[adc] Fix 11db deprecation warning (#6749)

esphome/components/adc/__init__.py

@@ -18,11 +18,23 @@
 
 CODEOWNERS = ["@esphome/core"]
 
+adc_ns = cg.esphome_ns.namespace("adc")
+
+
+"""
+From the below patch versions (and 5.2+) ADC_ATTEN_DB_11 is deprecated and replaced with ADC_ATTEN_DB_12.
+4.4.7
+5.0.5
+5.1.3
+5.2+
+"""
+
 ATTENUATION_MODES = {
     "0db": cg.global_ns.ADC_ATTEN_DB_0,
     "2.5db": cg.global_ns.ADC_ATTEN_DB_2_5,
     "6db": cg.global_ns.ADC_ATTEN_DB_6,
-    "11db": cg.global_ns.ADC_ATTEN_DB_11,
+    "11db": adc_ns.ADC_ATTEN_DB_12_COMPAT,
+    "12db": adc_ns.ADC_ATTEN_DB_12_COMPAT,
     "auto": "auto",
 }
 

esphome/components/adc/adc_sensor.cpp

@@ -62,7 +62,7 @@ extern "C"
   }
 
   // load characteristics for each attenuation
-  for (int32_t i = 0; i <= ADC_ATTEN_DB_11; i++) {
+  for (int32_t i = 0; i <= ADC_ATTEN_DB_12_COMPAT; i++) {
     auto adc_unit = channel1_ != ADC1_CHANNEL_MAX ? ADC_UNIT_1 : ADC_UNIT_2;
     auto cal_value = esp_adc_cal_characterize(adc_unit, (adc_atten_t) i, ADC_WIDTH_MAX_SOC_BITS,
                                               1100,  // default vref
@@ -118,8 +118,8 @@ void ADCSensor::dump_config() {
       case ADC_ATTEN_DB_6:
         ESP_LOGCONFIG(TAG, " Attenuation: 6db");
         break;
-      case ADC_ATTEN_DB_11:
-        ESP_LOGCONFIG(TAG, " Attenuation: 11db");
+      case ADC_ATTEN_DB_12_COMPAT:
+        ESP_LOGCONFIG(TAG, " Attenuation: 12db");
         break;
       default:  // This is to satisfy the unused ADC_ATTEN_MAX
         break;
@@ -183,12 +183,12 @@ float ADCSensor::sample() {
     return mv / 1000.0f;
   }
 
-  int raw11 = ADC_MAX, raw6 = ADC_MAX, raw2 = ADC_MAX, raw0 = ADC_MAX;
+  int raw12 = ADC_MAX, raw6 = ADC_MAX, raw2 = ADC_MAX, raw0 = ADC_MAX;
 
   if (channel1_ != ADC1_CHANNEL_MAX) {
-    adc1_config_channel_atten(channel1_, ADC_ATTEN_DB_11);
-    raw11 = adc1_get_raw(channel1_);
-    if (raw11 < ADC_MAX) {
+    adc1_config_channel_atten(channel1_, ADC_ATTEN_DB_12_COMPAT);
+    raw12 = adc1_get_raw(channel1_);
+    if (raw12 < ADC_MAX) {
       adc1_config_channel_atten(channel1_, ADC_ATTEN_DB_6);
       raw6 = adc1_get_raw(channel1_);
       if (raw6 < ADC_MAX) {
@@ -201,9 +201,9 @@ float ADCSensor::sample() {
       }
     }
   } else if (channel2_ != ADC2_CHANNEL_MAX) {
-    adc2_config_channel_atten(channel2_, ADC_ATTEN_DB_11);
-    adc2_get_raw(channel2_, ADC_WIDTH_MAX_SOC_BITS, &raw11);
-    if (raw11 < ADC_MAX) {
+    adc2_config_channel_atten(channel2_, ADC_ATTEN_DB_12_COMPAT);
+    adc2_get_raw(channel2_, ADC_WIDTH_MAX_SOC_BITS, &raw12);
+    if (raw12 < ADC_MAX) {
       adc2_config_channel_atten(channel2_, ADC_ATTEN_DB_6);
       adc2_get_raw(channel2_, ADC_WIDTH_MAX_SOC_BITS, &raw6);
       if (raw6 < ADC_MAX) {
@@ -217,25 +217,25 @@ float ADCSensor::sample() {
     }
   }
 
-  if (raw0 == -1 || raw2 == -1 || raw6 == -1 || raw11 == -1) {
+  if (raw0 == -1 || raw2 == -1 || raw6 == -1 || raw12 == -1) {
     return NAN;
   }
 
-  uint32_t mv11 = esp_adc_cal_raw_to_voltage(raw11, &cal_characteristics_[(int32_t) ADC_ATTEN_DB_11]);
+  uint32_t mv12 = esp_adc_cal_raw_to_voltage(raw12, &cal_characteristics_[(int32_t) ADC_ATTEN_DB_12_COMPAT]);
   uint32_t mv6 = esp_adc_cal_raw_to_voltage(raw6, &cal_characteristics_[(int32_t) ADC_ATTEN_DB_6]);
   uint32_t mv2 = esp_adc_cal_raw_to_voltage(raw2, &cal_characteristics_[(int32_t) ADC_ATTEN_DB_2_5]);
   uint32_t mv0 = esp_adc_cal_raw_to_voltage(raw0, &cal_characteristics_[(int32_t) ADC_ATTEN_DB_0]);
 
