Hello ST team, as mentioned in the title, I found that the power consumption when using the ULP library to drive TOF L1 is inconsistent with the manual. I'm using the official evaluation board vl53l1-satel. Below is my test code and power consumption data. In my code, you'll see both ULD and ULP libraries, but I'm actually only using the ULP library; the ULD library is commented out.
int main(void)
{
/* USER CODE BEGIN 1 */
uint16_t dev=0x52;
int status=0;
volatile int IntCount;
uint8_t byteData, sensorState=0;
uint16_t wordData;
uint8_t measurement_status;
uint16_t estimated_distance_mm;
uint16_t sigma_mm;
uint16_t signal_kcps;
uint16_t ambient_kcps;
/* USER CODE END 1 */
/* MCU Configuration--------------------------------------------------------*/
/* Reset of all peripherals, Initializes the Flash interface and the Systick. */
HAL_Init();
/* USER CODE BEGIN Init */
/* USER CODE END Init */
/* Configure the system clock */
SystemClock_Config();
/* USER CODE BEGIN SysInit */
/* USER CODE END SysInit */
/* Initialize all configured peripherals */
MX_GPIO_Init();
MX_I2C1_Init();
MX_TIM16_Init();
MX_LPUART1_UART_Init();
/* USER CODE BEGIN 2 */
HAL_TIM_PWM_Start(&htim16,TIM_CHANNEL_1);
printf("\r\n--MCU ready !--\r\n");
I2C_Scan(&hi2c1);
//while(1)I2C_Scan(&hi2c1);
status = VL53L1_RdByte(dev, 0x010F, &byteData);
printf("VL53L1X Model_ID: %X\r\n", byteData);
status = VL53L1_RdByte(dev, 0x0110, &byteData);
printf("VL53L1X Module_Type: %X\r\n", byteData);
status = VL53L1_RdWord(dev, 0x010F, &wordData);
printf("VL53L1X: %X\r\n", wordData);
while(sensorState==0){
status = VL53L1X_BootState(dev, &sensorState);
HAL_Delay(2);
}
printf("Chip booted\r\n");
/* This function must to be called to initialize the sensor with the default setting */
status = VL53L1X_ULP_SensorInit(dev);
if(status) printf("init error\r\n");
/* Optional functions to be used to change the main ranging parameters according the application requirements to get the best ranging performances */
//VL53L1X_ULP_SetInterruptConfiguration(dev,0,0);
status = VL53L1X_ULP_SetInterMeasurementInMs(dev, 500); /* in ms possible values [20, 50, 100, 200, 500] */
status = VL53L1X_ULP_SetMacroTiming(dev, 20); //MacropPeriod
status = VL53L1X_ULP_SetSignalThreshold(dev, 1600);
status = VL53L1X_ULP_SetSigmaThreshold(dev, 50);
// status = VL53L1X_SetOffset(dev,20); /* offset compensation in mm */
status = VL53L1X_ULP_SetROI(dev, 16); /* minimum ROI 4 */
// status = VL53L1X_CalibrateOffset(dev, 140, &offset); /* may take few second to perform the offset cal*/
// status = VL53L1X_CalibrateXtalk(dev, 1000, &xtalk); /* may take few second to perform the xtalk cal */
printf("VL53L1X Ultra Lite Driver Example running ...\r\n");
status = VL53L1X_ULP_StartRanging(dev); /* This function has to be called to enable the ranging */
if(status) printf("init error\r\n");
/* USER CODE END 2 */
/* Infinite loop */
/* USER CODE BEGIN WHILE */
while (1)
{
HAL_Delay(100);
status = VL53L1X_ULP_DumpDebugData(dev,
&measurement_status,
&estimated_distance_mm,
&sigma_mm,
&signal_kcps,
&ambient_kcps);
if (status == VL53L1X_ULP_ERROR_NONE)
{
printf("Measurement Status: %u ", measurement_status);
printf("Estimated Distance: %u mm ", estimated_distance_mm);
printf("Sigma: %u mm ", sigma_mm);
printf("Signal Rate: %u kcps ", signal_kcps);
printf("Ambient Rate: %u kcps\r\n", ambient_kcps);
}
else
{
printf("Error reading measurement: %d\r\n", status);
}
VL53L1X_ULP_StopRanging(dev);
HAL_Delay(2000);
VL53L1X_ULP_StartRanging(dev);
/* USER CODE END WHILE */
/* USER CODE BEGIN 3 */
//HAL_Delay(1000);
}
/* USER CODE END 3 */
}
My main confusion is why there are still several mA of power consumption when the TOF rangefinder completes measurement and enters software standby mode. My MCU and TOF have separate, independent power supplies. Also, when using the ULD library, the power consumption after ranging completion and entering software standby is only a few μA, which is consistent with the manual. Therefore, I don't think it's an issue with my hardware connections.When I pull XShut low, it consumes about 400μA of current at 2.8V supply (all my power to the TOF is 2.8V)。I also made some modifications to the circuit, removing the voltage regulator chip and shorting R21 to get closer to the device's power consumption, but I didn't remove the level shifting chip because when I removed this chip, the TOF stopped working for some reason.
This took me forever, but I have an idea.
I ranged and read out the data, printing it. I then cleared the interrupt.
On my power plot you can see the high-power ranging, then the lower-power I2C access. Then (after the print time) you can see the I2C command to clear the interrupt. Then the chip goes into lowest power.
I then ran the SAME code, but moving the clearing of the interrupt BEFORE I read out the I2C data.
Note how the power consumption goes nearly to zero just after the I2C reads.
So I'm suggesting you clear the interrupt before you do anything else.
My code looks like this:
void get_data_by_interrupt(uint16_t dev){
loop=0;
do
{
__WFI(); // Wait for interrupt
if(IntCount !=0 ){
IntCount=0;
VL53L4CD_ClearInterrupt(dev);
/* Read measured distance. RangeStatus = 0 means valid data */
VL53L4CD_GetResult(dev, &results);
printf("%4u Status= %2u, Distance= %5u mm, Signal= %6u, %6u kcps/spad, Spads= %3u Sigma= %2u\n",
loop,
results.range_status,
results.distance_mm,
results.signal_rate_kcps,
results.signal_per_spad_kcps,
results.number_of_spad,
results.sigma_mm);
loop++;
}while(1);
}The important bit being the ClearInterrupt() as soon as possible after detecting it.
Hopefully that works. My theory is the sensor will not go into lowest power until the interrupt is cleared.
BTW: I used the SATELLITE VL53L4CD sensor and the X-Nucleo-LPM01A with the STM32CubeMonitorPower software to do the analysis.
