adam/Meck
1
0
Fork 0
mirror of https://github.com/pelgraine/Meck.git synced 2026-07-28 20:32:47 +02:00
Code Issues Packages Projects Releases Wiki Activity

Watches: set the BMA423 step counter watermark, raise ODR to 100 Hz

Browse Source 
Both watches undercount steps. Two divergences from LilyGo's own configuration
for this hardware, applied to both boards:

1. Watermark. LilyGoLib calls setStepCounterWatermark(1) and SensorLib's
   BMA423_StepDetector example passes step_counter_wm = 1. Meck's raw-I2C enable
   never touched those bits, leaving whatever the config-file blob defaults to.
   The watermark is BMA423_STEP_CNTR_WM_MSK (0x03FF), spanning feature_config
   [0x36] as LSB and the low two bits of [0x37] as MSB; it does not collide with
   the enable bit, which is bit 4 of [0x37]. Both are now written in one
   read-modify-write. The bit arithmetic was checked exhaustively against
   bma423_step_counter_set_watermark() + feature_enable() over all 65536 starting
   states of cfg[0x36]/cfg[0x37]: identical.

2. ODR. SensorLib's step detector example runs the accelerometer at 100 Hz. Meck
   used 50 Hz. Everything else already matched: NORMAL, FS_2G, OSR2_AVG2,
   CIC_AVG_MODE. Reverting the single 100.0f literal to 50.0f A/Bs this against
   the watermark change.

Ruled out: the axis remap is BOTTOM_LAYER_TOP_RIGHT_CORNER on both, identical to
LilyGoLib, and the enable bit itself was already register-for-register what
bma423_feature_enable(BMA423_STEP_CNTR, TRUE) writes.

Not attempted: the 25-value pedometer parameter block at feature_config[0x04..0x35],
reachable via bma423_stepcounter_set_parameter() but with no reference values.
This commit is contained in:
meck
2026-07-10 10:28:14 +00:00
committed by pelgraine
parent 006d3bd00c
commit 755f883d66
2 changed files with 40 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
variants/lilygo_twatch_s3/TWatchS3Board.cpp
+20 -1
View File
@@ -16,6 +16,16 @@ void IRAM_ATTR TWatchS3Board::onTiltISR() { _tilt_flag = true; }
#define BMA423_FEATURE_LEN 64
#define BMA423_STEP_EN_BYTE 0x37 // BMA423_STEP_CNTR_OFFSET(0x36) + 1
#define BMA423_STEP_EN_BIT 0x10 // BMA423_STEP_CNTR_EN_MSK
// Step counter watermark: a 10-bit field (BMA423_STEP_CNTR_WM_MSK = 0x03FF)
// spanning feature_config[0x36] as LSB and the low two bits of [0x37] as MSB.
// It does not collide with the enable bit, which is bit 4 of [0x37] (bit 12 of
// the 16-bit word). Bosch's bma423_step_counter_set_watermark() writes exactly
// these bits; SensorLib's enableStepCounter() calls it with 1, and LilyGo's own
// firmware calls setStepCounterWatermark(1). Meck never set it, which is the
// prime suspect for the undercounting.
#define BMA423_STEP_WM_LSB_BYTE 0x36 // BMA423_STEP_CNTR_OFFSET
#define BMA423_STEP_WM_MSK 0x03FF // BMA423_STEP_CNTR_WM_MSK
#define BMA423_STEP_WM_LEVEL 1 // matches LilyGoLib and the SensorLib example
#define BMA423_REG_POWER_CONF 0x7C // BMA4_POWER_CONF_ADDR
#define BMA423_ADV_PWR_SAVE_BIT 0x01 // BMA4_ADVANCE_POWER_SAVE_MSK
@@ -49,6 +59,11 @@ static void bma423EnableStepCounter() {
uint8_t cfg[BMA423_FEATURE_LEN];
if (bma423ReadRegs(BMA423_REG_FEATURE_CONFIG, cfg, BMA423_FEATURE_LEN)) {
// Watermark first, then the enable bit, in a single read-modify-write.
uint16_t wm = ((uint16_t)cfg[BMA423_STEP_EN_BYTE] << 8) | cfg[BMA423_STEP_WM_LSB_BYTE];
wm = (wm & ~BMA423_STEP_WM_MSK) | (BMA423_STEP_WM_LEVEL & BMA423_STEP_WM_MSK);
cfg[BMA423_STEP_WM_LSB_BYTE] = (uint8_t)(wm & 0xFF);
cfg[BMA423_STEP_EN_BYTE] = (uint8_t)((wm >> 8) & 0xFF);
cfg[BMA423_STEP_EN_BYTE] |= BMA423_STEP_EN_BIT;
bma423WriteRegs(BMA423_REG_FEATURE_CONFIG, cfg, BMA423_FEATURE_LEN);
delay(1); // write settle
@@ -66,8 +81,12 @@ void TWatchS3Board::begin() {
_accel = new SensorBMA423();
if (_accel->begin(Wire, I2C_ADDR_ACCEL, PIN_BOARD_SDA, PIN_BOARD_SCL)) {
_accel->setRemapAxes(SensorRemap::BOTTOM_LAYER_TOP_RIGHT_CORNER);
// 100 Hz ODR: SensorLib's BMA423_StepDetector example runs the pedometer at
// 100 Hz, not the 50 Hz used here previously. Everything else already matched
// (NORMAL, FS_2G, OSR2_AVG2, CIC_AVG_MODE). Revert this one literal to 50.0f
// to A/B it against the watermark change above.
_accel->configAccelerometer(OperationMode::NORMAL, AccelFullScaleRange::FS_2G,
50.0f, AccelBandwidth::OSR2_AVG2, AccelPerfMode::CIC_AVG_MODE);
100.0f, AccelBandwidth::OSR2_AVG2, AccelPerfMode::CIC_AVG_MODE);
// INT1 pin electrical config: level trigger, active high, push-pull,
// output enabled. INT1_IO_CTRL resets to output-disabled, so without
// this the pin never drives and INPUT_PULLDOWN reads low forever.
variants/lilygo_twatch_s3_plus/TWatchS3PlusBoard.cpp
+20 -1
View File
@@ -17,6 +17,16 @@ void IRAM_ATTR TWatchS3PlusBoard::onTiltISR() { _tilt_flag = true; _tilt_isr_cou
#define BMA423_FEATURE_LEN 64
#define BMA423_STEP_EN_BYTE 0x37 // BMA423_STEP_CNTR_OFFSET(0x36) + 1
#define BMA423_STEP_EN_BIT 0x10 // BMA423_STEP_CNTR_EN_MSK
// Step counter watermark: a 10-bit field (BMA423_STEP_CNTR_WM_MSK = 0x03FF)
// spanning feature_config[0x36] as LSB and the low two bits of [0x37] as MSB.
// It does not collide with the enable bit, which is bit 4 of [0x37] (bit 12 of
// the 16-bit word). Bosch's bma423_step_counter_set_watermark() writes exactly
// these bits; SensorLib's enableStepCounter() calls it with 1, and LilyGo's own
// firmware calls setStepCounterWatermark(1). Meck never set it, which is the
// prime suspect for the undercounting.
#define BMA423_STEP_WM_LSB_BYTE 0x36 // BMA423_STEP_CNTR_OFFSET
#define BMA423_STEP_WM_MSK 0x03FF // BMA423_STEP_CNTR_WM_MSK
#define BMA423_STEP_WM_LEVEL 1 // matches LilyGoLib and the SensorLib example
#define BMA423_REG_POWER_CONF 0x7C // BMA4_POWER_CONF_ADDR
#define BMA423_ADV_PWR_SAVE_BIT 0x01 // BMA4_ADVANCE_POWER_SAVE_MSK
@@ -50,6 +60,11 @@ static void bma423EnableStepCounter() {
uint8_t cfg[BMA423_FEATURE_LEN];
if (bma423ReadRegs(BMA423_REG_FEATURE_CONFIG, cfg, BMA423_FEATURE_LEN)) {
// Watermark first, then the enable bit, in a single read-modify-write.
uint16_t wm = ((uint16_t)cfg[BMA423_STEP_EN_BYTE] << 8) | cfg[BMA423_STEP_WM_LSB_BYTE];
wm = (wm & ~BMA423_STEP_WM_MSK) | (BMA423_STEP_WM_LEVEL & BMA423_STEP_WM_MSK);
cfg[BMA423_STEP_WM_LSB_BYTE] = (uint8_t)(wm & 0xFF);
cfg[BMA423_STEP_EN_BYTE] = (uint8_t)((wm >> 8) & 0xFF);
cfg[BMA423_STEP_EN_BYTE] |= BMA423_STEP_EN_BIT;
bma423WriteRegs(BMA423_REG_FEATURE_CONFIG, cfg, BMA423_FEATURE_LEN);
delay(1); // write settle
@@ -67,8 +82,12 @@ void TWatchS3PlusBoard::begin() {
_accel = new SensorBMA423();
if (_accel->begin(Wire, I2C_ADDR_ACCEL, PIN_BOARD_SDA, PIN_BOARD_SCL)) {
_accel->setRemapAxes(SensorRemap::BOTTOM_LAYER_TOP_RIGHT_CORNER);
// 100 Hz ODR: SensorLib's BMA423_StepDetector example runs the pedometer at
// 100 Hz, not the 50 Hz used here previously. Everything else already matched
// (NORMAL, FS_2G, OSR2_AVG2, CIC_AVG_MODE). Revert this one literal to 50.0f
// to A/B it against the watermark change above.
_accel->configAccelerometer(OperationMode::NORMAL, AccelFullScaleRange::FS_2G,
50.0f, AccelBandwidth::OSR2_AVG2, AccelPerfMode::CIC_AVG_MODE);
100.0f, AccelBandwidth::OSR2_AVG2, AccelPerfMode::CIC_AVG_MODE);
// INT1 pin electrical config: level trigger, active high, push-pull,
// output enabled. INT1_IO_CTRL resets to output-disabled, so without
// this the pin never drives and INPUT_PULLDOWN reads low forever.