[DSP-10] Re: EZKit heat sinking

[boblark at proaxis.com]

Hi David,

Your pictures show a fine job of constructing the radio.  The heat sink
solution would really pull the heat from the challenged regulator.

A couple of asides. I tested my fingers once, and found that I could
continue to hold them on metal at 50C. At 55C it was still possible, but
starting to be painful!  At about 60C the protective reflex took over and
I pulled back.  Other fingerss may vary. The Fluke temperature probe shows
the unmodified regulator tab at 62C, running from 10.0 VDC (the US
supplied plug in supply). Running as you do with 9.0 coming in it will be
somewhat cooler.  Also the fingers act as a heat sink for something as
small as the tab!

The second aside, is the arrangement of parts that you show looks good.
Some other people have used that area near the L32 LP filter for wires and
connectors. This may be problematic, in that the pickup of noise or
spurious signals by L32 and other circuits in that area is possible. If
anyone is having such probles, they might try to move the wires near the
case and away from the board. The signals in the L32 area are very weak!

Thanks for sharing the info.

73,
Bob  W7PUA
> Hi Bob and All,
>
> Just thought you might like to see a couple of pics showing how I reduced
> the temp of the internal regulators in the EZKit enclosure.  They had felt
> a
> bit hot to touch (45deg?)  I ususally try to keep heat down as it tends to
> reduce reliability.
>
> I fitted a 1/2" by 1/4" ali bar between one regulator and the outer case
> A further piece of ali bar was drilled and tapped to get the heat from the
> other regulator to the top of the EZKit box, and the box lid is now 1/16"
> ali sheet instead of  FR4 material to further help conduction.
>
> Overall it just gets pleasantly warm to the touch rather than 45 deg as
> previously
>
> Another thought:
> So I can change over from internal 10MHz to external frequency reference I
> am going to fit a small 5V relay by the connector - this will be operated
> by
> a rear panel switch.  I had thought of fitting a circuit to detect the
> incoming 10MHz and switch automatically but perhaps that would be getting
> too complex?
>
>
> Pics at http://www.microwave.fsnet.co.uk/projects/projects-15.htm
>
>
> 73 David G6GXK
>
>