mirror of https://github.com/ArduPilot/ardupilot
MPU6000: use expected sample time, not actual sample time
this avoids problems in the scheduler turning up as errors in the DCM calculation. The MPU6k produces samples with more reliable timing than our scheduler does!
This commit is contained in:
parent
e0197a3f7c
commit
82b036282b
|
@ -185,7 +185,6 @@ int16_t AP_InertialSensor_MPU6000::_mpu6000_product_id = AP_PRODUCT_ID_NONE;
|
|||
|
||||
// variables to calculate time period over which a group of samples were collected
|
||||
static volatile uint32_t _delta_time_micros = 1; // time period overwhich samples were collected (initialise to non-zero number but will be overwritten on 2nd read in any case)
|
||||
static volatile uint32_t _delta_time_start_micros = 0; // time we start collecting sample (reset on update)
|
||||
static volatile uint32_t _last_sample_time_micros = 0; // time latest sample was collected
|
||||
|
||||
// DMP related static variables
|
||||
|
@ -253,8 +252,7 @@ bool AP_InertialSensor_MPU6000::update( void )
|
|||
_count = 0;
|
||||
|
||||
// record sample time
|
||||
_delta_time_micros = _last_sample_time_micros - _delta_time_start_micros;
|
||||
_delta_time_start_micros = _last_sample_time_micros;
|
||||
_delta_time_micros = count * _micros_per_sample;
|
||||
SREG = oldSREG;
|
||||
|
||||
count_scale = 1.0 / count;
|
||||
|
@ -391,15 +389,18 @@ void AP_InertialSensor_MPU6000::hardware_init(Sample_rate sample_rate)
|
|||
case RATE_50HZ:
|
||||
rate = MPUREG_SMPLRT_50HZ;
|
||||
default_filter = BITS_DLPF_CFG_20HZ;
|
||||
_micros_per_sample = 20000;
|
||||
break;
|
||||
case RATE_100HZ:
|
||||
rate = MPUREG_SMPLRT_100HZ;
|
||||
default_filter = BITS_DLPF_CFG_42HZ;
|
||||
_micros_per_sample = 10000;
|
||||
break;
|
||||
case RATE_200HZ:
|
||||
default:
|
||||
rate = MPUREG_SMPLRT_200HZ;
|
||||
default_filter = BITS_DLPF_CFG_42HZ;
|
||||
_micros_per_sample = 5000;
|
||||
break;
|
||||
}
|
||||
|
||||
|
|
|
@ -75,6 +75,7 @@ private:
|
|||
// ensure we can't initialise twice
|
||||
bool _initialised;
|
||||
static int16_t _mpu6000_product_id;
|
||||
uint32_t _micros_per_sample;
|
||||
|
||||
// dmp related methods and parameters
|
||||
static void dmp_register_write(uint8_t bank, uint8_t address, uint8_t num_bytes, uint8_t data[]); // Method to write multiple bytes into dmp registers. Requires a "bank"
|
||||
|
|
Loading…
Reference in New Issue