A bunch of pictures

[Saber]

Could the case have some kind of perimeter venting(basically a slit that extends around the exterior) like they have on the Microsoft Surface? Not all the way around but on the sides disguised by a ridge pattern?

 

[Linux-SWAT]

The CPU being on it's own card, at least it can be extended with some thermal-conductive material.

Spreading out is another thing. Maybe on the edges, as suggested.

 

[EvilDragon]

Making any open holes in the case won't help much with the heating unless you include a fan to specifically lead it out there.

It's not like heat will magically leave through small holes if you put them there.

That's also why smartphones and tablets have no venting holes.

 

[Linux-SWAT]

I meant the thermal conductor being long enough to reach the edges of the device, therefore being in "contact" with the outside.

_____

_____|

_____| <----conductor

_____|

 

[Saber]

Making any open holes in the case won't help much with the heating unless you include a fan to specifically lead it out there.


It's not like heat will magically leave through small holes if you put them there.


That's also why smartphones and tablets have no venting holes.

What if they were magic holes?

Any way to add some metal in the stylus slot or better still moving the slot to below the PCB stack?

Further brainstorming, if metal strip lines were spread like say railroad tracks around the area to the USB ports/Sim bay, that would draw some warmth away from the heat core. Probably no need to do this though since you're already bridging to the SD cages.

 

[Grench]

I just had another idea on this...

On the Pandora there is a thin sheet of plastic between the motherboard and the battery. It's part of the lower plastics piece.

What if this layer were replaced with a thin sheet of aluminum extending to the left, front, right edges?

Something like this pad between the back of the SoC and the aluminum sheet:

http://www.amazon.com/Cool-Silver-Pad-Overclocking-performance/dp/B002VM7GO8/

At the edges of the Pyra, the aluminum sheet could wrap down around the front/rear plastics.

Again, it may be that none of these twists and tricks are even needed - ED & team will figure it out with real world testing.

What I would use as a basis/scenario:

Under full (or typical) load, can it shed enough heat without air flow in a constant temp, stagnant box at ambient 40*C to prevent throttling back?

 

[EvilDragon]

That would give more heat to the battery, since plastic is an insulator.

 

[Grench]

That would give more heat to the battery, since plastic is an insulator.

It would combine the heat from the SoC and heat from the battery - and give it a direction to go out of the case and dissipate to the air more efficiently than the case walls.

Yes, if the SoC was significantly hotter than the battery, it would use the battery as a heat sink along the way.

It's getting the heat -out- of the case that concerns me. Otherwise it's a continual heat soak and build up - with no end. The thing could go supernova.

 

[DrHAX]

That would give more heat to the battery, since plastic is an insulator.

It would combine the heat from the SoC and heat from the battery - and give it a direction to go out of the case and dissipate to the air more efficiently than the case walls.

Yes, if the SoC was significantly hotter than the battery, it would use the battery as a heat sink along the way.


It's getting the heat -out- of the case that concerns me. Otherwise it's a continual heat soak and build up - with no end. The thing could go supernova.

I'd rather have a CPU die then a battery explode in my hands.. IMHO..I like having hands

 

[EvilDragon]

Let's try to prevent both

 

[Grench]

Let's try to prevent both

Exactly why that thin sheet of aluminum between the SoC and battery transporting heat outside the case as I've described above makes so much sense to me. Help the heat from both the SoC and the battery exit the case.

 

[EvilDragon]

It won't work like that, that's the problem.


Heat evenly distributes on a heat sink.


Without a heat sink, the SoC might have 110C.


With a 5x5cm heatsink, it might go down to 80C evenly distributed on the heatsink.


With a 10x10cm heatsink, it might go down to 50C evenly distributed on the heatsink.


Now, if you have a small opening in the case, there won't be much heat. And it won't go out, as only a tiny little fraction of the full heatsink will be outside.


Ever opened a Laptop?


The heatsink usually has a small copper pipeline to the side of the case, with lots of small copper plates besides the case opening (most of the heat goes there, as this is the largest area of the heatsink).


