Merge pull request #2012 from Ralim/issue-1946-thermal-runaway-update

Rewrite Thermal Runaway detection

.gitignore

@@ -215,3 +215,4 @@ Logo GUI/TS100 Logo Editor/TS100 Logo Editor/bin/
 
 # Tests/linters/sanitizers
 source/check-style.log
+.ash_history

source/Core/BSP/MHP30/configuration.h

@@ -103,8 +103,8 @@
 #define DETAILED_IDLE      0 // 0: Disable 1: Enable - Default 0
 
 // Due to large thermal mass of the PCB being heated we need to pull this back a bit
-#define THERMAL_RUNAWAY_TIME_SEC 45
+#define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   3
+#define THERMAL_RUNAWAY_TEMP_C   2
 
 #define CUT_OUT_SETTING          0  // default to no cut-off voltage
 #define RECOM_VOL_CELL           33 // Minimum voltage per cell (Recommended 3.3V (33))

source/Core/BSP/Miniware/configuration.h

@@ -109,7 +109,7 @@
 #define DETAILED_IDLE      0 // 0: Disable 1: Enable - Default 0
 
 #define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   10
+#define THERMAL_RUNAWAY_TEMP_C   3
 
 #define CUT_OUT_SETTING          0  // default to no cut-off voltage
 #define RECOM_VOL_CELL           33 // Minimum voltage per cell (Recommended 3.3V (33))

source/Core/BSP/Pinecil/configuration.h

@@ -104,7 +104,7 @@
 #define DETAILED_IDLE      0 // 0: Disable 1: Enable - Default 0
 
 #define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   20
+#define THERMAL_RUNAWAY_TEMP_C   3
 
 #define CUT_OUT_SETTING          0  // default to no cut-off voltage
 #define RECOM_VOL_CELL           33 // Minimum voltage per cell (Recommended 3.3V (33))

source/Core/BSP/Pinecilv2/configuration.h

@@ -105,7 +105,7 @@
 #define DETAILED_IDLE      0 // 0: Disable 1: Enable - Default 0
 
 #define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   20
+#define THERMAL_RUNAWAY_TEMP_C   3
 
 #define CUT_OUT_SETTING          0  // default to no cut-off voltage
 #define RECOM_VOL_CELL           33 // Minimum voltage per cell (Recommended 3.3V (33))

source/Core/BSP/Sequre/configuration.h

@@ -133,7 +133,7 @@
 #define TEMP_uV_LOOKUP_S60
 #define USB_PD_VMAX              12 // Maximum voltage for PD to negotiate
 #define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   10
+#define THERMAL_RUNAWAY_TEMP_C   3
 
 #define HARDWARE_MAX_WATTAGE_X10 600
 
@@ -167,7 +167,7 @@
 #define TEMP_uV_LOOKUP_S60
 #define USB_PD_VMAX              20 // Maximum voltage for PD to negotiate
 #define THERMAL_RUNAWAY_TIME_SEC 20
-#define THERMAL_RUNAWAY_TEMP_C   10
+#define THERMAL_RUNAWAY_TEMP_C   3
 
 #define HARDWARE_MAX_WATTAGE_X10 600
 
@@ -218,8 +218,8 @@
 
 #define TIP_THERMAL_MASS         30 // X10 watts to raise 1 deg C in 1 second
 #define TIP_THERMAL_INERTIA      10 // We use a large inertia value to smooth out the drive to the tip since its stupidly sensitive
-#define THERMAL_RUNAWAY_TIME_SEC 60
+#define THERMAL_RUNAWAY_TIME_SEC 30
-#define THERMAL_RUNAWAY_TEMP_C   3
+#define THERMAL_RUNAWAY_TEMP_C   2
 
 #define COPPER_HEATER_COIL 1  // Have a heater coil that changes resistance on us
 #define TIP_RESISTANCE     52 // PCB heater, measured at ~19C. Will shift by temp a decent amount

source/Core/Threads/PIDThread.cpp

@@ -30,7 +30,7 @@ int32_t                    powerSupplyWattageLimit = 0;
 bool                       heaterThermalRunaway    = false;
 
 static int32_t getPIDResultX10Watts(TemperatureType_t set_point, TemperatureType_t current_value);
-static void    detectThermalRunaway(const TemperatureType_t currentTipTempInC, const TemperatureType_t tError);
+static void    detectThermalRunaway(const TemperatureType_t currentTipTempInC, const uint32_t x10WattsOut);
 static void    setOutputx10WattsViaFilters(int32_t x10Watts);
 static int32_t getX10WattageLimits();
 
