KK: We'll use heat pipes and a custom cooling solution - PSU could sold separately

I have read reports of the Xbox 360 itself overheating, and this being due to the hard drive on top blocking airflow into the system. Most users seem to be able to rectify the problem by removing the hard drive, and as soon as they replace it the problem returns.

 

OT: judging by that photo ive gotta say the ps3 looks like an ugly beast, the xb360 looks way nicer (which certainly was an improvement on the xb1)

 

I thought the original PS2 looked bad when I first saw pictures of it. Shrink it down, though, and it doesn't look that bad.

 

I seriously doubt this is true. Not only are there ventilation openings in the HDD module's top AND bottom side, fact is there are plenty other openings around various faces of the xbox that aren't blocked by the HDD in any way whatsoever. Considering how the internal ducting works in the 360, it doesn't really matter from where the air enters anyway, since there's only way for it to exit - through the duct and out the back.

What caused you to post that spurious and irrelevant thought into THIS thread in particular? Maybe you need to read through this post again, if you ever did so to begin with.

 

I have been looking at the internals of the 360 and trying to figure out how airflow works. Anandtech took one apart and has some good pictures.

The GPU has a pretty puny heat sink on it, that's the first thing I notice. Second, is that the CPU is going to hog all the airflow. What's worse is that the hard drive covers the vents closest to the GPU. The ductwork also doesn't appear to help the situation. If I had to design the cooling system, I'd have taken XCPU's heatsink, made it twice as deep (fins are twice as long, not twice as many) and put the GPU and CPU in series, sharing the same heatsink.

 

You call taht a PSu? ...Its a whole other system.LOL.

 

Obviously. My point was that an increase in ambient temp for the PSU and an increase in heat dissipation inside the case (for the PSU) does not lead one to believe that the PSU could be moved inside the case without changes to the existing cooling solution. It is quite possible that either the PSU would require a larger heatsink, or the main cooling system would require a beefier fan, or both. Any of which, when combined with the resultant larger case, might have led to an unfavorable result (in terms of both production cost and marketability) than the current solution.

 

Sometimes heat transfer is not intuitive. It doesn't make sense that moving hotter air over a surface as opposed to letting it sit in cooler air will cool it better. There is a break point where your fluid is moving too slow or the temperature is too close to that of what you're trying to cool.

Obviously the cooling system would have to change. Components would have to be moved to accomodate new ones. The PSU wouldn't need a larger heatsink, it could use a smaller one instead.

 

Turbulence has a big effect on cooling. There is a thin layer of static air next to a surface which is a very good insulator. Surface turbulence greatly improves cooling by reducing the thickness of this layer (which is why car radiators have serrated cooling fins).

What this means is that it is better to have a fan blow fast turbulent air directly onto each heatsink than to have a fan extracting air from the case. The advantage of a heat pipe is that you can have a single fan do this for a number of heatsinks.

Something sitting in still air (like the PSU) will be cooled a lot less efficiently than a fan cooled PSU even if the air is cooler.

 

I was trying to avoid talking about boundary layers and mixing due to turbulence.

 

I think my undergraduate and graduate classes in thermodynamics, fluid dynamics, heat transfer, computational fluid dynamics, etc. provide me with all the "intuition" I need in this area.

My point stands. Thanks for agreeing.

 

Well, if you're so well versed in heat transfer, then you should understand the effects of convective heat transfer coefficient. I used the same equations you learn in thermo and heat and mass to try to estimate how much the 360 would be heating up ambient air.

I maintain that we're at the point where warm moving air is better than cool non-moving air. Increasing h has a much greater effect than increasing theta in this case. If Q = h * theta it's easier to increase h, it grows with fluid velocity. The only way to increase theta is to make the chip hotter. I'd rather decrease theta and increase h because that's now a variable I can control in the system.

 

Are you sure you aren't confusing my posts with other similar ones in this thread on this topic?

While that may work fine for the PSU, and I think I covered that simply enough many posts ago, it still adds more heat into the system and increase ambient temp inside the case. Which requires a more efficient cooling solution for the CPU/GPU than what already exists. Which means larger heat sink or more powerful fan. Which means more money, and perhaps more noise. And it means a larger case. After all those which's, you get back to my initial, very simple point: after MS turned the engineering, manufacturing, marketing and economic cranks on each of the options they found it more desirable to put the PSU outside the case.

 

Yeah, I looked back at my original post to see where the confusion may have originated. I think I may have found it here:

This was worded poorly as I didn't really finish that thought. To minimize the effect of the PSU heat output on the existing CPU/GPU cooling solution, you would have to segregate the case, thus the PSU would indeed be in a natural convection environment, but one with relatively worse ventillation. This doesn't really accomplish much at all, which is why I continued:

(if a barrier were left in place)

(which is the obviously more likely scenario).