With the fan running, the air outside runs through the copper plates and exchanges the heat, thus cooling the copper and heating the air that gets blown out again.


If you disable the fan, the laptop will overheat within a short time frame.


The heat doesn't go out by itself. If you plan holes for heat exchange, you also need a fan.


The trick for us is to create a copper area as huge as possible (multiple PCB layers) to distribute the heat as much as possible.


The bigger the area, the lower the overall temperature.


The device will get handwarm and the whole case surface will exchange the heat with its' surroundings.

 

[rygD]

The device will get handwarm and the whole case surface will exchange the heat with its' surroundings.

An optional aluminum case would help with this. 

 

[Linux-SWAT]

So the flame logo takes all its meaning ^^.

 

[comradekingu]

Heat does not dissipate evenly on a heatsink. It is hotter in the centre, or closest to where the dies are underneath, which produce the heat. Then it tapers off as it travels distance over which it is cooled by air.

We must obey the laws of thermodynamics. What kind of household would we be if we didnt?

 

[fusion_power]

It won't work like that, that's the problem.

Heat evenly distributes on a heat sink.

Without a heat sink, the SoC might have 110C.

With a 5x5cm heatsink, it might go down to 80C evenly distributed on the heatsink.

With a 10x10cm heatsink, it might go down to 50C evenly distributed on the heatsink.

Now, if you have a small opening in the case, there won't be much heat. And it won't go out, as only a tiny little fraction of the full heatsink will be outside.

Ever opened a Laptop?

The heatsink usually has a small copper pipeline to the side of the case, with lots of small copper plates besides the case opening (most of the heat goes there, as this is the largest area of the heatsink).

With the fan running, the air outside runs through the copper plates and exchanges the heat, thus cooling the copper and heating the air that gets blown out again.

If you disable the fan, the laptop will overheat within a short time frame.

The heat doesn't go out by itself. If you plan holes for heat exchange, you also need a fan.

The trick for us is to create a copper area as huge as possible (multiple PCB layers) to distribute the heat as much as possible.

The bigger the area, the lower the overall temperature.

The device will get handwarm and the whole case surface will exchange the heat with its' surroundings.

Heatpipes should work with an passive solution, there are PC's with only passive cooling, no fans at all and the heat transportation still works. So even without fans, an heat transportation through holes (or radiator copper outside of the case) would not be completely useless imho. Air tends to move if there is an temp difference -this is how wind is generated. Of course I know that there is not really space for something like heatpipes within the Pyra. ^^" Overall, it's a very dense "Sandwich" with very limmited space for cooling.

 

[EvilDragon]

Can you show me a picture of a PC which has a passive heatsink with holes in the case to let out the heat?

The only passive ones I know are simply heatsinks that are fully inside of the device.

 

[fusion_power]

Can you show me a picture of a PC which has a passive heatsink with holes in the case to let out the heat?

The only passive ones I know are simply heatsinks that are fully inside of the device.

It does not need "open" holes, only a radiator-thingy that goes through the case to the outside somehow. However we can stop all the speculation if you show us your solution for that heat problem, I'm not an expert and only suggest suggestions like others.
For your question (Yes I know these cases are metal but the principe is the same)

http://www.meisterkuehler.de/forum/vorstellung-stromspar-pcs/32352-vorstellung-stromspar-passiv-htpc-selber-gebaut.html

http://quietpc.de/zen-nt-tx2000

 

[bzar]

What heat problem? So far no problem has been shown to exist or found to be probable as far as I'm aware. Please correct me if one has presented itself.

 

[klapse]

Can you show me a picture of a PC which has a passive heatsink with holes in the case to let out the heat?


The only passive ones I know are simply heatsinks that are fully inside of the device.

It does not need "open" holes, only a radiator-thingy that goes through the case to the outside somehow. However we can stop all the speculation if you show us your solution for that heat problem,

ED already explained how heat will be dissipated - from CPU throughout the whole PCB, then to the case then to the outside world. At least that is how I understood him.