I have a device that works on **3V battery** (2 D size cells) now I want to debug it so that all the devices work correctly. My device consist of **atmega 328p**, running on 08 MHz Crystal which gets readings from ADXL345 for 2 seconds and then store it in 25Q64FVSIG flash memory and then RFM96 checks if server is available if yes then send the current timestamp, the data stored in flash and current battery otherwise goes to deep sleep mode. There are two 22 pF, one 100uF25V and one 104 pF (to power up radio) capacitors and 10k resistor. This device is enclose in a hard plastic box and then water packed so that not even air goes in or out of that box.



Now the main issue is that most of devices lose battery faster and die after a week or two ,but then one or two keeps on going for 2 to 3 months, yet **same code** is uploaded in all of them. I'm attaching reference picture of my device circuit and let me know if you need anything else.



I have tried to debug through checking the voltage and current drop across various components yet got the same values for every device as compare to the perfect one. I have not checked the capacitance of capacitors let me know if that can be crucial as well.



Let me know if you have any questions related to this in which I can help. Thanks

 

The radio module is a power hog. Most adjust RF power for stable comms. So maybe so you have marginal signals on some devices. Just a guess.

 

Here is a WAG: Do you have any pins that are not connected and set as inputs?

 

Here is a WAG: Do you have any pins that are not connected and set as inputs?

No

 

The radio module is a power hog. Most adjust RF power for stable comms. So maybe so you have marginal signals on some devices. Just a guess.

I considered that too will testing and kept them at the same location while testing.