@@ -92,8 +92,8 @@ void startPIDTask(void const *argument __unused) {
           PIDTempTarget = TipThermoModel::getTipMaxInC();
         }
 
-        detectThermalRunaway(currentTipTempInC, PIDTempTarget - currentTipTempInC);
         x10WattsOut = getPIDResultX10Watts(PIDTempTarget, currentTipTempInC);
+        detectThermalRunaway(currentTipTempInC, x10WattsOut);
       } else {
         detectThermalRunaway(currentTipTempInC, 0);
       }
@@ -228,31 +228,59 @@ int32_t getPIDResultX10Watts(TemperatureType_t set_point, TemperatureType_t curr
 #endif
 }
 
-void detectThermalRunaway(const TemperatureType_t currentTipTempInC, const TemperatureType_t tError) {
+/*
-  static TemperatureType_t tipTempCRunawayTemp   = 0;
+ * Detection of thermal runaway
-  static TickType_t        runawaylastChangeTime = 0;
+ * The goal of this is to handle cases where something has gone wrong
+ * 1. The tip MOSFET is broken, so power is being constantly applied to the tip
+ * a. This can show as temp being stuck at max
+ * b. Or temp rising when the heater is off
+ * 2. Broken temperature sense
+ * a. Temp is stuck at a value
+ * These boil down to either a constantly rising temperature or a temperature that is stuck at a value
+ * These are both covered; but looking at the eye/delta between min and max temp seen
+ */
+void detectThermalRunaway(const TemperatureType_t currentTipTempInC, const uint32_t x10WattsOut) {
 
-  // Check for thermal runaway, where it has been x seconds with negligible (y) temp rise
+  static TemperatureType_t tiptempMin         = 0xFFFF; // Min tip temp seen
-  // While trying to actively heat
+  static TemperatureType_t tipTempMax         = 0;      // Max tip temp seen while heater is on
+  bool                     thisCycleIsHeating = x10WattsOut > 0;
+  static TickType_t        heatCycleStart     = 0;
 
-  // If we are more than 20C below the setpoint
+  static bool haveSeenDelta = false;
-  if ((tError > THERMAL_RUNAWAY_TEMP_C)) {
 
-    // If we have heated up by more than 20C since last sample point, snapshot time and tip temp
+  // Check for readings being pegged at the top of the ADC while the heater is off
-    TemperatureType_t delta = currentTipTempInC - tipTempCRunawayTemp;
+  if (!thisCycleIsHeating && (getTipRawTemp(0) > (0x7FFF - 16))) {
-    if (delta > THERMAL_RUNAWAY_TEMP_C) {
+    heaterThermalRunaway = true;
-      // We have heated up more than the threshold, reset the timer
+  }
-      tipTempCRunawayTemp   = currentTipTempInC;
+
-      runawaylastChangeTime = xTaskGetTickCount();
+  if (haveSeenDelta) {
-    } else {
+    return;
-      if ((xTaskGetTickCount() - runawaylastChangeTime) > (THERMAL_RUNAWAY_TIME_SEC * TICKS_SECOND)) {
+  }
-        // It has taken too long to rise
+
+  if (currentTipTempInC < tiptempMin) {
+    tiptempMin = currentTipTempInC;
+  }
+  if (thisCycleIsHeating && currentTipTempInC > tipTempMax) {
+    tipTempMax = currentTipTempInC;
+  }
+  if (thisCycleIsHeating) {
+    if (heatCycleStart == 0) {
+      heatCycleStart = xTaskGetTickCount();
+    }
+  } else {
+    heatCycleStart = 0;
+  }
+
+  if ((xTaskGetTickCount() - heatCycleStart) > (THERMAL_RUNAWAY_TIME_SEC * TICKS_SECOND)) {
+    if (tipTempMax > tiptempMin) {
+      // Have been heating for min seconds, check if the delta is large enough
+      TemperatureType_t delta = tipTempMax - tiptempMin;
+      haveSeenDelta           = true;
+
+      if (delta < THERMAL_RUNAWAY_TEMP_C) {
         heaterThermalRunaway = true;
       }
     }
-  } else {
-    tipTempCRunawayTemp   = currentTipTempInC;
-    runawaylastChangeTime = xTaskGetTickCount();
   }
 }
 

source/Core/Threads/UI/logic/Soldering.cpp

@@ -138,9 +138,7 @@ OperatingMode gui_solderingMode(const ButtonState buttons, guiContext *cxt) {
 
   // Draw in the screen details
   if (getSettingValue(SettingsOptions::DetailedSoldering)) {
-
     ui_draw_soldering_power_status(cxt->scratch_state.state2);
-
   } else {
     ui_draw_soldering_basic_status(cxt->scratch_state.state2);
   }