On Thu, 16 Aug 2001, David VanHorn wrote:

> >
> >The device will be connected to anything from 10 to 26 VDC with a power
> >supply requirement of 20 mA @ 5 VDC.  All of the hideous automotive-type
> >abuse will be present; e.g. reversed battery, load dump, battery disconnect
> >and voltage peaks.  In addition, one of our customers has specified a 2
> >minute, 80 volt survival test.
>
> Your dissipation, at 26V with a linear reg will be 0.42W, no matter how you
> do it.
> At 80V, 1.5W
>
> A zener supply might work, if the rest of the components can handle the
> 3W/min of heat during the test, and live with the .42W continuous.   This
> will depend on the total thermal mass, since you can dump heat into the
> potting compound in the short term. Make sure that the resistor dosen't get
> hot enough to decompose the potting compound though! :)
>
  You can't use one resistor from 10V to 80V input voltage variation !
You need a high voltage current source ( one pnp + one npn which will
dissipate indeed about 1.5W, this sort of current source are used to
supply from 24 to 240V ac or dc any low current boards ) You don't need
any linear reg. A zenner is enough. But you'll have some thermal stress
there...
Vasile

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