[sdiy] Voltage regs + diode drops.

[Mike]

> At 12:07 PM 2/21/2009, Neil Johnson wrote:
>
> >My own opinion: avoid if possible.
>
> Just because some small number of situations are problematic, it should
> always be avoided?  Not logical, captain.  :-)
>
> Ian

Perhaps a better rule might be: if you have alow or  invariant load, then
it's OK. If the load is going to change significantly, then modify the
output in other ways.

A quick check of the ST micro datasheet for the 7805:
http://www.rapidonline.com/netalogue/specs/47-3290e.pdf
shows a range of ways to increase the output of a 7805 easily.

I'd be more worried about the 79xx series though, these seem to be less
stable and might get more upset by a non-linear device in their output line.

I once put a diode in the output of a 7805 to combine the output with that
of a 3.6V standby battery, (also with a diode). The simple test breadboard
worked fine. With 700mA worth of old fashioned TTL logic hooked up, things
got quite nasty, because I'd 'slugged' the very low output impedence of the
regulator with a decidedly non linear 1N4001.  The power rail was moving
quite visibly on the 'scope, and some LF ringing seemed to be occurring. The
answer was to make a shut down circuit and just power the part of the
circuit that needed backup.

For a small 'quick and dirty voltage drop, I've never had a problem.

||\/||ike