Blinkie power
Aug. 30th, 2011 06:55 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
johnridleyAs always, power source is what drives blinkie design. A single lithium cell worked well for the 2011 design, but 3 volts won't cut it for blue or yellow LEDs, too much forward voltage drop.
Today I've been thinking that a single AAA cell into a 5 volt boost power supply might be doable. I priced out the parts and the with battery holder it comes to about $1.50. The downside is that it's mostly surface mount parts; the boost regulator is only really available in SMT, and most parts are way cheaper in SMT (the exception being the inductor, which is way more expensive in SMT, but an axial inductor could easily be surface mounted manually by just bending the leads down.
But then it turns out that 1-5 volt to 5 volt circuit boards with a USB socket soldered on cost < $4 shipped from China. Simply removing the USB socket then soldering it to the board as a sub-assembly would eliminate sourcing all those parts and more importantly would eliminate that surface mount soldering. It might be the way to go.
A single AAA alkaline cell contains about 1.5 watt hours of power; the CR2032 I used last year has about 0.675 or less than half that power. There is some inefficiency involved in the boost converter of course, but it's still a good deal, and AAA alkalines in 100 packs are cheap, 15 cents each or so.
Today I've been thinking that a single AAA cell into a 5 volt boost power supply might be doable. I priced out the parts and the with battery holder it comes to about $1.50. The downside is that it's mostly surface mount parts; the boost regulator is only really available in SMT, and most parts are way cheaper in SMT (the exception being the inductor, which is way more expensive in SMT, but an axial inductor could easily be surface mounted manually by just bending the leads down.
But then it turns out that 1-5 volt to 5 volt circuit boards with a USB socket soldered on cost < $4 shipped from China. Simply removing the USB socket then soldering it to the board as a sub-assembly would eliminate sourcing all those parts and more importantly would eliminate that surface mount soldering. It might be the way to go.
A single AAA alkaline cell contains about 1.5 watt hours of power; the CR2032 I used last year has about 0.675 or less than half that power. There is some inefficiency involved in the boost converter of course, but it's still a good deal, and AAA alkalines in 100 packs are cheap, 15 cents each or so.
- dermot dobson - I wonder if you could mod this sort of simple circuit to get 5V from 1.5 - it has the advantage that it draws just about every scrap of energy out of the cell, but the efficiency loss my be self-defeating in terms of overall life. http://www.emanator.demon.co.uk/bigclive/joule.htm
- John Ridley - That's pretty much what the circuit I'm looking at is doing, though the Joule Thief circuit uses the forward voltage drop triggering of the LED to set the output voltage, whereas I need actual regulated voltage.
I just stole the circuit from the Arduino LilyPad circuit here https://www.sparkfun.com/datasheets/DevTools/LilyPad/LilyPad-PowerSupply.pdf
It uses an NCP1400 boost regulator. - dermot dobson - Whereas you could use the joule thief cct with a diode to pick off the flyback volts from the collector and then shunt regulate, you might be able to use part of your micro to provide feedback to the oscillator drive to close loop regulate. That would result in no SMT and still make use of a 'homebuilt' inductor while keeping the efficiency up.
- John Ridley - I had thought about using the micro to regulate, but isn't that a bit of chicken and egg? I didn't know if the micro could get booted up in time to regulate its own supply in the time it has as the voltage is ramping up.
I'm thinking that I'm going to have at least one SMT kit this year anyway simply to have one. I'm also thinking I'm going to have something besides just blinkies to build. - dermot dobson - How long does the micro take to boot up? A simple capacitor should enable softstart on the inverter, hopefully giving enough time to close the loop.
