AngusMcDoon
Well-Known Member
Could you consider doing the same for NMEA, if I can figure out how it does it?
May be in the future, but I need to limit the features if it's ever going to see the light of day. I'm not aware of a MOB NMEA-0183 message.
The sleep mode power consumption of the processor seems quite variable between different samples of the same type - 13uA to 18uA. These are the power calculations I have done using those figures for the victim unit...
Victim unit sleep mode varies between 13uA and 18uA.
In transmit mode takes 8.4mA.
Transmit takes 7ms every 4000ms.
Proportion of time transmitting=7/4000=0.00175.
Proportion of time not transmitting=1-0.00175=0.99825
Total Amps = (18/1000000)*0.99825 + (8.4/1000)*0.00175
33uA
Charge storage in CR2450 570 mAh
Estimated life 1.9 years
Charge storage in CR2032 210 mAh
Estimated life 0.7 years
Charge storage in AAA alkaline cell 1000 mAh
Estimated life 3.5 years
There are 2 problems I envisage if I went the CR2032 route...
1) It won't make my target 1 year life
2) My calculations are just that - estimates. I won't have the time before completion to leave one running for a year, so I want a good redundancy in my power calculations.
I really don't want to have to start drilling holes in boxes, because it is difficult to get right and the inevitable cock-ups cost me dearly. Maybe a later version could be smaller and neater, but to start with I want to keep it cheap and simple. As it will be open source, anyone else could have a go at a better victim unit.