   // Contribution of each value, in range 0-2048 (12 bit ADC) or 0-4096 (13 bit ADC)
-  uint32_t c11 = std::min(raw11, ADC_HALF);
+  uint32_t c12 = std::min(raw12, ADC_HALF);
   uint32_t c6 = ADC_HALF - std::abs(raw6 - ADC_HALF);
   uint32_t c2 = ADC_HALF - std::abs(raw2 - ADC_HALF);
   uint32_t c0 = std::min(ADC_MAX - raw0, ADC_HALF);
   // max theoretical csum value is 4096*4 = 16384
-  uint32_t csum = c11 + c6 + c2 + c0;
+  uint32_t csum = c12 + c6 + c2 + c0;
 
   // each mv is max 3900; so max value is 3900*4096*4, fits in unsigned32
-  uint32_t mv_scaled = (mv11 * c11) + (mv6 * c6) + (mv2 * c2) + (mv0 * c0);
+  uint32_t mv_scaled = (mv12 * c12) + (mv6 * c6) + (mv2 * c2) + (mv0 * c0);
   return mv_scaled / (float) (csum * 1000U);
 }
 #endif  // USE_ESP32

esphome/components/adc/adc_sensor.h

@@ -1,19 +1,34 @@
 #pragma once
 
-#include "esphome/core/component.h"
-#include "esphome/core/hal.h"
-#include "esphome/core/defines.h"
 #include "esphome/components/sensor/sensor.h"
 #include "esphome/components/voltage_sampler/voltage_sampler.h"
+#include "esphome/core/component.h"
+#include "esphome/core/defines.h"
+#include "esphome/core/hal.h"
 
 #ifdef USE_ESP32
-#include "driver/adc.h"
 #include <esp_adc_cal.h>
+#include "driver/adc.h"
 #endif
 
 namespace esphome {
 namespace adc {
 
+#ifdef USE_ESP32
+// clang-format off
+#if (ESP_IDF_VERSION_MAJOR == 4 && ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(4, 4, 7)) || \
+    (ESP_IDF_VERSION_MAJOR == 5 && \
+     ((ESP_IDF_VERSION_MINOR == 0 && ESP_IDF_VERSION_PATCH >= 5) || \
+      (ESP_IDF_VERSION_MINOR == 1 && ESP_IDF_VERSION_PATCH >= 3) || \
+      (ESP_IDF_VERSION_MINOR >= 2)) \
+    )
+// clang-format on
+static const adc_atten_t ADC_ATTEN_DB_12_COMPAT = ADC_ATTEN_DB_12;
+#else
+static const adc_atten_t ADC_ATTEN_DB_12_COMPAT = ADC_ATTEN_DB_11;
+#endif
+#endif  // USE_ESP32
+
 class ADCSensor : public sensor::Sensor, public PollingComponent, public voltage_sampler::VoltageSampler {
  public:
 #ifdef USE_ESP32

esphome/components/adc/sensor.py

@@ -1,3 +1,5 @@
+import logging
+
 import esphome.codegen as cg
 import esphome.config_validation as cv
 import esphome.final_validate as fv
@@ -19,16 +21,29 @@
     ATTENUATION_MODES,
     ESP32_VARIANT_ADC1_PIN_TO_CHANNEL,
     ESP32_VARIANT_ADC2_PIN_TO_CHANNEL,
+    adc_ns,
     validate_adc_pin,
 )
 
+_LOGGER = logging.getLogger(__name__)
+
 AUTO_LOAD = ["voltage_sampler"]
 
 
+_attenuation = cv.enum(ATTENUATION_MODES, lower=True)
+
+
 def validate_config(config):
     if config[CONF_RAW] and config.get(CONF_ATTENUATION, None) == "auto":
         raise cv.Invalid("Automatic attenuation cannot be used when raw output is set")
 
+    if config.get(CONF_ATTENUATION) == "11db":
+        _LOGGER.warning(
+            "`attenuation: 11db` is deprecated, use `attenuation: 12db` instead"
+        )
+        # Alter value here so `config` command prints the recommended change
+        config[CONF_ATTENUATION] = _attenuation("12db")
+
     return config
 
 
@@ -47,7 +62,6 @@ def final_validate_config(config):
     return config
 
 
-adc_ns = cg.esphome_ns.namespace("adc")
 ADCSensor = adc_ns.class_(
     "ADCSensor", sensor.Sensor, cg.PollingComponent, voltage_sampler.VoltageSampler
 )
@@ -65,7 +79,7 @@ def final_validate_config(config):
             cv.Required(CONF_PIN): validate_adc_pin,
             cv.Optional(CONF_RAW, default=False): cv.boolean,
             cv.SplitDefault(CONF_ATTENUATION, esp32="0db"): cv.All(
-                cv.only_on_esp32, cv.enum(ATTENUATION_MODES, lower=True)
+                cv.only_on_esp32, _attenuation
             ),
         }
     )