Thanks Bob - that's a fantastic help. It's great to have advice from someone with experience in this. My plan wasn't to have two battery packs, but rather to tap the five cell pack at the three cell mark for the lower voltage. The choice of batteries was more due to the fact that I already have them rather than anything else. Component availability is an issue for me - this is a hobbyist project rather than intended for production. I'm also avoiding SMT where I can since the li'l buggers are so hard to solder. I love your switching supply suggestion - I'll definitely look into trying an adjustable switching regulator and battery charger like that. You said you used a FET to switch the circuit. Sorry to appear stupid here but did you use a p-channel FET on the positive side or an n-channel FET switching the ground? Or something else? The reason I ask is that I tried a design using an n-channel FET to switch the positive side and ran into issues relating to the floating voltage level. I wasn't too keen on switching the ground but now I think about it that might be ok. What's the preferred way of doing this? Cheers, Zik On 10/04/06, Bob Axtell wrote: > Zik Saleeba wrote: > > >I'm designing a handheld PIC-based device (a handheld GPS / nav. > >unit). It'll be using a PIC18LF2620 and will be battery powered. I'd > >like to give it a software power-on/off feature so I can use one of > >the control buttons for power on/off rather than having a big ugly > >power switch. I was wondering what kind of circuit people would > >suggest for the power on/off feature? > > > >Just to complicate the issue some parts of the circuit operate at 5V > >and others need 3.3V. The 18LF2620 can operate at down to 2.0V but > >when the circuit's in full-power mode it needs to be running at 5V to > >talk to some of its peripherals. My plan was to use a 6V Ni-MH battery > >pack with MAX883 and MAX884 regulators. These have a low dropout > >voltage and provide a shutdown mode. I was thinking I could power the > >PIC through a schottky diode off the 5V normally and then have an > >"off" mode where it's powered on ~3.6V (three Ni-MH cells) through a > >different schottky diode. The "off" mode would power only the PIC > >rather than the whole circuit. The PIC would be in sleep mode while > >the device is "off". A button press on one of the interrupt-on-change > >lines would wake the CPU and it would raise an output line to tell the > >regulators to power up again. > > > >Does this sound line a feasible plan? Is there an easier way of doing this? > > > >Thanks, > >Zik > >(Apologies if this appears multiple times - I've been having some > >problems with the mailing list) > > > > > > > I ran into one of those power up / power down problems before. I used an > SC70-sized (approx SOT23-6) > set of flipflops to actuate the PIC and other peripherals, switching at > the battery level. The power to the > PIC and the other stuff is then supplied thru a switched FET. > > Now, the battery system... > > .. I am not trying to be critical here... the diode drops etc... but I > think you are missing a lot about the way > batteries work. NiMH cells are nominally 1.2V. When you say 6V, do you > really mean 5 cells (5*1.2=6V)? > That's a hefty, barely manageable pack, now you are adding another pack > just for the OFF mode? > > If it were me, I'd design the system around a pack operating with two > Li-Ion 3.6V cells in series at 7.2V. I > say this because there are some reasonably-priced HIGH QUALITY battery > charger chips made for two > Li-Ion cell packs. Then you design a simple switching regulator that > will supply 4.75V reliably with minimal losses > FROM THE 7.2V supply. That means that your power from the pack is > efficiently converted to 4.75V, with a > further pass 3.6V regulation during ON mode. Then, in the OFF mode, you > can simply ADJUST THE > SWITCHER to deliver 3.6V during the off mode. Look how it would work.... > You now can get FLAWLESS > power out of a pack that can fall as flat as 2.6V per cell, yet the > switcher will STILL deliver 5V as long as > 5V is at the input. The advantages are: EXTREMELY long battery charge > life. Dramatic reduction in size > and weight. Charges up quickly... Li-Ion cells can be charged in a few > minutes to a useable amount. > > Now, to carry it through the last step... Linear Technology sells a > switching battery charger chip. You can > use a wallwart or automobile as a charging source, yet the chip > generates almost NO heat. That being > the case, you can design your product with the battery charger INSIDE > YOUR PRODUCT. Very > convenient. The wallwart can be ANYTHING from 9V to 19V, and a cigarette > lighter plug is inside that > range perfectly.... > > That entire system I just described can fit into a 1.5" square, the > downswitcher and the downswitching > charger. > > Did that help? > > --Bob > > -- > Note: To protect our network, > attachments must be sent to > attach@engineer.cotse.net . > 1-520-850-1673 USA/Canada > http://beam.to/azengineer > > -- > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist