#6 screw as heatsink

[Rich Grise]

If something is working for you go for it. I just wanted to keep it
clear that there are choices and alternatives and tradeoffs.

Speaking of aluminum heatsinks, is there a measurable difference
in thermal resistance (HS-to-ambient) between plain aluminum and
black-anodized aluminum, all other things being equal?

Thanks,
Rich

 

[John Devereux]

Speaking of aluminum heatsinks, is there a measurable difference
in thermal resistance (HS-to-ambient) between plain aluminum and
black-anodized aluminum, all other things being equal?

What is the thermal resistance of solder resist?

 

[MooseFET]

Don't skimp on the lockwasher. The FR-4 will cold-flow in these
situations.

I think the metal does it more so but yes always-always use a
lockwasher. If it gets even a little loose, vibration will make it
very loose quickly.

 

[MooseFET]

On Mon, 10 Mar 2008 06:56:22 -0700 (PDT), MooseFET [ascii art snipped]
I have used this basic design quite a few times. The lower weight of
the aluminum, in this case, I think dominates the vibration question.
If something is working for you go for it. I just wanted to keep it
clear that there are choices and alternatives and tradeoffs.

Speaking of aluminum heatsinks, is there a measurable difference
in thermal resistance (HS-to-ambient) between plain aluminum and
black-anodized aluminum, all other things being equal?

What is the thermal resistance of solder resist?

IIRC about 10x better than air.

 

[John Popelish]

Speaking of aluminum heatsinks, is there a measurable difference
in thermal resistance (HS-to-ambient) between plain aluminum and
black-anodized aluminum, all other things being equal?

There is a radiant component of heat sink performance, but
it goes up at the 4th power of absolute temperature
difference between the sink and its visible surroundings,
and in proportion to its emissivity. So an emissivity near
zero (shiny aluminum) pretty well eliminates this component.
But even if the emissivity approaches 1, it still takes a
quite hot sink surface to make this effect very significant.

So if you intend to operate the sink in still air at
something like 125 C in a 25 C environment, you might get
your money's worth out of the black dye job. If you have
forced convection or can spend that few percent on a
slightly larger heat sink, you will probably do better with
that.

Also, if the sink is made of parallel fins, almost all that
radiated energy is just passed back and forth between
adjacent fins. So it only helps on parts of the sink that
"see" the cooler environment.

 

[John Popelish]

What is the thermal resistance of solder resist?

What is the thermal resistance of the air boundary layer
attached to the resist or a metal surface?

 

[John Devereux]

What is the thermal resistance of the air boundary layer attached to
the resist or a metal surface?

Actually I was wondering in the context of a metal block clamped to a
groundplane. I have an application with lots of thermal vias coupling
heat from one side of a PCB. Then there is a metal block coupling the
heat to, effectively, a heatsink.


Heatsink

//////////////////////////////////////
======================================
| Metal block |
| |
====][====][====][====][====][====][== PCB with thermal vias

^^^^^^^^^^^
Heat Source

As an experiment I flow soldered the PCB so as to fill the vias. It
worked nicely, except a couple of vias have a bit of solder leaking
through to the top which creats a slight "bump". Very small but I know
it would not take much irregularity to destroy the thermal conduction
to the metal block.

So I was going to try putting solder resist on the "metal block" side
in the next revision.

(I am mainly doing this for interest since I am confident the
conductivity is enough without filling the vias at all).

 

[Rich Grise]

There is a radiant component of heat sink performance, but
it goes up at the 4th power of absolute temperature
difference between the sink and its visible surroundings,
and in proportion to its emissivity. So an emissivity near
zero (shiny aluminum) pretty well eliminates this component.
But even if the emissivity approaches 1, it still takes a
quite hot sink surface to make this effect very significant.

So if you intend to operate the sink in still air at
something like 125 C in a 25 C environment, you might get
your money's worth out of the black dye job. If you have
forced convection or can spend that few percent on a
slightly larger heat sink, you will probably do better with
that.

Also, if the sink is made of parallel fins, almost all that
radiated energy is just passed back and forth between
adjacent fins. So it only helps on parts of the sink that
"see" the cooler environment.

Thanks!
Rich

 

[Don Klipstein]

There is a radiant component of heat sink performance, but
it goes up at the 4th power of absolute temperature
difference between the sink and its visible surroundings,
and in proportion to its emissivity. So an emissivity near
zero (shiny aluminum) pretty well eliminates this component.
But even if the emissivity approaches 1, it still takes a
quite hot sink surface to make this effect very significant.

It is not 4th power of the difference, but the difference between the
4th powers of the two temperatures.

So if you intend to operate the sink in still air at
something like 125 C in a 25 C environment, you might get
your money's worth out of the black dye job. If you have
forced convection or can spend that few percent on a
slightly larger heat sink, you will probably do better with
that.

Keep in mind that the black dye is not the whole story - a good part is
the oxide layer formed in anodizing. The radiation will mostly be at
wavelengths 5 to 20 micrometers, and what is black or white or
transparent to visible light is not necessarily so to these longer
wavelengths. However, bare metal is mostly bright and shiny to these
longer wavelengths.

Also, if the sink is made of parallel fins, almost all that
radiated energy is just passed back and forth between
adjacent fins. So it only helps on parts of the sink that
"see" the cooler environment.

John Popelish

- Don Klipstein ([email protected])