I was looking through the fabfin catalog from r theta (great place to learn about heatsink properties and calculations btw!) and started getting some ideas.
these look kind of interesting except that I have no idea how much they cost. looking at an mft245b(10"x5"), equipped with a panasonic fbaF12G12L1a (120mm, 69cfm, 30dbA) running at minimum speed (less airflow and noise) would produce 5-7 m/s of total airflow before pressure derating. effectively half of that flowing across the fins(i think?). giving the sink an effective <.08 C/W sink to ambient rating. (there are some assumptions here with regards to the free area, actual output of the fan at reduced speed, and pressure differential...i used 50cfm as fan output, 3000-4000mm^2 for CSA, and 0mmHg) a panaflo H1A fan may need to be substituted for its better pressure specs.
That sounds good for about 12 aleph output devices at 25 watts each with a junction to sink of 1*C/W the case will be 25* above the sink or 50* above ambient. Or just connect them directly to the sink with some silver compound.
which brings me to next set of compromises. the devices would have to be attached via a clamp since it would be pretty difficult to get bolt and a nut worked in between the fins. Might be unreliable. The fan would have to blow into the sink unless its made out of a metal and has bearings made out of some space age polymer that wont melt or expand. Also anything directly in front of the fan if it were in suck mode would be likely to melt too. 300W is not a little heat...
The things i am concerned about right now are: What is this source area? My guess is that its either the optimal, or required mounting area. It would be pretty difficult to fit more than one transistor into that area. I am guessing that all the calculations above would only be valid in that area. How bad would they be out of that area? I can imagine the edges being pretty hot. Assuming the edges aren't hot enough to hurt the mosfets, how would sound quality be affected if the fets on the edges are 10* hotter than the ones in the middle? 2nd concern is what if that fan goes! costly experience for sure. I wonder how much noise the air moving over the serrated fins would make on its own?Other concerns are of course cost and availability. Any ideas on that? A quick google search didn't turn up anything, and i've never even seen a bonded fin sink anywhere besides cpu heatsinks. They weren't cheap either...
If i end up doing this, the heat sink assembly would be enclosed in an insulated, sound dampened section of the chassis to further reduce noise. I would rig up some sort of LED that would let me know if the fan is on since hopefully it will be difficult to tell with a quiet fan and damping.
I get so carried away with these things...
jt
these look kind of interesting except that I have no idea how much they cost. looking at an mft245b(10"x5"), equipped with a panasonic fbaF12G12L1a (120mm, 69cfm, 30dbA) running at minimum speed (less airflow and noise) would produce 5-7 m/s of total airflow before pressure derating. effectively half of that flowing across the fins(i think?). giving the sink an effective <.08 C/W sink to ambient rating. (there are some assumptions here with regards to the free area, actual output of the fan at reduced speed, and pressure differential...i used 50cfm as fan output, 3000-4000mm^2 for CSA, and 0mmHg) a panaflo H1A fan may need to be substituted for its better pressure specs.
That sounds good for about 12 aleph output devices at 25 watts each with a junction to sink of 1*C/W the case will be 25* above the sink or 50* above ambient. Or just connect them directly to the sink with some silver compound.
which brings me to next set of compromises. the devices would have to be attached via a clamp since it would be pretty difficult to get bolt and a nut worked in between the fins. Might be unreliable. The fan would have to blow into the sink unless its made out of a metal and has bearings made out of some space age polymer that wont melt or expand. Also anything directly in front of the fan if it were in suck mode would be likely to melt too. 300W is not a little heat...
The things i am concerned about right now are: What is this source area? My guess is that its either the optimal, or required mounting area. It would be pretty difficult to fit more than one transistor into that area. I am guessing that all the calculations above would only be valid in that area. How bad would they be out of that area? I can imagine the edges being pretty hot. Assuming the edges aren't hot enough to hurt the mosfets, how would sound quality be affected if the fets on the edges are 10* hotter than the ones in the middle? 2nd concern is what if that fan goes! costly experience for sure. I wonder how much noise the air moving over the serrated fins would make on its own?Other concerns are of course cost and availability. Any ideas on that? A quick google search didn't turn up anything, and i've never even seen a bonded fin sink anywhere besides cpu heatsinks. They weren't cheap either...
If i end up doing this, the heat sink assembly would be enclosed in an insulated, sound dampened section of the chassis to further reduce noise. I would rig up some sort of LED that would let me know if the fan is on since hopefully it will be difficult to tell with a quiet fan and damping.
I get so carried away with these things...
jt
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.