Heat Sink Rating

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
P

plasnu

Member
Joined 2011
I'm trying to find heat sinks for Aleph J. Based on information on this forum, NP recommends 25 to 30 degree, and Aleph J dissipates roughly 100W par ch, so 0.25-0.30 c/w par ch is required for Aleph J. The heat sink of the 4U chassis sold at DIYAUDIO Store is rated 0.31 c/w at the size of 300x165x40, and I'm looking for the equivalent heat sinks for my own custom chassis.


I checked Aavid, and and the equivalent heat sink seems to be

Thermal Engineering and Heat Sink Manufacturing - Online Store

Although 80345 is made with 6000 Aluminum (best material) and it has about the same number of fins as 4U chassis, it has only 0.53 C/W at the length of 165mm (4U chassis), which is apparently inadequate for my purpose. Based on Aavid calculator, 500mm length is required for 0.3 C/W (3 times longer than 4U heat sink)


I'm wondering if I misunderstand something here...
 
Last edited:
E

evanc

Member
Joined 2009
Putting theory aside. I built an f5 using 2 7" (about 178mm) pieces of the heatsink usa 10.08" extrusion and it was OK.
 

Attachments

  • DSC01404.JPG
    DSC01404.JPG
    142.9 KB · Views: 539
  • DSC01409.JPG
    DSC01409.JPG
    134.8 KB · Views: 529
P

plasnu

Member
Joined 2011
I measured my PASS LAB Aleph 5 (not DIY) dimensions.

A5 dissipates probably 250-300W, so the rating of A5 heat sink should be 0.4-0.5 c/w for each side (280w x 250h x 50d). The Aavid heat sink closest to A5's is 62185, which c/w is 0.65 for 250x250x50. 62185 is about 10% narrower than A5, and considering the distributed load of A5's FET (10% better performance expected, AAvid says), Aleph 5's heat sink real world performance should be around 0.52 c/w based on Aavid calculation method, which seems to be reasonable.

http://www.aavid.com/products/extrusion-heatsinks/62185

—————

Now I feel Aavid information is pretty reliable. then I recalculate Diyaudio chassis with the Aavid method.

I realized that the DIYAUDIO 4U heat sink has thicker bass plate than Aavid 80345 that I mentioned on my first post, so this time, I chose 70165 which has slightly more density and slightly longer fins than 4U heat sink. Considering the distributed loads of FET (10% better), I guess 4U heat sink should perform about 0.38 C/W in real world based on Aavid calculation, which is acceptable for AJ, although not perfect. I think DIYAUDIO heat sink calculated by different type of method. (I’m not criticizing about it, I’m just looking for universal calculation method that I can use to find adequate heat sink for AJ. Also, I feel that Aavid method is somewhat pessimistic, DIY is slightly optimistic, so the real world performance would be somewhere between them.)

http://www.aavid.com/products/extrusion-heatsinks/70165


In the other thread, someone was saying 0.3 c/w on paper is more like 0.4 c/w, so Aavid calculation matches to his comment as well.
 
Last edited:
P

plasnu

Member
Joined 2011
10.080" profiles 6"long should be 0.584c/w according to AAVID, and 65W*0.584 = 37c above ambient is their estimate. They would say 10-12 inch is the correct size for 65W. They are just trying to sell more heat sink than what we actually need?, I don't know...
 
A

AudioSan

Member
Joined 2009
cut out from FAQ at heatsinkUSA:
"Question: What are the °C/W ratings for your heatsinks?

HeatsinkUSA does not offer specific advice regarding the thermal properties of our heatsinks. However, we have supplied Approximate °C/W ratings for each profile for your reference."
in other words. C/W ratings from heatsinkUSA is nothing but guess work and not an exact science.
 
P

plasnu

Member
Joined 2011
hmm,

I was looking at heatsinkUSA site yesterday, and realized that the rating of 10.080" seems to be somewhat less efficient compare to the other heat sink they sell. I compared their heat sinks with AAVID equivalent ones. All of USA heat sink are made with 6063-T5 except 10.080" which is made with unknown material, so I was wondering if 10.080" is made with some less desirable material, since I read making higher fins with 6063 is somewhat more costly and more difficult somewhere...
 
Last edited:
P

plasnu

Member
Joined 2011
I think you misunderstand, sorry for my English.

I mean 2x 4.600" are connected horizontally (9.2 width), 0.7 c/w/3 , cheaper, smaller, more efficient than 1x 10.080". Only small difference, though. I think this is due to serrated surface of the fins (like DIYAUDIo 5U fins), and maybe materials difference.
 
Last edited:
rayma

rayma

Moderator
Joined 2011
dantwomey said:
I believe I fondly remember a post where Papa suggests that you calculate what you need and then double it!
Click to expand...

Bear in mind that heat sink ratings are based on UNIFORM thermal loading over the entire interface surface, with the fins vertical and unimpeded.
This is seldom the case in actual use. The effective thermal rating will have to be decreased significantly when just a few, relatively small devices
are mounted on the sink, since it is not being fully used.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom