diff --git a/port/boards/labplus_for_xuejing/bsp_audio_board.c b/port/boards/labplus_for_xuejing/bsp_audio_board.c
index 67dc2371..71d92039 100644
--- a/port/boards/labplus_for_xuejing/bsp_audio_board.c
+++ b/port/boards/labplus_for_xuejing/bsp_audio_board.c
@@ -53,6 +53,11 @@ static const audio_codec_ctrl_if_t *codec_ctrl_if = NULL;
 static const audio_codec_if_t *codec_if = NULL;
 static esp_codec_dev_handle_t codec_dev = NULL;
 
+static int s_play_sample_rate = 16000;
+static int s_play_channel_format = 1;
+static int s_bits_per_chan = 16;
+
+
 static esp_err_t bsp_i2s_init(uint32_t sample_rate, int channel_format, int bits_per_sample)
 {
     esp_err_t ret_val = ESP_OK;
@@ -274,6 +279,63 @@ esp_err_t bsp_audio_play(const int8_t  *data, int length, TickType_t ticks_to_wa
 
     return ret;
 }
+esp_err_t bsp_audio_play2(const int16_t* data, int length, TickType_t ticks_to_wait)
+{
+    size_t bytes_write = 0;
+    esp_err_t ret = ESP_OK;
+    if (!codec_dev) {
+        return ESP_FAIL;
+    }
+
+    int out_length= length;
+    int audio_time = 1;
+    audio_time *= (16000 / s_play_sample_rate);
+    audio_time *= (2 / s_play_channel_format);
+
+    int *data_out = NULL;
+    if (s_bits_per_chan != 32) {
+        out_length = length * 2;
+        data_out = malloc(out_length);
+        for (int i = 0; i < length / sizeof(int16_t); i++) {
+            int ret = data[i];
+            data_out[i] = ret << 16;
+        }
+    }
+
+    int *data_out_1 = NULL;
+    if (s_play_channel_format != 2 || s_play_sample_rate != 16000) {
+        out_length *= audio_time;
+        data_out_1 = malloc(out_length);
+        int *tmp_data = NULL;
+        if (data_out != NULL) {
+            tmp_data = data_out;
+        } else {
+            tmp_data = data;
+        }
+
+        for (int i = 0; i < out_length / (audio_time * sizeof(int)); i++) {
+            for (int j = 0; j < audio_time; j++) {
+                data_out_1[audio_time * i + j] = tmp_data[i];
+            }
+        }
+        if (data_out != NULL) {
+            free(data_out);
+            data_out = NULL;
+        }
+    }
+
+    if (data_out != NULL) {
+        ret = esp_codec_dev_write(codec_dev, (void *)data_out, out_length);
+        free(data_out);
+    } else if (data_out_1 != NULL) {
+        ret = esp_codec_dev_write(codec_dev, (void *)data_out_1, out_length);
+        free(data_out_1);
+    } else {
+        ret = esp_codec_dev_write(codec_dev, (void *)data, length);
+    }
+
+    return ret;
+}
 
 /* 做为afe is_get_raw_channel = true :16K 16bit 左声道 + 右声道，无ref, 
           is_get_raw_channel = false :16K 16bit 左声道 + 右声道 + 一路ref,*/
diff --git a/port/boards/labplus_for_xuejing/mpconfigboard.cmake b/port/boards/labplus_for_xuejing/mpconfigboard.cmake
index a05e46c9..3f50d884 100644
--- a/port/boards/labplus_for_xuejing/mpconfigboard.cmake
+++ b/port/boards/labplus_for_xuejing/mpconfigboard.cmake
@@ -42,6 +42,7 @@ set(MICROPY_SOURCE_BOARD
     ${MPY_PORT_DIR}/drivers/startup/oled.c
     ${MPY_PORT_DIR}/drivers/startup/startup.c 
     ${MPY_PORT_DIR}/drivers/rfid/mfrc522.c
+    ${MPY_PORT_DIR}/drivers/tts/tts_pcm.c
     ${MPY_PORT_DIR}/lib/utils/pyexec.c
     ${MPY_PORT_DIR}/builtins/modmusictunes.c
     ${MPY_PORT_DIR}/builtins/modmusic.c
@@ -51,6 +52,7 @@ set(MICROPY_SOURCE_BOARD
     ${MPY_PORT_DIR}/builtins/modframebuf.c
     ${MPY_PORT_DIR}/builtins/modaudio.c
     ${MPY_PORT_DIR}/builtins/modrfid.c
+    ${MPY_PORT_DIR}/builtins/modtest.c
     # ${MPY_PORT_DIR}/builtins/mod_audio/audio_player.c
     # ${MPY_PORT_DIR}/builtins/mod_audio/audio_recorder.c 
     # ${MPY_PORT_DIR}/builtins/mod_audio/vfs_stream.c 
@@ -60,6 +62,7 @@ set(MICROPY_SOURCE_BOARD
 set(MICROPY_SOURCE_BOARD_DIR
     ${MPY_PORT_DIR}/drivers
     ${MPY_PORT_DIR}/drivers/audio/include
+    ${MPY_PORT_DIR}/drivers/tts
     ${MPY_PORT_DIR}/lib
     ${MPY_PORT_DIR}/builtins
     ${MPY_PORT_DIR}/boards
diff --git a/port/boards/labplus_for_xuejing/partitions-8MiB.csv b/port/boards/labplus_for_xuejing/partitions-8MiB.csv
index 3c955e12..4a8942d4 100644
--- a/port/boards/labplus_for_xuejing/partitions-8MiB.csv
+++ b/port/boards/labplus_for_xuejing/partitions-8MiB.csv
@@ -1,10 +1,9 @@
 # Notes: the offset of the partition table itself is set in
 # $IDF_PATH/components/partition_table/Kconfig.projbuild.
-# Name,   Type, SubType,   Offset,  Size, Flags
-nvs,      data, nvs,       0x9000,  0x6000,
-phy_init, data, phy,       0xf000,  0x1000,
-factory,  app,  factory,   0x10000, 0x200000,
-vfs,      data, spiffs,    0x210000, 0x1f0000,
-font,     data, nvs,      0x400000, 0x200000,
-user_nvs, data, nvs,       0x7fc000,  0x4000,
-
+# Name,   Type, SubType, Offset,  Size, Flags
+nvs,      data, nvs,     0x9000,  0x6000,
+phy_init, data, phy,     0xf000,  0x1000,
+factory,  app,  factory, 0x10000, 0x2f0000,
+vfs,      data, fat,     0x300000, 0x100000,
+font, 	  data, fat,   0x400000, 0x3fc000,
+user_nvs, data, nvs,     0x7fc000,  0x4000,
diff --git a/port/builtins/modaudio.c b/port/builtins/modaudio.c
index 8be6a369..61be6b44 100755
--- a/port/builtins/modaudio.c
+++ b/port/builtins/modaudio.c
@@ -135,6 +135,7 @@ static MP_DEFINE_CONST_FUN_OBJ_0(audio_loudness_obj, audio_loudness);
 
 static mp_obj_t audio_record(size_t n_args, const mp_obj_t *pos_args, mp_map_t *kw_args)
 {
+    audio_init();
     enum { ARG_file_name, ARG_record_time, ARG_bits_per_sample, ARG_channels, ARG_sampleRate};
     static const mp_arg_t allowed_args[] = {
         { MP_QSTR_file_name,    MP_ARG_REQUIRED | MP_ARG_OBJ },
@@ -193,4 +194,4 @@ const mp_obj_module_t mp_module_audio = {
 };
 
