diff --git a/esphome/components/rtttl/__init__.py b/esphome/components/rtttl/__init__.py
index 6163129529..10f1313408 100644
--- a/esphome/components/rtttl/__init__.py
+++ b/esphome/components/rtttl/__init__.py
@@ -12,6 +12,7 @@ from esphome.const import (
     CONF_PLATFORM,
     CONF_TRIGGER_ID,
     CONF_SPEAKER,
+    CONF_GAIN,
 )
 
 _LOGGER = logging.getLogger(__name__)
@@ -38,6 +39,7 @@ CONFIG_SCHEMA = cv.All(
             cv.GenerateID(CONF_ID): cv.declare_id(Rtttl),
             cv.Optional(CONF_OUTPUT): cv.use_id(FloatOutput),
             cv.Optional(CONF_SPEAKER): cv.use_id(Speaker),
+            cv.Optional(CONF_GAIN, default="0.6"): cv.percentage,
             cv.Optional(CONF_ON_FINISHED_PLAYBACK): automation.validate_automation(
                 {
                     cv.GenerateID(CONF_TRIGGER_ID): cv.declare_id(
@@ -98,6 +100,8 @@ async def to_code(config):
         out = await cg.get_variable(config[CONF_SPEAKER])
         cg.add(var.set_speaker(out))
 
+    cg.add(var.set_gain(config[CONF_GAIN]))
+
     for conf in config.get(CONF_ON_FINISHED_PLAYBACK, []):
         trigger = cg.new_Pvariable(conf[CONF_TRIGGER_ID], var)
         await automation.build_automation(trigger, [], conf)
diff --git a/esphome/components/rtttl/rtttl.cpp b/esphome/components/rtttl/rtttl.cpp
index 199a373785..0bdf65b7bd 100644
--- a/esphome/components/rtttl/rtttl.cpp
+++ b/esphome/components/rtttl/rtttl.cpp
@@ -16,7 +16,7 @@ static const uint16_t NOTES[] = {0,    262,  277,  294,  311,  330,  349,  370,
                                  1109, 1175, 1245, 1319, 1397, 1480, 1568, 1661, 1760, 1865, 1976, 2093, 2217,
                                  2349, 2489, 2637, 2794, 2960, 3136, 3322, 3520, 3729, 3951};
 
-static const uint16_t I2S_SPEED = 1600;
+static const uint16_t I2S_SPEED = 1000;
 
 #undef HALF_PI
 static const double HALF_PI = 1.5707963267948966192313216916398;
@@ -136,7 +136,7 @@ void Rtttl::loop() {
         if (this->samples_per_wave_ != 0 && this->samples_sent_ >= this->samples_gap_) {  // Play note//
           rem = ((this->samples_sent_ << 10) % this->samples_per_wave_) * (360.0 / this->samples_per_wave_);
 
-          int16_t val = 8192 * sin(deg2rad(rem));
+          int16_t val = (49152 * this->gain_) * sin(deg2rad(rem));
 
           sample[x].left = val;
           sample[x].right = val;
@@ -269,7 +269,7 @@ void Rtttl::loop() {
     }
     if (this->output_freq_ != 0) {
       this->output_->update_frequency(this->output_freq_);
-      this->output_->set_level(0.5);
+      this->output_->set_level(this->gain_);
     } else {
       this->output_->set_level(0.0);
     }
@@ -278,18 +278,23 @@ void Rtttl::loop() {
 #ifdef USE_SPEAKER
   if (this->speaker_ != nullptr) {
     this->samples_sent_ = 0;
-    this->samples_count_ = (this->sample_rate_ * this->note_duration_) / I2S_SPEED;
-    // Convert from frequency in Hz to high and low samples in fixed point
+    this->samples_gap_ = 0;
+    this->samples_per_wave_ = 0;
+    this->samples_count_ = (this->sample_rate_ * this->note_duration_) / 1600;  //(ms);
+    if (need_note_gap) {
+      this->samples_gap_ = (this->sample_rate_ * DOUBLE_NOTE_GAP_MS) / 1600;  //(ms);
+    }
     if (this->output_freq_ != 0) {
       this->samples_per_wave_ = (this->sample_rate_ << 10) / this->output_freq_;
-    } else {
-      this->samples_per_wave_ = 0;
-    }
-    if (need_note_gap) {
-      this->samples_gap_ = (this->sample_rate_ * DOUBLE_NOTE_GAP_MS) / I2S_SPEED;
-    } else {
-      this->samples_gap_ = 0;
+
+      // make sure there is enough samples to add a full last sinus.
+      uint16_t division = ((this->samples_count_ << 10) / this->samples_per_wave_) + 1;
+      uint16_t x = this->samples_count_;
+      this->samples_count_ = (division * this->samples_per_wave_);
+      ESP_LOGD(TAG, "play time old: %d div: %d new: %d %d", x, division, this->samples_count_, this->samples_per_wave_);
+      this->samples_count_ = this->samples_count_ >> 10;
     }
+    // Convert from frequency in Hz to high and low samples in fixed point
   }
 #endif
 
diff --git a/esphome/components/rtttl/rtttl.h b/esphome/components/rtttl/rtttl.h
index e09b0265be..bf089ce980 100644
--- a/esphome/components/rtttl/rtttl.h
+++ b/esphome/components/rtttl/rtttl.h
@@ -15,7 +15,7 @@ namespace esphome {
 namespace rtttl {
 
 #ifdef USE_SPEAKER
-static const size_t SAMPLE_BUFFER_SIZE = 256;
+static const size_t SAMPLE_BUFFER_SIZE = 512;
 
 struct SpeakerSample {
   int16_t left{0};
@@ -31,6 +31,13 @@ class Rtttl : public Component {
 #ifdef USE_SPEAKER
   void set_speaker(speaker::Speaker *speaker) { this->speaker_ = speaker; }
 #endif
+  void set_gain(float gain) {
+    if (gain < 0.1f)
+      gain = 0.1f;
+    if (gain > 1.0f)
+      gain = 1.0f;
+    this->gain_ = gain;
+  }
   void play(std::string rtttl);
   void stop();
   void dump_config() override;
@@ -60,6 +67,7 @@ class Rtttl : public Component {
   uint16_t note_duration_;
 
   uint32_t output_freq_;
+  float gain_{0.6f};
 
 #ifdef USE_OUTPUT
   output::FloatOutput *output_;
@@ -68,13 +76,13 @@ class Rtttl : public Component {
   void play_output_();
 
 #ifdef USE_SPEAKER
-  speaker::Speaker *speaker_;
-  void play_speaker_();
+  speaker::Speaker *speaker_{nullptr};
   int sample_rate_{16000};
   int samples_per_wave_{0};
   int samples_sent_{0};
   int samples_count_{0};
   int samples_gap_{0};
+
 #endif
 
   CallbackManager<void()> on_finished_playback_callback_;