Barry Gershenfeld wrote:

>>Not the most elegant or cheapest solution, but you might also
>>consider small-V 5W zeners instead of a series R. I have one
>>item where I do this, but is Vz = 4.3v @ .3A, so only about
>>1.5W.
>>
>
>No matter how you do it, if you aren't switching then you are
>dissipating it.  And the combination of resistors or zeners
>or pass transistors all add up to that 7 watts of heat and
>you will have to get rid of it somehow.  And heat sinks
>aren't free, either.
>
>And a mobile unit that's too hot to touch is rather embarrassing
>(don't ask how I know this:)
>

Yes, my scheme was simply a way to distribute the heat,
so the 7805 wasn't dissipating all of it. In my system,
[15-5v]*.3A = 3W in 7805 w/o distribution. Removing 1W+
to the zener helped cool everything down "enough" to apply
Roman's test --> take a baby's finger and hold it on the
hottest part of the ckt for 30 seconds; measure sound output
at 3 feet using Radio Shack sound meter; if >=105 dB, turn off
power. [jus kidding about the baby, folks].

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.