# Looking for efficient battery supply to power 3.3V circuit

I have been searching everywhere for the best way to power my circuit and I can't find straight forward information.

My circuit needs 3.3V and the maximum current it draws is 50mA. It will run at 50mA approx 1% of the time and 99% it will run at .004mA (not including the draw of the regulator).

I'd like the physical size to be something close to 3 AA batteries (or smaller).

It has to be rechargeable, have high capacity 3200mAh or more (5000mAh would be great).

So far the best information I see is to use a li-ion rechargeable battery and a TPS63031 voltage regulator. I would like some input regarding if this will even work and if it is an efficient way to power my circuit or if there are better ways.

• Is it practical to let the voltage 'droop' below 3.3V? Many microcontrollers have quite a wide range of operating voltages, and you might be able to take advantage of that too. – gbulmer Sep 19 '14 at 2:24
• I noticed when it gets below 2.8v it starts to have problems. – user3151798 Sep 19 '14 at 2:59

At such low currents, it would be much more efficient to use a linear regulator. If you use a lithium based battery with it you will not be able to get 100% from it because the circuit would stop running once you get below 3.4V on the battery. I'm a fan of TI's TLV70033 LDO.

• Thanks for the info. I was under the impression the regulator will step up the voltage when the battery gets below 3.3v. Does it make sense to use a lithium based battery? What would you use when mAh is a high priority? – user3151798 Sep 19 '14 at 3:04
• How many days should the battery last on one charge? The average current you are looking at is: 50mA x 1% + 0.004mA x 99% = 0.5mA. A 3200mAH battery will last approximately, 267 days. At 267 days, this limits your battery options. A NiMH rechargable has too high of a discharge rate. So the best option is lithium based. Also, by the time you reach 3.4V, not much is left on the battery. Here is a good link describing how much juice is left on a lithium battery. The reason I am hesitant about switching supplies is that they carry – user36770 Sep 19 '14 at 13:10
• - a lot of overhead when you are supplying 0.004mA. – user36770 Sep 19 '14 at 13:13
• Thanks for the information. I just ordered samples of the LDO you recommended. Any recommendations or advice for a battery? – user3151798 Sep 19 '14 at 15:34

A LP2985-33 doesn't need an inductor, will be much less noisy, and is much cheaper. It probably won't be as efficient when the battery is full, but will excel when it reaches 70-80% capacity (which is useful when the battery doesn't get fully charged in order to prolong product life).

• Would I still get the total mAh that the lithium battery is rated for with this regulator? I'm not stuck on using a lithium battery/regulator combo. How would you power this circuit when mAh is a major concern? – user3151798 Sep 19 '14 at 3:06
• No, you won't. But the switcher won't either. If I wanted to eke every last bit of energy from the battery then I'd make sure that the circuit can work with 3.0-3.6V and then use a 3.6V LDO which will ride the battery down to 3.2V or lower. – Ignacio Vazquez-Abrams Sep 19 '14 at 3:10
• It will handle 3.0 - 3.6v. Is the lithium-ion battery a good choice? I was thinking something like this [link] (dx.com/p/…) About how much of the 2600 mAh would you expect to get out of it? – user3151798 Sep 19 '14 at 3:25
• It's hard to say without doing the math, but I would estimate... somewhere between 1600 and 2000mAh? – Ignacio Vazquez-Abrams Sep 19 '14 at 3:27