diff --git a/firmware/controllers/actuators/alternator_controller.cpp b/firmware/controllers/actuators/alternator_controller.cpp
index 6181fa6fc1..f3d26b4fc4 100644
--- a/firmware/controllers/actuators/alternator_controller.cpp
+++ b/firmware/controllers/actuators/alternator_controller.cpp
@@ -69,7 +69,7 @@ void AlternatorController::onFastCallback() {
 		alternatorControl.setSimplePwmDutyCycle(0);
 	} else {
 		currentAltDuty = alternatorPid.getOutput(targetVoltage, vBatt.Value, FAST_CALLBACK_PERIOD_MS / 1000.0f);
-    	// huh, why do we bump based on button not based on actual A/C relay state?
+    	// see "idle air Bump for AC" comment
 		int acDutyBump = engine->module<AcController>().unmock().acButtonState ? engineConfiguration->acRelayAlternatorDutyAdder : 0;
 		currentAltDuty += acDutyBump;
 		if (engineConfiguration->isVerboseAlternator) {
diff --git a/firmware/controllers/actuators/idle_thread.cpp b/firmware/controllers/actuators/idle_thread.cpp
index 109ff0941e..d79eaac205 100644
--- a/firmware/controllers/actuators/idle_thread.cpp
+++ b/firmware/controllers/actuators/idle_thread.cpp
@@ -26,8 +26,10 @@
 int IdleController::getTargetRpm(float clt) {
 	targetRpmByClt = interpolate2d(clt, config->cltIdleRpmBins, config->cltIdleRpm);
 
-	// Bump for AC
-	// huh, why do we bump based on button not based on actual A/C relay state?
+	// idle air Bump for AC
+	// Why do we bump based on button not based on actual A/C relay state?
+	// Because AC output has a delay to allow idle bump to happen first, so that the airflow increase gets a head start on the load increase
+	// alternator duty cycle has a similar logic
 	targetRpmAcBump = engine->module<AcController>().unmock().acButtonState ? engineConfiguration->acIdleRpmBump : 0;
 
 	return targetRpmByClt + targetRpmAcBump;