> How about MOSFETs on the output of the regulator to disconnect it 
> from the battery when no input voltage is present?

Would have to be connected 'backwards" to stop reverse flow via body 
diode when off. so when power present the off FEt would be the body 
diode and turning it on wuld make it somewhat better. As the blocking 
diode is not a major perfomance issue this is not likely to be cost 
effective compared toma diode.

\> How about the old LM723, cheap with built in current limiting, 
through will need an external pass transistor for the kind of currents 
you require.

LM723 is still an extremely useful device. My first ever voltage 
regulator was a uA723 in a metal TO5 (?TO39?) can.

Add a cheap pass transistor and not much else and it would handle this 
task and wuth a little care you could use foldback to achieve both 
boost charging and float.

About $US0.30/1000 for the regulator which is good for 150 mA and 
about 7V differential by itself. It's surprising how long sucha  dated 
design can still compete. The ootb reference accuracy is only +/-5% so 
calibration would be neweded but long term stability looks easily good 
enough.

For those who haven't met it the LM723 can implement most forms of 
regulators including switching, shunt and series and even floating 
negative and positive topologies which exceed the ICs maximum voltage 
rating. It does this by providing "building block" access to its 
internals - it's the "555" of the early regulatr regulator world. I 
had my first introduction to switching regulators in a uA723 app note 
and they probably still publish the same app note today - but with a 
newer date on it.


        Russell McMahon

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist