This is a follow-up post to this one from Project 14. The goal was to analyze the battery life of my Dirty Smart Button diaper logger. I meant to post this before the end of the month but left this sitting partially done in my "Drafts" until it was too late. I still wanted to post the follow up, so here it is!
Well it looks like I got 500 button presses exactly in the first round of tests. The MeArm was very useful in doing the lion's share of the work by laboriously pressing buttons time after time. I checked the SQL database and see /exactly/ 500 presses since I replaced the batteries last month (prepping for the Home Automation contest).
My estimates at the time of that article were based around 1000 mAh per battery x 4; so 4000 mAh available. That worked out to ~over 4 years of idle or roughly 2-4000 button presses. That seemed very optimistic at the time which is why I decided to actually test it. So since I got 25% of the estimate in the first round of testing, I started digging deeper.
Once the unit stopped waking up when the button was pressed, I measured the voltage battery pack. It was 2.65 VDC for 4x AAA batteries. Each cell was at .65 volts. So this was clearly lower that was is required to run the linear regulator of the Pro Mini at 3.3 volts. I didn't check the voltage until the next day, and I could see the until trying to do its power-up sequence as the LEDs kept trying to come on; but then it would fail. The DATASHEET for the voltage regulator on the Pro Mini (MIC5205) notes that dropout voltage; while load dependent; should be V-Out + ~.275 mV. So perhaps the Pro Mini was operating outside its spec of 3.3v for a while and/or the battery continued to drain while stuck in this loop before I got back to check it. I checked with my bench power supply and found that the system could work at 3.00 VDC but would fail at 2.9; so it was likely stuck in a power draining situation overnight before I came back to check it.
The next step was to pull the actual data sheet for the battery. I was running a Duracell AAA, model Alkaline battery. Here is the datasheet.
According to this chart, a 1.5V AAA cell would really have more like 0.75-0.9 mAh instead of the 1,000 mAh that Google showed me. So a big fail on my part for not looking up the actual specs and trusting a quick google search. If I calculate the total run time of the system, 500 button presses @ 37 seconds works out to 308 minutes of run time; or 5.13 hours. If we look at the graph of current usage for my application, it looks like I fall between 100 and 250 mA. So that would probably land me right smack where I should be for a load of about 200mAh. Another calculation mistake I had done was adding the current for each battery. However since they were in series instead of parallel, that shouldn't have been done.
Additionally, according to some websites like this one, newborns can use about 10 diapers per day. Our actual usage was higher than that, but not all was logged through the Dirty Smart Button. With 500 presses, that gets 50 days or just shy of 2 months of exclusive use.
So I actually came out right where I should be, although I would still have liked to see some more usage available.
A big thanks to the entire E14 team for sponsoring the contest! I received a finisher package of a RaspberryPi, GraspIO, some stickers, and my favorite - a wind-up action figure - which I've wanted for some time now!