 MP_REGISTER_MODULE(MP_QSTR_audio, mp_module_audio);
-// #endif //#if MICROPY_PY_AUDIO
\ No newline at end of file
+// #endif //#if MICROPY_PY_AUDIO
diff --git a/port/builtins/modtest.c b/port/builtins/modtest.c
new file mode 100644
index 00000000..aaf41bb8
--- /dev/null
+++ b/port/builtins/modtest.c
@@ -0,0 +1,39 @@
+#include "py/obj.h"
+#include "py/runtime.h"
+#include "esp_tts.h"
+#include "tts_pcm.h"
+
+#ifndef STATIC
+#define STATIC static
+#endif
+
+// MicroPython 绑定函数
+STATIC mp_obj_t mp_tts_generate(mp_obj_t text_obj) {
+    // 获取文本参数
+    const char* text = mp_obj_str_get_str(text_obj);
+
+    // 直接调用 text_to_speech 播放音频
+    text_to_speech(text);
+
+    return mp_const_none;  // 播放完成后返回 None
+}
+
+STATIC MP_DEFINE_CONST_FUN_OBJ_1(mp_tts_generate_obj, mp_tts_generate);
+
+
+// 模块定义
+STATIC const mp_rom_map_elem_t tts_module_globals_table[] = {
+    { MP_ROM_QSTR(MP_QSTR___name__), MP_ROM_QSTR(MP_QSTR_tts) },
+    { MP_ROM_QSTR(MP_QSTR_generate), MP_ROM_PTR(&mp_tts_generate_obj) },
+};
+STATIC MP_DEFINE_CONST_DICT(tts_module_globals, tts_module_globals_table);
+
+// 模块结构体
+const mp_obj_module_t tts_module = {
+    .base = { &mp_type_module },
+    .globals = (mp_obj_dict_t*)&tts_module_globals,
+};
+
+// 注册模块
+MP_REGISTER_MODULE(MP_QSTR_tts, tts_module);
+
diff --git a/port/builtins/modtest.c.bk b/port/builtins/modtest.c.bk
new file mode 100644
index 00000000..7d0c7b84
--- /dev/null
+++ b/port/builtins/modtest.c.bk
@@ -0,0 +1,44 @@
+#include "py/obj.h"
+#include "py/runtime.h"
+#include "esp_tts.h"
+#include "tts_pcm.h"
+
+#ifndef STATIC
+#define STATIC static
+#endif
+
+// MicroPython绑定函数
+STATIC mp_obj_t mp_tts_generate(mp_obj_t text_obj) {
+    // 获取参数
+    const char* text = mp_obj_str_get_str(text_obj);
+    
+    // 生成 PCM 数据
+    int pcm_len = 0;
+    short* pcm_data = text_to_pcm(text, &pcm_len);
+
+    if (!pcm_data || pcm_len == 0) {
+        return mp_const_none;  // 失败返回 None
+    }
+
+    // 返回 PCM 数据为字节数组
+    return mp_obj_new_bytearray_by_ref(pcm_len * sizeof(short), pcm_data);
+}
+
+STATIC MP_DEFINE_CONST_FUN_OBJ_1(mp_tts_generate_obj, mp_tts_generate);
+
+// 模块定义
+STATIC const mp_rom_map_elem_t tts_module_globals_table[] = {
+    { MP_ROM_QSTR(MP_QSTR___name__), MP_ROM_QSTR(MP_QSTR_tts) },
+    { MP_ROM_QSTR(MP_QSTR_generate), MP_ROM_PTR(&mp_tts_generate_obj) },
+};
+STATIC MP_DEFINE_CONST_DICT(tts_module_globals, tts_module_globals_table);
+
+// 模块结构体
+const mp_obj_module_t tts_module = {
+    .base = { &mp_type_module },
+    .globals = (mp_obj_dict_t*)&tts_module_globals,
+};
+
+// 注册模块
+MP_REGISTER_MODULE(MP_QSTR_tts, tts_module);
+
diff --git a/port/drivers/audio/include/bsp_audio.h b/port/drivers/audio/include/bsp_audio.h
index 916c72af..5483936a 100644
--- a/port/drivers/audio/include/bsp_audio.h
+++ b/port/drivers/audio/include/bsp_audio.h
@@ -67,6 +67,7 @@ esp_err_t bsp_codec_dev_open(int sample_rate, int channel_format, int bits_per_s
 esp_err_t bsp_codec_dev_close(void);
 
 esp_err_t bsp_audio_play(const int8_t *data, int length, TickType_t ticks_to_wait);
+esp_err_t bsp_audio_play2(const int16_t *data, int length, TickType_t ticks_to_wait);
 
 /**
  * @brief Get the record pcm data.
diff --git a/port/drivers/audio/include/recorder.h b/port/drivers/audio/include/recorder.h
index 4eb59c52..024398dc 100644
--- a/port/drivers/audio/include/recorder.h
+++ b/port/drivers/audio/include/recorder.h
@@ -1,6 +1,7 @@
 #ifndef __RECORDER_
 #define __RECORDER_
 #include "freertos/FreeRTOS.h"
+#include "freertos/semphr.h"
 #include "freertos/event_groups.h"
 #include "ringbuf.h"
 #include "py/obj.h"
@@ -27,5 +28,6 @@ typedef struct
 void recorder_record(const char *filename, wav_fmt_t fmt, int time);
 
 extern recorder_handle_t *recorder;
+extern SemaphoreHandle_t xBinarySemaphore;
 
 #endif
diff --git a/port/drivers/audio/include/wav_codec.h b/port/drivers/audio/include/wav_codec.h
index 72ef9281..fc77cdef 100644
--- a/port/drivers/audio/include/wav_codec.h
+++ b/port/drivers/audio/include/wav_codec.h
@@ -23,6 +23,9 @@
 #include "py/obj.h"
 #include "freertos/ringbuf.h"
 #include "wave_head.h"
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+
 
 #ifdef __cplusplus
 extern "C" {
diff --git a/port/drivers/audio/recorder.c b/port/drivers/audio/recorder.c
index 1704bbe7..60002434 100755
--- a/port/drivers/audio/recorder.c
+++ b/port/drivers/audio/recorder.c
@@ -11,6 +11,8 @@
 #include <ctype.h>
 #include <math.h>
 #include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "freertos/semphr.h"
 #include "esp_system.h"
 #include "esp_log.h"
 #include "py/stream.h"
@@ -23,6 +25,7 @@
 #include "player.h"
 #include "msg.h"
 
+
 // #define TAG "recorder"
 
 recorder_handle_t *recorder = NULL;
@@ -65,7 +68,10 @@ void recorder_record(const char *filename, wav_fmt_t fmt, int time)
     recorder->total_frames = recorder->time * (recorder->wav_fmt.sampleRate * recorder->wav_fmt.channels * recorder->wav_fmt.bits_per_sample / 8) / READ_RINGBUF_BLOCK_SIZE;
     // ESP_LOGE(TAG, "record total frame: %d, time: %d", recorder->total_frames, recorder->time);
     recorder->file_uri = filename;
+    xBinarySemaphore = xSemaphoreCreateBinary();
     xTaskCreatePinnedToCore(&stream_i2s_read_task, "stream_i2s_read_task", 4 * 1024, (void*)recorder, 8, NULL, CORE_NUM1);
+    xSemaphoreTake(xBinarySemaphore, portMAX_DELAY);
+    vSemaphoreDelete(xBinarySemaphore);
 }
 
 void record_deinit(void)
diff --git a/port/drivers/audio/wav_codec.c b/port/drivers/audio/wav_codec.c
index bf883f16..6a919be6 100644
--- a/port/drivers/audio/wav_codec.c
+++ b/port/drivers/audio/wav_codec.c
@@ -19,6 +19,7 @@
 #include "esp_log.h"
 #include "freertos/FreeRTOS.h"
 #include "freertos/task.h"
+#include "freertos/semphr.h"
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -29,6 +30,8 @@
 #include "player.h"
 #include "recorder.h"
 
+SemaphoreHandle_t xBinarySemaphore;  // 定义信号量
+
 static const char *TAG = "wav_codec";
 
 wav_codec_t *wav_file_open(const char *path_in, int mode)
@@ -99,7 +102,7 @@ void wav_file_read_task(void *arg)
 		if(!wav_codec){ break; }
 		player->wav_codec = wav_codec;
 
