@scalz
You made a good point by comparing the self discharge rate of nimhs to circuit power consumption. There are also "low selfdischarge" nimhs on the market, for example panasonic/sanoy eneloops. Specs give them discharge rate of 5% per year.

I try to achieve 1-4 month battery life for the high consumption nodes. These nodes will spend most of the time sleeping, but when action is requested, they consume a lot in short time frame.

You still made a good point, that I didn't consider much before. I will re-evaluate what is actually relevant with power optimization. Thanks

@Charles
Thanks for quick reply.

After studying more your design, I believe that there is no need for hw battery protection. The workaround you mentioned works well.
Right after the booster activation a sw battery protection can handle things. This consumes so little energy that it can run for long time and provides sufficient protection to rechargeables, even in very low charge situation.

Looking forward to your release!

Hi Charles,
You have done very thorough work with power optimization. I'm impressed!

I need this is kind of ultra low power wireless node for many applications and have done some designs, but I'm light years behind you After reading your blog, I think I'll use your board.

One battery related question:
Some of my nodes might consume quite a lot power, even if they are well optimized. I planned to use AAA rechargeable NIMHs on these nodes. Rechargeable NIMHs don't like if every bit of juice is squeezed out of them. I understood that your design will drain the batteries completely? Does your design allow to enable some sort of battery protection when using rechargeables, so that they won't be destroyed after few cycles?