Wouter van Ooijen wrote:
> I think a counter-argument could be that the loss in the converter is at
> least partially 'constant-voltage', which would hit a low input voltage
> harder than a higher input voltage.

I don't really agree with this.  A single cell presumably has a lower
voltage than the desired output, so I'm assuming a boost converter.  For a
boost converter, the fixed voltage inefficiency is on the output side.  The
input sees a inductor in series with the switch, and nothing else.  The
inefficiencies here are the resistance of the inductor and the voltage drop
accross the switch.  At low voltages suitable FET switches are cheaply and
readily available.  These act like resistor, not voltage sources like
bipolar transistors do.  Values below 10mOhm are quite possible.

This means essentially all inefficiencies in the input part of a low voltage
boost converter are resistive losses, not fixed voltage losses.  This in
turn means boost converters can scale to low input voltages like those of
common single battery cells.

The trick with a single cell power supply would be to generate the 5V or so
for the control circuitry without being a significant drain on the battery
at low output power.  I'd probably use a pure analog circuit to create a low
power boost supply to run the control PIC, then have the control PIC manage
the exported power output.

All that being said though, I'm not convinced that a single cell power
supply is worth all the trouble to get around the multiple cell issues.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist