colin wrote:
"Richard H." wrote in message
...
To step-up a single-cell 1.5v battery to ~3.3v, is there a better option
than using a boost regulator like an LM2623?
(http://www.national.com/ds.cgi/LM/LM2623.pdf)
I'm trying to hack out power options for a simple battery-operated LED
gizmo, and it's turning out more complex than the gadget itself.
BTW, load is 1mA to ~100mA depending on LED duty cycle, with a lot of
"off" time. (The power needs to stay on, but the microcontroller will be
quiescent - I'm not sure that a simple transistor+inductor scheme will
give much battery life.)
depends how well it is designed,
if it reduces the frequency of operation at idle it can
increase the efficiency becuase it lowers switching losses.
a hysteretic type would be ideal for this.
some others go into whats called burst mode.
you would need a very low power controller,
ive not had to look for such before.
maybe you could use the led pwm if its high enough frequency.
an alternative is the capacitor charge pump.
100ma is quite a bit, there are many wich are fairly low power,
not sure about 100ma.
depends what idle power consumption you can tolerate.
if its only a few microamps its difficult.
I used one of those micropwer op amps in a rf sniffer probe,
didnt have space for an on/off switch as well as a battery.
but got it down to 5 yrs life with no switch.
Colin =^.^=
Hmmm. The application here is something along the lines of Activate -
blink LEDs for a while - shutoff until re-activated. Activation is
likely to be manual (maybe a pushbutton or capacitive sensor) but maybe
also periodic timer (which could be another microcontroller in the power
module).
So, it's possible between cycles to fully quiesce the power regulator
for max battery life. I'm not sure an MCU will be required, but either
way perhaps triggering shutdown on the regulator or cutting off its
power completely with a MOSFET. (But then this calls for a separate
low-power stage to feed the control circuitry. Seems a lot for a
battery widget, but maybe not.)