A closed case will need very powerful fans, drawing a lot of power, decreasing global efficiency.
not necessarily pallas ...
just a tunnel - or vacuum which automatically forces air through ...
it is beneficial for a powerful fan to be placed at the beginning - but not necessarily fundamental for the operation ...
a larger slower moving and less power consuming fan ion the back will suffice to keep the airflow moving at high rates ...
#crysx
This is untrue. Cases are leaky and fans are inefficient as cases aren't designed as a sealed highflow system. You're better off with positive airflow then negative as it will leak throw the cracks in the case. You need more powerful fans to have the same effect in the back of the system as the front.
But yeah, rackmount systems generally have highflow fans on the front of them which are extremely noisy to push A/C air through them in data centers, which is why they're designed the way that they are. That doesn't work for home use.
not entirely bensam ...
we have 6 of them in the rack at the moment and ALL of them are highly efficient as airflow systems - BUT with powerful fans ... ones in the back not as powerful as the front ... they are all chenbra and supermicro chassis and servers ... all of them are 'standard' servers not high heat miner systems ... yet
![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
...
but for high heat dissipation systems like our miners - i must agree ... which is THE reason our frames that we are currently designing have a vacuum style of airlow in them - and it seems to be working ...
no leakage - now spaces - just a complete sealed system that forces airflow through two fans ... one powerful highflow fan in the front - one large lower powered one ( yet same airflow ) in the back ...
once this is perfected ( for our farm - cant speak for anyone else here ) - then we will move onto the funneling of the heat rich exhaust out of the room ...
though there isnt much an open air system cant do in comparison to a sealed one ... most of the farm is open air - and very fiddly to deal with ...
IF - and that is a big IF - we can get the design of the closed system close to a perfect airflow - then all should be well ... if not - then back to the open air system frames ...
either way - we are just testing when there is time ... i am very time poor at the moment due to personal issues that require my attention - but work on it every chance i get ...
im actually just wondering about the vertical style system of venting airflow also - though that also will require a complete design change ... and soon - the help that is here now will soon be gone - which will make it more a 'me' and 'i' rather than a 'we' ...
i see there are some major improvements to the ccminer-spmod fork also ( with some help from pallas and others ) - amongst other forks like epsylon3 and djm34 and klaust ... seems i need to get my ass moving to publish the donation servers and have them active ... they are working and running ( bar two sevrers which are being redone this week ) so some public use / testing would be great ...
ill update them here -
https://bitcointalk.org/index.php?topic=1089744 ... but will also do a quick notify here in this thread also when they are finished - for those interested in mining anonymously for donations for the aforementioned devs and others ...
#crysx