David Duffy (AVD) wrote:
> For one of the projects we have on the go at present, I'm using a
> 78L05 regulator to drop the 24V incoming supply down to 5V for the
> PIC and MAX1485. The board's relay is a low power 24V type, so only
> draws about
> 8.5mA when energised.
>
> With the 78L05, the idle current draw from the 24V supply was about
> 4.7mA. Swapping it out for an LM2931-5 regulator dropped that to about
> 1.4mA, quite an improvement as the relay is off most of the time.
>
> That's good enough for this application, but it got me thinking about
> the few wasted mA * 19V (24V - 5V) in the linear regulator. This
> design is through hole, but had to be compact so I didn't even try to
> fit a switching regulator in.

How much current does the 5V supply need when nothing is happening?  A
linear regulator is quite inefficient when dropping 24V down to 5V, but it
can be really good at quiescient current.  If most of the time is spent
doing nothing, a linear regulator may be the better answer.

> Are there any small, low-current, hi-efficiency switching regulators
> out there that are still reasonably priced? I realise that a 78L05 is
> hard to beat price wise, but are there good alternatives out there?

This is probably not appropriate for your case, but I'll mention it because
others might find it useful.  I sometimes use a 10F204 as the controller for
a switching power supply.  All the parts are dirt cheap except the inductor,
but that would be required either way.  Nice switching power supply chips
are surprisingly expensive.  It's also often useful to treat startup
differently than main operation and to be able to limit duty cycle.  This
essentially makes the power supply current limiting without any direct
current sensing.  It can usually be arranged to survive a short without
damage or excessive current draw all the while guaranteeing not to saturate
the inductor.  This and intelligent startup are not easy to get from cheap
PWM chips with analog feedback.

A neat trick I've used a bunch of times now is to put a PNP transistor
around a LDO linear post-regulator to detect the switcher output regulation
threshold.  The B-E junction drop happens to be just the right target
voltage for the LDO input above its output.  The MCP1700 LDO is a nice part
for this since it's pretty cheap, works with 0 ESR caps, and the B-E
junction drop above its output is solidly above its dropout voltage but
below its maximum input of 6V.

Once example of this is on page 1 of http://www.embedinc.com/ioext/io2.pdf.
In this case the 5.6V output has to be cabable of 3/4 amp, so a few mA
quiescient current was no issue.  Having the 5.6V supply availabe is useful
since you can give various subsections their own LDO, and other parts just a
diode if the 5V supply doesn't need to be too tight.  It also makes it easy
to isolate supply segments so that one driven externally doesn't drive the
others.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist