At 11:25 AM 11/24/99 -0500, Bob Landman wrote:
>One of the problems with effective cooling of several PC boards is
>that if they are stacked close to one another they radiate heat to one
>another and convection is made difficult.  Now if ALL boards had some
>holes in them to allow air to flow through them then convection
>(whether forced or natural) would, it seems to me, to be improved.
>Now, for a significiant amount of air to move THROUGH a board, the
>holes have to be fairly large and in that case, surface area of the
>radiators are being sacrificed to air flow.  There must be a trade-off
>between the two and where that is, I have not a clue.

I'm going to keep pressing this point: adding holes, unless the holes are
*too* large, *increases* surface area, provided that the holes are
plated-through. In the example I gave, a pattern of 50 mil holes on 100 mil
centers decreased the top surface area by 20 percent. In other words, the
holes become 20 percent of the board. I'd think that would allow some air
flow. But the hole walls added 100 percent, and the holes allow the other
side to become a heat sink as well, so the net increase is 160 percent;
that's more than double the surface area.

What the optimum point is I don't know. I wouldn't want much less board
material left between holes. If the pattern were offset or if the holes
were in a hexagonal pattern, one might make the holes a little larger; I
haven't studied the math....

>But, if you
>have to stack boards horizontally then you need a way for air to flow
>THROUGH the boards (I'm not advocating that - a bad idea but my Compaq
>is laid out this way as are most tower PCs and the "forced" air
>cooling is minimal as air is drawn in through the power supply and
>this hot air then pressurizes the case blowing out through the disk
>drive and a series of holes punched at the top rear of the case and
>the lower portion of the sides of the case.  How enough air gets
>around to the boards to do any good is a good question.

If the board surface is hot air will rise from it. If there are holes in
that surface, air will be drawn from the bottom of the board. I'd expect it
to work pretty well....

I remember saving a company I worked for a sum greater than my yearly
salary by noting that a heat sink planned for a board was not necessary. I
was not using holes in the board; but heat sinks are expensive compared to
holes in a board; a hundred holes would be what? ten cents?

Who is going to do the research to give solid numbers to this technique? I
would not expect the heat sink companies to do it!

[log in to unmask]
Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433