heatsinks for aleph 2.

[BrianGT]

I picked up some aluminum heatsinks today from Austin Electronics, a electronics surplus store north of Atlanta, GA today. The dimensions are:

8" x 9" x 2.6"
baseplate - 1/4" thick
19 fins - 8" long x 1/8" thick each

Will these work good for the Aleph 2? Would one work for a channel? or should I use 2, with 6 mosfets on each?

Here are some pictures. I sized them down quite a bit. Hopefully they are not too big.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

If you are in the Atlanta Area, if you call them, they can get more in. They cost $24 each. It is north of the city on 85, off Jimmy Carter. Here is their contact information:

AUSTIN ELECTRONICS
6050 CENTER PL NW, NORCROSS, GA 30093, (770) 449-8697

--
Brian
gte619j@prism.gatech.edu

 

[cp642]

There's no such thing as too much heatsink for a class A amp like Aleph. I would be using 4 pieces of those heatsinks per ch. if I were you.....

 

[BrianGT]

4 of these per channel, with 3 mosfets per heatsink? Isn't that overkill?

I was figuring that I would use 2 heatsinks per channel, with 6 mosfets on each. Does this sound like it would work?

They also have heatsinks that are about 12x14x1.5 I figured that they would not work as well.

Anyone have any suggestions on heatsinks, with regard to what they used for their designs?

--
Brian
gte619j@prism.gatech.edu

 

[BrianGT]

There seems to be a problem with the pictures coming up.

I put a mirror of the pictures up at:

http://brian.prohosting.com/heatsinks/

--
Brian
gte619j@prism.gatech.edu

 

[cp642]

Overkill? I'm not so optimistic.....try comparing your heatsink dimensions with the dimensions of Aleph 2.

 

[13DoW]

Brian,

I reckon each of your sinks is ~0.5 C/W.
Comparing the profile to the Wakefield extrusion catalogue the nearest match was 0.75C/W (3" length) but yours have slightly deeper fins.
So, conservatively, yours will at least as good and then adding the length factor of ~ 0.7x for an 8" lenght cf. a 3" length gives ~0.5C/W overall.

Now you can figure out how many you want. I don't know the aleph 2 rating but 4 heatsinks for 200W will give you a 25C rise. On top of that you must consider the device junction to heatsink rise (spreading the power between more devices helps).

Regards
13 Duke of Wymbourne

 

[BrianGT]

The aleph 2 was 12" x 12" x 10.5" high

An externally hosted image should be here but it was not working when we last tested it.

I guess that 4 heatsinks would be the best bet. I didn't realize that one channel putting out 100W would take so much.

So, 4 heatsinks with 3 mosfets spaced out pretty good on each heasink will work?

--
Brian
gte619j@prism.gatech.edu

 

[13DoW]

Brian,

just checked Aleph2 and the standing dissipation is 300W per channel (for 100W output). With 12x output devices that's 25W each. The junction to sink thermal resistance will be at least 1 C/W (depends on insulating pad)- let's assume 1.5C/W then each device junction will be 38C hotter than the sink.
If the total sinking (4 pieces) is ~0.125C/W then the sink will be 38C hotter than ambient. So, for an ambient temperature of 20C the device junctions will be at ~96C.
96C is acceptable but lower is always better for both longevity and performance (MOSFET gain decreases as temperature rises).

Regards
13DoW

 

[promitheus]

96 is theoreticaly good. I think its too hot for the fets.
If the themal resistance of the heatsinks is only 0.5K/W than I agree that 4 heatsinks is better but you could use 2 per channel on lower power. You can use pads with lower than 0.1 K/W thermal resistance.

 

[BrianGT]

What kind of pads are best for this application? If I use good pads, and 4 heatsinks (per channel), will it run reliably as an aleph2 at full power?

--
Brian
gte619j@prism.gatech.edu

 

[grataku]

Yep, definitely 4 per channel.

 

[13DoW]

Promethius,

please post some details on the best insulator pads. My quick search revealed a value of junction to sink for a to-247 package with a Kapton pad of 1C/W. Other options were higher.
The junction to case for the package is 0.3C/W.

Thanks
13DoW

 

[Nelson Pass]

Pass sez: 4