In {IDF_TARGET_NAME}, the digital-to-analog converter (ADC) compares the input analog voltage to the reference, and determines each bit of the output digital result. By design, the ADC reference voltage for {IDF_TARGET_NAME} is 1100 mV. However, the true reference voltage can range from 1000 mV to 1200 mV among different chips. This guide introduces the ADC calibration driver to minimize the effect of different reference voltages, and get more accurate output results.
The ADC calibration driver provides ADC calibration scheme(s). From the calibration driver's point of view, an ADC calibration scheme is created for an ADC calibration handle :cpp:type:`adc_cali_handle_t`.
:cpp:func:`adc_cali_check_scheme` can be used to know which calibration scheme is supported on the chip. If you already know the supported schemes, this step can be skipped. Just call the corresponding function to create the scheme handle.
If you use your custom ADC calibration schemes, you could either modify this function :cpp:func:`adc_cali_check_scheme`, or just skip this step and call your custom creation function.
{IDF_TARGET_NAME} supports :c:macro:`ADC_CALI_SCHEME_VER_LINE_FITTING` scheme. To create this scheme, set up :cpp:type:`adc_cali_line_fitting_config_t` first.
There is also a configuration :cpp:member:`adc_cali_line_fitting_config_t::default_vref`. Normally this can be simply set to 0. Line Fitting scheme does not rely on this value. However, if the Line Fitting scheme required eFuse bits are not burnt on your board, the driver will rely on this value to do the calibration.
You can use :cpp:func:`adc_cali_scheme_line_fitting_check_efuse` to check the eFuse bits. Normally the Line Fitting scheme eFuse value is :c:macro:`ADC_CALI_LINE_FITTING_EFUSE_VAL_EFUSE_TP` or :c:macro:`ADC_CALI_LINE_FITTING_EFUSE_VAL_EFUSE_VREF`. This means the Line Fitting scheme uses calibration parameters burnt in the eFuse to do the calibration.
When the Line Fitting scheme eFuse value is :c:macro:`ADC_CALI_LINE_FITTING_EFUSE_VAL_DEFAULT_VREF`, you need to set the :cpp:member:`esp_adc_cali_line_fitting_init::default_vref`. Default vref is an estimate of the ADC reference voltage provided as a parameter during calibration.
This function may fail due to reasons such as :c:macro:`ESP_ERR_INVALID_ARG` or :c:macro:`ESP_ERR_NO_MEM`. Especially, when the function returns :c:macro:`ESP_ERR_NOT_SUPPORTED`, this means the calibration scheme required eFuse bits are not burnt on your board.
{IDF_TARGET_NAME} supports :c:macro:`ADC_CALI_SCHEME_VER_CURVE_FITTING` scheme. To create this scheme, set up :cpp:type:`adc_cali_curve_fitting_config_t` first.
-:cpp:member:`adc_cali_curve_fitting_config_t::chan`, this member is kept here for extensibility. The calibration scheme only differs by attenuation, there is no difference among different channels.
-:cpp:member:`adc_cali_curve_fitting_config_t::chan`, the ADC channel that your ADC raw results are from. The calibration scheme not only differs by attenuation but is also related to the channels.
After setting up the configuration structure, call :cpp:func:`adc_cali_create_scheme_curve_fitting` to create a Curve Fitting calibration scheme handle. This function may fail due to reasons such as :c:macro:`ESP_ERR_INVALID_ARG` or :c:macro:`ESP_ERR_NO_MEM`. Especially, when the function return :c:macro:`ESP_ERR_NOT_SUPPORTED`, this means the calibration scheme required eFuse bits are not burnt on your board.
When the ADC calibration is no longer used, please delete the calibration scheme driver from the calibration handle by calling :cpp:func:`adc_cali_delete_scheme_curve_fitting`.
If you want to use your custom calibration schemes, you could provide a creation function to create your calibration scheme handle. Check the function table ``adc_cali_scheme_t`` in ``components/esp_adc/interface/adc_cali_interface.h`` to know the ESP ADC calibration interface.
After setting up the calibration characteristics, you can call :cpp:func:`adc_cali_raw_to_voltage` to convert the ADC raw result into calibrated result. The calibrated result is in the unit of mV. This function may fail due to an invalid argument. Especially, if this function returns :c:macro:`ESP_ERR_INVALID_STATE`, this means the calibration scheme is not created. You need to create a calibration scheme handle, use :cpp:func:`adc_cali_check_scheme` to know the supported calibration scheme. On the other hand, you could also provide a custom calibration scheme and create the handle.
ADC calibration is only supported under :c:macro:`ADC_ATTEN_DB_0` and :c:macro:`ADC_ATTEN_DB_11`. Under :c:macro:`ADC_ATTEN_DB_0`, the attenuation of ADC is set to 0 dB, and input voltage higher than 950 mV is not supported. Under :c:macro:`ADC_ATTEN_DB_11`, the attenuation of ADC is set to 11 dB, and input voltage higher than 2800 mV is not supported.
The factory function :cpp:func:`esp_adc_cali_new_scheme` is guaranteed to be thread-safe by the driver. Therefore, you can call them from different RTOS tasks without protection by extra locks.
Other functions that take the :cpp:type:`adc_cali_handle_t` as the first positional parameter are not thread-safe, you should avoid calling them from multiple tasks.
-:ref:`CONFIG_ADC_CAL_EFUSE_TP_ENABLE` - disable this to decrease the code size, if the calibration eFuse value is not set to :cpp:type:`ADC_CALI_LINE_FITTING_EFUSE_VAL_EFUSE_TP`.
-:ref:`CONFIG_ADC_CAL_EFUSE_VREF_ENABLE` - disable this to decrease the code size, if the calibration eFuse value is not set to :cpp:type:`ADC_CALI_LINE_FITTING_EFUSE_VAL_EFUSE_VREF`.
-:ref:`CONFIG_ADC_CAL_LUT_ENABLE` - disable this to decrease the code size, if you do not calibrate the ADC raw results under :c:macro:`ADC_ATTEN_DB_11`.
The {IDF_TARGET_NAME} ADC is sensitive to noise, leading to large discrepancies in ADC readings. Depending on the usage scenario, you may need to connect a bypass capacitor (e.g., a 100 nF ceramic capacitor) to the ADC input pad in use, to minimize noise. Besides, multisampling may also be used to further mitigate the effects of noise.
