Why Does My NiMH Battery Project Run Longer After a Rest With PIC18F27J13?

I've got a small project on the go with a PIC18F27J13 powered by 6 rechargeable NiMH duracell AA batteries using a LDO MCP1702-4002E .. It applies a steady 3.9V for a week or so, before dropping away. I know that 4V for the PIC18F27J13 is above spec and I should really be using a 3.3V LDO MCP1702.

My project utilizes the PIC18F27J13 HLVD facility so that when the voltage drops below 3.4V an interrupt occurs and my project closes down cleanly.

No problems with any of the above.


What I don't understand is that if I power up the project again with the same batteries without having recharged them my project can carry on for another day or two. What's happening? How is my project managing to squeeze the extra juice from the battery after having a short period of rest?

I seem to remember that when I used to test commercial lamps and batteries (lead acid and Zinc ) that most batteries are able to recover after resting .I think it all comes down to what was called depolarisation of anode and cathodes in each cell.However you may be doing the cells harm when running after the first shut down as some of the cells in series may be going into reverse bias, which will shorten the life of the reversed cell considerably.

Conrad is right, if you have a rechargeable torch and it runs down completely, and you leave it for a few hours then switch it on again, the lamp will glow briefly,  same principle. Also Conrad is right that you may be reverse polarising some cells as your dropout battery voltage will be around 4.5V, a long way below the 6 x 1.1V = 6.6V at which you should recharge your NiMHsI don't know the details of your application but you may get better battery life using a lithium battery and running the PIC at a lower voltage - if 3.0 or 3.1V would be acceptable this would work with your existing LDO. Also, have a look at the datasheet for the LDO.  The dropout voltage is only at maximum for loads of around 200mA+.  At maximum load of 100mA it is only around 0.35V:
This is for the Vout = 2.8V part.  As the Vout gets higher the Vdo gets less, but they only show curves for Vout= 1.8 / 2.8 and 5.0V.  It sounds like your application is pretty frugal so you might find a 18650 battery will last up to a month or more.You could probably use a Vout of 3.3V without a problem, depending on you maximum current draw.

Thanks for your replies. Makes is clear especially with the example of the rechargeable torch. I have a torch where when I've switched it off for only short time then back on again I get enough usable light, albeit briefly, to finish the task at hand.My application is frugal using about 7mA for about 3 secs then sleeps for about 20mins when it should  only be using about 20 uA. Having said that I've just noticed that version 2 of my circuit is taking a lot more amps than it should do when sleeping, but still less than when running in non-sleep mode. In changing things must have left something on during sleep mode, need to investigate whether it's the circuit or the program.

.

Got a previous reply to this thread stuck in moderation. Can't remember what I wrote other than
I've seen the phenomenon of the glowing lamp. Never related it to what I was seeing in my circuit. So accept the explanation .
I thought that my problems could be due to poor connections on the breadboard.I seem to have stabilised my circuit by just pushing in the wires as far as they can go.My circuit includes the a PIC18F27J13, a 24LC256, a DS18B20 and a LED. Uses about 4mA for 5 secs then no more than 20uA for 20 minutes. Extremely frugal.
Now using 3 AAA batteries. The 24LC256 operates between 2 2.5v -> 5.5v which is why I've opted for the 3 AAA batteries. I'm using the HLVD option of the PIC18F27J13 to with a setting to close down when it reaches 3v.ThanksAlex