Skip to content

AA Battery Pack to 3.3V for IoT Projects

AASwitcherTest1I tweeted about my experiments with the AAT1217 switching supply controller a few days ago.  I found the chip in a product design I was evaluating. It looked like a great answer to the power problem in an IoT project I’m working on. 

I ordered a few from Mouser and some breakout boards to try.  My first experiment was to fire it up on a breadboard with a load and a AA battery pack.

I recorded the performance while the batteries drained into a constant load.


Result Summary

  • Rock Solid 3.3VDC at 75mA continuous 
  • More than 15 hours. I missed the very end, it went longer.
  • From a pair of AA Duracell Alkaline batteries.
  • Simple to use, 1 inductor, 2 caps and a resistor


Time hr:min Input Watts In Output Watts Out Efficiency
0 ? Fresh ? 3.3V@75mA 0.258 ?
1:00 2.93V@105mA 0.308 3.3V@75mA 0.258 84%
1:15 107mA   3.3V@75mA 0.258  
3:25 2.65V@118mA 0.313 3.3V@75mA 0.258 82%
12:00 2.30V@140.5mA 0.323 3.3V@75mA 0.258 80%
14:15 2.10V@149.5mA 0.314 3.3V@75mA 0.258 82%


Some data is missing above. I was doing this all by hand and just stopping by now and then. The test ran much longer than I expected it would.  I had to sleep.  The 12:00 reading was an unexpected wake up with the 14:15 being a return to bed mark.  The pack was all but dead at this point, but I don’t know how much longer it took to die completely.

Test Circuit

I used the example circuit from Skyworks for the AAT1217 breadboard. The part was a pain to solder to the adapter I had. The adaptor was for a SOT type package with different spacing, but I made it work. 

SchTestThe AAT1217 only needs an inductor and two caps.  A resistor is used to hold the enable high. The output was 3.3V with about 100mV of high-frequency noise/ripple.  I added a 0.15uF ceramic cap to clean that up a bit.



Parts Used

Part Qty Source
AAT1217ICA-3.3-T1 1 Mouser
4.7uH Inductor 1 Mouser
4.7uF Al Elec Caps 2 bench stock
0.15 Ceramic Cap 1 bench stock
1MOhm Resistor 1 bench stock
22Ohm Resistor 2 bench stock


The 22Ohm resistors were used to make a 44 ohm load. At 3.3V this would be a continuous 75mA load.


The AAT1217 works great.  I will do some more test, but I plan to use it in future projects as the battery power solution with 2 AA battery packs. 

Battery lifetime was considerably longer than expected. I probably have a flaw in my test somewhere.

Based on discharge curves an Alkaline AA would give about 2.4A/hr @100mA and discharge to about 1V. The charge after about 1V is not significant. I estimate that the batteries should have drained in less than 8 hours.

The AAT1217 is specified to run down to 0.8V which would pull the pair of series AA batteries down to 0.4V each.  That might yield some more time but not double the estimate.

My assumptions are that a AA battery is all but dead when it’s at 1V. But that might be wrong while under load. The load was close to the 100mA load a AA is designed for.

Efficiency was much less than expected. I was hoping to see closer to 90% which would make a significant difference in battery life.  I’m thinking the inductor used is not the best for this application. I chose it to be easy to use in a solderless breadboard.  I plan to get a few better parts when I do a more controlled test.

I used a new Rigol DM3058E and a Fluke 175 for measurements.  I hoped to use just the Rigol as it should be able to display two measurements, voltage, and current, at the same time but I did not figure that out.

The Rigol DMM also has an easy to use serial output so I could record one of the data values simply by logging it. I might create a QT app to do a DMM logger, especially if I can get the meter to do both measurements.

When I repeat this test, I’ll pull out a device  I designed and built if for some consulting work recently. It could be adjusted to be an ideal tool for this kind of testing.

Posted in Uncategorized.

0 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

Some HTML is OK

or, reply to this post via trackback.