diff options
| author | Selene ToyKeeper | 2019-11-14 01:52:50 -0700 |
|---|---|---|
| committer | Selene ToyKeeper | 2019-11-14 01:52:50 -0700 |
| commit | 8b59e880614455bd7bc7d8595de847dd6fe9b5b2 (patch) | |
| tree | a7426296b9fa1471ba2841eba0415bbfaed3e5c7 /spaghetti-monster/fsm-adc.h | |
| parent | fixed some compile issues related to delay_4ms() (diff) | |
| download | anduril-8b59e880614455bd7bc7d8595de847dd6fe9b5b2.tar.gz anduril-8b59e880614455bd7bc7d8595de847dd6fe9b5b2.tar.bz2 anduril-8b59e880614455bd7bc7d8595de847dd6fe9b5b2.zip | |
refactored how interrupts work...
set a flag and return immediately,
then handle the actual logic later during a less-critical code path
Enables smarter responses to standby wakeups.
Seems to fix missed button presses during standby,
and most of the too-fast sleep ticks.
Also eliminated waits from button state measurement, so it can happen easier during standby.
(also eliminates the chance of an infinite loop on extra-noisy hardware)
Also might improve timing-sensitive interrupts like attiny85 PWM channel 4,
or a PWM-DSM hybrid technique I'd like to try.
BUT this change also appears to break the thermal sensor, so that needs to be fixed.
Diffstat (limited to '')
| -rw-r--r-- | spaghetti-monster/fsm-adc.h | 8 |
1 files changed, 7 insertions, 1 deletions
diff --git a/spaghetti-monster/fsm-adc.h b/spaghetti-monster/fsm-adc.h index 274fb4d..acb3da6 100644 --- a/spaghetti-monster/fsm-adc.h +++ b/spaghetti-monster/fsm-adc.h @@ -38,9 +38,15 @@ #define VOLTAGE_FUDGE_FACTOR 5 #endif #endif + +volatile uint8_t irq_adc = 0; // ADC interrupt happened? +uint16_t adc_value; // last ADC measurement +uint8_t adcint_enable = 0; // is the current ADC result needed? +void ADC_inner(); // do the actual ADC-related calculations + volatile uint8_t voltage = 0; -volatile uint8_t adcint_enable; // kludge, because adc auto-retrigger won't turn off void low_voltage(); + #ifdef USE_BATTCHECK void battcheck(); #ifdef BATTCHECK_VpT |