[Lowfer] Homemade IC chip heatsink

[C. Turner]

For DIP-style IC's, it's usually enough to simply attach a chunk of 
scrap aluminum from, say, metal front of a Radio Shack project box - one 
of those in which you get both a plastic and metal front - but you use 
the plastic cover.  (Very thin aluminum such as that from a soda can is 
really too thin to conduct much heat.)

Bending them into a flat-bottomed "U" shape - the bottom being about the 
size of the flat top of the IC itself with the sides extending 1/2"-1" - 
and then using a metal-filled epoxy such as "JB-Weld" - which has quite 
low thermal resistance - works nicely and can dissipate about as much as 
the epoxy case itself will conduct anyway.  Slightly roughing the bottom 
side of the aluminum "U" and the top of the IC with fine-grit sandpaper 
just before attaching helps to make a durable bond as well.

In other projects I've found that the JB-Weld stuff does a respectable 
job, having used it to transfer 10's of watts of heat from surface-mount 
semiconductor devices to aluminum and copper heat sinks.  The key is - 
as with all heat sinks - is to have as flat and perfect a mechanical 
mating contact as possible before gluing so that the epoxy layer is as 
thin as possible.  After all, the epoxy isn't what's supposed to be 
transferring the heat, but rather keeping the heat sink in contact and 
filling the (few and small as possible) voids that inevitably result.

73,

Clint
KA7OEI
MedFER CT


Andy - KU4XR wrote:
> Nothing that's going to revolutionize the electronics world.
> The TDA7052B audio chip that I have been using as the PA stage
> for my current transmitter is a potent little chip. It has
> built in Over Temperature protection circuits, and Over current
> protection.. It pretty much takes care of itself, and will shut
> down if you try to push it past its internally set parameters.
> It is rated for 1 watt output, and at that level; it will get
> fairly warm to the touch, and start to back down its output.
> I did a search for IC chip heatsinks, and well, they are
> available to purchase, but way too expensive.. I saw that all
> they were was a " wrap around the chip " design, some were
> aluminum, some were a gold anodized plating.. Aluminum is a
> very common material for a heatsink, so here is the experiment
> that I tried, with a good success.. I still have lots of the
> Aluminum foil tape for HVAC use, so I cut a short piece from the
> roll to start my experiment. I cut a very, very, narrow strip
> of the tape, and wrapped it lengthwise around the chip, making
> sure that the tape did not touch any pins. I cut it long enough
> so that it would overlap itself, and I took the excess piece and
> folded it to itself, and then stuck it to the overlapped piece.
> A bad description I know, but the homemade aluminum foil tape
> heatsink has done a very good job a transfering the chips heat
> into the air. Observations after the heatsink: The power output
> from the chip running at 1 watt now stays steady with no
> fluctuation. Before; it would go up and down, especially when the
> heat, or air came on, but now it makes no difference. It surprised
> me what that little piece of aluminum foil did. Further observations,
> To keep it steady before the homemade heatsink; I had to blow air
> across it with a small muffin fan... Not now; - however, with the
> fan, I can get a sustained, steady output over 2 watts from the
> chip. It will go over the 200 Milliamp range of the meter, and
> stay there. Distorted as heck I'm sure, but the heatsink, and fan
> allowed the amp to generate more power, and hold it.
> I hope someone has found this to be at the least interesting reading.
>
> 73 to all:
>
>
>
> Andy - KU4XR - EM75xr - Friendsville, TN. USA
>
>