Completed MJOLNIR: The minimalistic – but not boring – 9.7L Case

[wolfatpno]

Maybe the two magnets don't need to meet. A small gap of 0.3mm or so should prevent cracks.

I think the magnets need to be slightly recessed anyway to guarantee a flush transition from the panel to the rest of the case.

Good point.

I was about to write down the third way.

Exactly what you said.

The only concern is that as time goes by, the glue will fall off due to the tensile force of the magnet.

 

[wolfatpno]

$193,551

pledged of $61,611 goal
463
backers


It's getting closer to 200,000

 

[keepcoding]

With regards to 240 AIO. Xspc TX series is 20.5mm thick and Silverstone's is 22mm.
Difference is that xspc rads are able to keep up with likes of higher priced 30mm rads from likes of Hardware labs etc

Problem is not the performance of the Radiator, it is the lack of a 2nd fan that makes a 240mm AIO unfeasible in this case.

I would definitely vote for scraping the 240mm AIO support all together and lower the MoBo slightly for more riser cable clearance.

 

[Goatee]

Problem is not the performance of the Radiator, it is the lack of a 2nd fan that makes a 240mm AIO unfeasible in this case.

I would definitely vote for scraping the 240mm AIO support all together and lower the MoBo slightly for more riser cable clearance.

For my use, I'm interested in the height in green?

If thats >31mm then a slim 120mm fan (10mm) should fit in with a TX radiator and my planned custom loop will work nicely. If not then I might need to get creative with 40mm fans.

Secondly, @AlexTzone would you consider adjusting the the rear design to allow a customisable plate to be inserted above the PCIE card holder? I seem to remember some other premium cases have this and its nice to have something completely unique. Excuse the crude paint example.

Those backer plates could then be optional (paid) extra's or a stretch goals? Perhaps with the normal GPU holder bracket meshed with screw holes to allow airflow? Backer plates could be sent separately (while cases being made?) from the case so you don't need to worry about trying to get the right plate in the right case box?

 

[For_Science]

For my use, I'm interested in the height in green?

23 mm, remember we can't even fit a normal 25 mm fan there.

 

[Goatee]

23 mm, remember we can't even fit a normal 25 mm fan there.

Guess I'm getting creative with 40mms then

 

[For_Science]

Guess I'm getting creative with 40mms then

I think the problem with custom loops in this kind of chassis is the space gained on the GPU compartment side is wasted. Obviously the cooling capacity of the system is dependent on how many watts are being cooled vs the available surface area. The conversion of an aircooled GPU to a watercooled one gets you space in the form of 1~1.5 slots of PCIe expansion card, but that space cannot be used to house a radiator/fan of any sort. And so in effect you've lost a nice chunk of heatsink only to be replaced by a smaller radiator, which may or may not be having to deal with a CPU as well.

On the other hand, the CPU cooler supported here are small, whereas a 120 mm radiator can provide similar cooling to the likes of a mid-tier tower cooler and provide clear benefit for their implementation.

 

[Goatee]

For me its more about moving the heat to an area its easier to move away from the case. I understand your point but there are options to utilise that GPU space, for example hybrid GPU's like a POSEIDON-GTX 1080TI.

 

[AlexTzone]

Out of curiosity how tall are the case feet? It's another example of a form vs. function trade-off as with the panel vents. The lower the case is to a desk surface the better it looks, but the harder (and possibly noisier) it is for fans to pull in air. My personal feeling here is having the gap the same height as the thickness of the case would be visually pleasing and probably sufficiently roomy. Maybe the feet are already this height? (5mm)

I worry slightly about the new magnetic panels being less stable than 4 screws, but eliminating lightbleed is a pretty great perk. If it's cheap I think I would feel better about them with a magnet on each side.

So the 240 AIO is, as was predicted, a disaster and no one should do it under any circumstance. Absolutely not surprised there. As a result though I think you should eliminate compatibility for it completely. It's not a situational option; It's a BAD option and given the ability to choose poorly, people always will and then complain about it. An even greater reason to do so though... I saw comments regarding Ali's review video on Reddit and Youtube about the evident wear and tear on the PCIe riser cable as a result of sliding it in and out repeatedly. Maybe he could've been more careful, but most of the community trusts that Ali puts in atleast above average effort into building SFFPCs and furthermore there's no longer a reason for the riser to be that close. IIRC, an early revision of the Mjolnir had the mobo lower in the case and it was raised for the purpose of adding 240mm compatibility. Now that we've established using a 240mm AIO is definitively a bad idea the mobo can be lowered to its original position giving a couple mm's of leeway back to the riser cable. (If the raised mobo improves compatibility for particularly long 120mm AIOs maybe it's still worth keeping.)

The feet are 5mm tall, but the distance to the radiator/fans is 15 mm. I never see the feet when sitting normally (it kind of looks like MJOLNIR is floating slightly above the table). We're super OCD when it comes to dimensions. If thickness of the shell is 5mm, then the height of the feet and the Glass panel / O2 panel frames have to be 5mm wide etc etc ?

Magnets. The are different grades of neodymium magnets, and it looks like we're going for the strongest ones just to have that extra support and avoid things like @keepcoding mentioned (AIO tubes pressing against the panels).

