrounding sheet metal from that case, and put it in a '95 ProLiant Server case which I modded in 2007. But you couldn't simply "rearrange" things and put the PSU somewhere else. I probably should have made provisions for that during the modding work, but then -- who would expend the energy and time?

I think if the CFM throughput for the case is more than adequate (#1), and #2 -- certain sources of heat and therefore heated air are segregated and immediately exhausted, it wouldn't matter so much having a PSU in the top-rear.

But I still guess I'd prefer a bottom-mount PSU. That being said, this Stacker 830 case is "NOT FOR SALE." I've had this case either in use or sitting around for eight years, and I've suddenly discovered more ways to use it in the future. This includes what could be some very neat water-cooling configurations and air-cooling configurations.

I just wish it had only two USB 2.0 ports on the front-panel instead of four, and I wish it had USB 3.0 and eSATA ports. I solved that problem, with a 3'5" 2-port USB 3.0 panel in front, and an eSATA 2-port panel installed at the very case-top front in a recessed area that is supposed to be a "lifting" handle.

By the way -- here's a nifty little adapter for about $7:

http://www.newegg.com/Product/Product.aspx?Item=9SIAB274850329

Who needs the PCI plate? I had a beige 3.5" plastic bezel from an old Gateway case from the '90s; cut the holes for the eSATA plugs and drilled holes for the screws; then built around the plastic bezel with foam art-board for a piece that simply pops into the "handle-recess" of the Stacker and locks in place with an interference fit. Cut the SATA holes at the back of the recess, connected them to internal SATA ports, and "Voila!"

Yes -- we have fun with seemingly tedious projects. . . .

 

[mv2devnull]

CM Stacker STC-T01 (published 2004?) had two places for PSU's: the default at the top and a second at the bottom. The formal idea was to have additional PSU. (As said, "big" PSU were not that big yet, dual CPU was still a thing before multi-core, and there were plenty of HDD slots.) Furthermore, you could reassemble it as BTX case.

Anyway, I did (eventually) move the PSU down and add two 80mm exhaust fans at the top PSU's place.


Bit old, bit large, bit noisy, but still in use. (Not as primary though.)


Heat, mass, stability, looks ... could it be that a need to be different also did support the bottom-mounted PSU-design? In the market of beige and standard anything that deviates is bound to catch attention.

 

[BonzaiDuck]

CM Stacker STC-T01 (published 2004?) had two places for PSU's: the default at the top and a second at the bottom. The formal idea was to have additional PSU. (As said, "big" PSU were not that big yet, dual CPU was still a thing before multi-core, and there were plenty of HDD slots.) Furthermore, you could reassemble it as BTX case.

Anyway, I did (eventually) move the PSU down and add two 80mm exhaust fans at the top PSU's place.


Bit old, bit large, bit noisy, but still in use. (Not as primary though.)


Heat, mass, stability, looks ... could it be that a need to be different also did support the bottom-mounted PSU-design? In the market of beige and standard anything that deviates is bound to catch attention.

Yes -- that was the "original" Stacker I'd looked at before getting a '95 ProLiant Server case for free which I modded extensively.

After that, given the size of my room, the number of computers, I never used full-tower (or bigger-than-full-tower) after that. But apparently, with the Stacker 830 midtower, I'm not sure that an alternative location was part of the design. I'd actually have to check, because there's a "BTX" option for the midtower Stacker. Frankly, I don't much care for the unlatchable, removable motherboard tray and back-panel. I couldn't see an advantage to it: if you want to "work on" your motherboard, you'll either do so with it in the case, or you'll remove it from the case. If you remove it, you still have to unplug everything. So what good is that?

I think the "good" of it was the alternative BTX configuration, and that's why the L-shaped mobo-tray is also "removable."

Just looked again at the guide/instruction PDF in the CM archives:

http://www.coolermaster.com/xresserver01-DLFILE-P1302180011eadb-F130323001191c4.html

Changing to BTX leaves the PSU in the top position.