battery driven stuff

[Martin Czech]

>From time to time one needs battery supply, e.g. if you want to play
your little synth at the beach... (WASP), or if you need a little mic
preamp for outdoor recording etc..  Usually severall NiCd or other
elements are cascaded to get a reasonable voltage.

The problem is that if one element is weaker then the others the voltage
will collapse, in the beginning everything looks ok, but after a few
minutes it's over even if there is enough charge in the other elements.

Another thing is voltage regulation, I have no problems to waste several
W for a linear (and thus quiet) regulator, but for battery stuff power
consumption is an issue.


Therefore I'd suggest the following:

Instead of serial cascade, just use the elements in a parallel way.
The output voltage of NiCd is stable over discharge, so I hope there
will be no large currents as the elemets will fight against each other.

I don't know much about batteries and such, but I'm getting better. Could this 
parallelism be a problem for element lifetime or performance?

If an element gets weak in serial connection, there will be also
currents through the element chain (if the circuit still works).
This will also stress the weak element further.



Ok, if this works we have a rather stable 1.2V supply that should work
until the last element has lost it's charge. In order to get to some usefull
voltage I would suggest a step-up switcher. Yes I know, noise and problems.

Do you think that a 1.2V to 5V or even 1.2V to 20V ratio is possible
with good performance?

I guess the usuall switcher ICs won't work with such a low voltage (1.2V),
so at least the oscillator must be homebrew, the regulator can then live
from the higher output voltage.


Any battery gurus out there?
What's the proper name for rechargeable (NiCd) elements and
how do you call non rechargeable elements (MgZn and the like)?


m.c.