240mm AIO. That 23mm space can be used somewhat creatively (mod in compatibility for 25mm thick fans, fan shrouds, etc) so I'm not totally sure yet if this should be decreased. As long as you tuck in the riser before inserting the Core into the Shell, it's going to be fine.

Looks cool, I assume while now in theory you could change the panels while the core is inserted, the idea is still to remove the core, and pop them out from the inside right? I assume the tolerances of the panels will be such that trying to pry the panel out with nails are going to end with broken nails

I'm completely happy with the flat feet

Yeah indeed, but it's not going to be easy! ? You can mount something to the slots in the performance panels that lets you pull them out (like the GHOST's top panel), but I don't know how this would be done with the TG panels. 100% sure nobody here has 0.125mm thick nails made of adamantium either, so the most convenient way to remove them will probably be by extracting the core first.

Agree with @thekinghippo, scrap the 240mm AIO compatibility, it is just not reasonable to further promote it.

Also, I think magnetic panels are generally a nice thing, but I am a bit concerned about stability. Since the case is only press-fitted (there is not a single screw holding the case together), the screwed-in side panels gave me some piece of mind about stability.

Another concern is pressure towards the side panels from internal components. What if the AIO tube sticks out a bit and exercises pressure against the panels? I think it could fall off.

See further up for a comment on your concern regarding pressure from the AIO tubes, etc

Don't underestimate the press fit, I can guarantee it's a LOT stronger than the tiny M2 screws that held the panels in place before. Had to go down to a factory to remove one of the press-fitted dowels we tested with. Now think that the case if held together with 16 of them, I don't think it's even possible to separate it without some special equipment. The flexible spiral roll pins gave us a hard time as it is (separating the parts before sending them out for anodizing) ?

Maybe the two magnets don't need to meet. A small gap of 0.3mm or so should prevent cracks.

I think the magnets need to be slightly recessed anyway to guarantee a flush transition from the panel to the rest of the case.

Good point.

I was about to write down the third way.

Exactly what you said.

The only concern is that as time goes by, the glue will fall off due to the tensile force of the magnet.

There will be a tiny gap between the magnets (0.25mm) ?

 

[For_Science]

For me its more about moving the heat to an area its easier to move away from the case. I understand your point but there are options to utilise that GPU space, for example hybrid GPU's like a POSEIDON-GTX 1080TI.

The Poseidon series would be indeed an alright solution, however it is too tall for the MJOLNIR so won't fit anyway.

 

[Goatee]

Completely, I was a bit unclear there I was just using it as an example.

 

[Sharknose]

With regards to 240 AIO. Xspc TX series is 20.5mm thick and Silverstone's is 22mm.
Difference is that xspc rads are able to keep up with likes of higher priced 30mm rads from likes of Hardware labs etc

The XSPC TX240 radiator is also 125mm wide - the case can fit radiators up to 120mm, so it won't fit. As far as I've found, the Silverstone is the ONLY 240mm radiator that would fit.

 

[keepcoding]

Optimum Tech's final review is out:



Looks like PSU bracket flip is the way to go to get the best cooling performance.

@AlexTzone: Speaking of bracket flip, did you have time to check the price for the slot cutouts?

 

[Goatee]

The XSPC TX240 radiator is also 125mm wide - the case can fit radiators up to 120mm, so it won't fit. As far as I've found, the Silverstone is the ONLY 240mm radiator that would fit.

That's a good spot. Looking at a model of the TX even trimming the fan shroud on both sides (bits in red) will only reduce it to 123mm wide which I think is the width of the end tanks anyway.

 

[For_Science]

Dug it out from page 22, list of common AIOs and custom radiators that will or won't fit in the MJOLNIR, red and purple will not fit, green should fit according to the manufacturer dimensions.

 

[Goatee]

I think this double rad should work: https://www.aliexpress.com/item/1094652141.html? - 270x24x120 mm

Looks decent enough and I think the thickness of 24mm is based on the highest point of the rad, which looks to be the G1/4 ports.

 

[AlexTzone]

Heey guys, updated the tempered glass panels:

Snap-in rather than glue-on (and magnetic ofc) ? It was painful seeing the old ones on Optimum video today...

 

[For_Science]

Heey guys, updated the tempered glass panels:

Snap-in rather than glue-on (and magnetic ofc) ? It was painful seeing the old ones on Optimum video today...

Can the glass be non-destructively removed from the frame? I was hoping to mod one of the glass panels to be a mesh panel...I was intending to heatgun the glue out, but now without it, will the glass panel have any way or taking it out, short of smashing it?

 

[For_Science]

As an additional question, does this mean the glass is not flush this he shell? looks like it will be recessed. Personally i think flush glass would look better.

Also with thid new magnetic design, any scope to make the panels larger (since you dont need space for the screw knobs)? does this look any better?

 

[Aviswa]

As an additional question, does this mean the glass is not flush this he shell? looks like it will be recessed. Personally i think flush glass would look better.

Also with thid new magnetic design, any scope to make the panels larger (since you dont need space for the screw knobs)? does this look any better?

like that idea, could leave more room for increased ventilation