In a GoodWatch30 prototype with a CC430F6147, the following high
current draw was observed in EnergyTrace. The prototype has no radio,
but ought to gracefully degrade its consumption rather than draw so
much.
GoodWatch Current Consumption
0.00274 +---------------------------------------------------------+
| + + + A + |
0.00272 |-+ "energytrace.txt" A +-|
| A A |
0.0027 |-+ A AA AA A A A AA +-|
|A AA A A A A A A A A A A AAA A |
0.00268 |A+ AAAA AAAA AA A A AAA A AAAAA AA AAAA A+-|
|AA AAAAA AA AAAAAAAAAAA AAA AAA AAAAAA AAAAAA AAA|
0.00266 |AAAAAA AAAAAAAAAAAA AAAA AAAAAAAAAAAA AAAAAAAAAAAAAAAA AA|
|AAAAAAAAAAAAAAAAAAAAAAAAA A AAAAAAAAAAAAAAAAA AAAAAAAAAA|
|AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA|
0.00264 |AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA|
|AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA|
0.00262 |AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA|
|AAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAA AAAA|
0.0026 |AAA AA A AAA AA A AA AAAAA AA AAAA|
| + + + AAA |
0.00258 +---------------------------------------------------------+
10 20 30 40 50 60
Seconds
../bin/batterylife.py <energytrace.txt
2631.487862 µA average consumption
0.052022 months of CR2016 battery life.
Disabling the radio module fixed the problem, implying that we
do not properly disable the radio hardware when no high frequency
crystal is present.
GoodWatch Current Consumption
0.0000048 +--------------------------------------------------------+
| + + + + |
| "energytrace.txt" A |
0.0000047 |-+ AA AA AA +-|
| |
| |
0.0000046 |AAA AA AAAA AAA AA AAAAAAA AA AAA-|
| |
| AA AAAAA AAAA AA AA AAAA AAA AA|
0.0000045 |-+ +-|
| |
0.0000044 |-+ AA AAA AA AAA AA A AA +-|
| |
| |
0.0000043 |AA AAA AA AAAAAA AA AA AAA AA AA AA AA+A|
| |
| + + + + |
0.0000042 +--------------------------------------------------------+
10 20 30 40 50 60
Seconds
../bin/batterylife.py <energytrace.txt
4.471645 µA average consumption
30.614092 months of CR2016 battery life.
This is caused by improper sequencing in radio_off() and radio_strobe() exiting early if !has_radio.
In a GoodWatch30 prototype with a CC430F6147, the following high current draw was observed in EnergyTrace. The prototype has no radio, but ought to gracefully degrade its consumption rather than draw so much.
Disabling the radio module fixed the problem, implying that we do not properly disable the radio hardware when no high frequency crystal is present.
This is caused by improper sequencing in
radio_off()andradio_strobe()exiting early if!has_radio.