Too little surface area, same problems as direct die cooling, but a bit safer, but yet introducing two more thermal materials.
Printable View
Too little surface area, same problems as direct die cooling, but a bit safer, but yet introducing two more thermal materials.
Look at the storm, not that much SA but the impingement makes up for it, and in this case the drastic increase in flow. And how does it have the same problems when its sealed up (aka a "waterblock")?Quote:
Originally Posted by n00b 0f l337
Not trying to sound stubborn or anything, just wondering about the performance potential of some idea in this area. I might try this
Impingement creates turbulence, and the impingement effect, more flow just means your rushing a bunch more liquid across the top of a flat surface. Ever see those blocks with a tube soldered or brazed to a flat peice of copper? Your doing the same thing.
Flow will not make up for everything.
what about leak testing it? say with o-rings and such?
i wan't to make one so bad... nice guide!
impingement does work but is highly influenced by the size of the holes. For the same geometry (relative size of components) smaller holes will ALWAYS have higher performance. You want to get the smallest drill that is usable for the holes and think about mounting distance in terms of diameters of that hole. Beyond 15D its not jet impingement as the jet has slowed down.
the storm works well because it uses confined jet in hole impingement. When the first nozzle hits the base it creates and extremely fast and thin wall jet which when it hits the other wall creates a impingement jet effect all the way around the bottom rim of the hole. This gives you approximately 3x the performance of a standard jet for heat transfer.
I believe that jet impingement blocks will scale much better with flow than "normal" blocks for a higher flow rates.
I don’t believe that direct die cooling has been well done as a whole based on the methods I have seen. Also FC72 has lower cooling performance (grashoff number for example) than water which can make a huge difference to cooling power.
wall jet = a fast moving thin region of flow between a wall and a slower moving body of fluid. It has extremely thin velocity boundary layers due to its thickness. The thickness for a jet can be approximated using same flow area. Indeed the flow structure of a vanilla jet impingement can be derived directly from the navier stokes equations if your that way inclined.
nice guide, I really learned a lot
A question, what are the distances from the mounting holes for a socket 775 and the die center ??