-		if(!wav_codec_ringbuff){
+		if(!wav_codec_ringbuff){/
 			wav_codec_ringbuff = xRingbufferCreate(STREAM_OUT_RINGBUF_SIZE, RINGBUF_TYPE_BYTEBUF);
 			if(!wav_codec_ringbuff){ break; }
 		}else{
@@ -291,5 +294,6 @@ void stream_i2s_read_task(void *arg)
         free(stream_buff);
     }
     // ESP_LOGE(TAG, "stream in task end, RAM left: %ld", esp_get_free_heap_size());
+    xSemaphoreGive(xBinarySemaphore);
     vTaskDelete(NULL);
 }
diff --git a/port/drivers/tts/tts_pcm.c b/port/drivers/tts/tts_pcm.c
new file mode 100644
index 00000000..0eec16f8
--- /dev/null
+++ b/port/drivers/tts/tts_pcm.c
@@ -0,0 +1,92 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "esp_system.h"
+#include "freertos/semphr.h"
+#include "esp_tts.h"
+#include "esp_tts_voice_xiaole.h"
+#include "esp_tts_voice_template.h"
+#include "esp_tts_player.h"
+#include "ringbuf.h"
+
+#include "esp_partition.h"
+#include "esp_idf_version.h"
+#include "bsp_audio.h"
+
+typedef struct {
+    char text[256];
+    SemaphoreHandle_t done_semaphore;
+} tts_task_param_t;
+
+void tts_task(void* param) {
+    tts_task_param_t* task_param = (tts_task_param_t*)param;
+    const esp_partition_t* part = esp_partition_find_first(ESP_PARTITION_TYPE_DATA, ESP_PARTITION_SUBTYPE_ANY, "font");
+    if (part == NULL) {
+        printf("Couldn't find voice data partition!\n");
+        vTaskDelete(NULL);
+        return;
+    }
+
+    const void* voicedata;
+    esp_partition_mmap_handle_t mmap;
+    esp_err_t err = esp_partition_mmap(part, 0, part->size, ESP_PARTITION_MMAP_DATA, (const void**)&voicedata, &mmap);
+    if (err != ESP_OK) {
+        printf("Couldn't map voice data partition!\n");
+        vTaskDelete(NULL);
+        return;
+    }
+
+    esp_tts_voice_t* voice = esp_tts_voice_set_init(&esp_tts_voice_template, (int16_t*)voicedata);
+    if (voice == NULL) {
+        printf("TTS voice init failed!\n");
+        return;
+    }
+    esp_tts_handle_t* tts_handle = esp_tts_create(voice);
+    if (!tts_handle || !task_param->text) {
+        printf("Invalid TTS handle or text!\n");
+        return;
+    }
+
+    if (esp_tts_parse_chinese(tts_handle, task_param->text)) {
+	int len[1]={0};
+	bsp_codec_dev_delete();
+	bsp_codec_dev_create();
+	bsp_audio_set_play_vol(50);
+	bsp_codec_dev_open(16000,2,32);
+        do {
+            short* pcm = esp_tts_stream_play(tts_handle, len, 2);
+            bsp_audio_play2(pcm, len[0]*2, portMAX_DELAY);
+           // vTaskDelay(pdMS_TO_TICKS(10));
+        } while (len[0] > 0);
+	bsp_codec_dev_close();
+	bsp_codec_dev_delete();
+    }
+    
+    esp_tts_stream_reset(tts_handle);
+    xSemaphoreGive(task_param->done_semaphore);
+    vTaskDelete(NULL);
+}
+
+void text_to_speech(const char* text) {
+    static SemaphoreHandle_t done_semaphore = NULL;
+    if (done_semaphore == NULL) {
+        done_semaphore = xSemaphoreCreateBinary();
+    }
+
+    tts_task_param_t* task_param = malloc(sizeof(tts_task_param_t));
+    if (!task_param) {
+        printf("Failed to allocate memory for task parameters\n");
+        return;
+    }
+
+    strncpy(task_param->text, text, sizeof(task_param->text) - 1);
+    task_param->text[sizeof(task_param->text) - 1] = '\0';
+    task_param->done_semaphore = done_semaphore;
+
+    xTaskCreatePinnedToCore(tts_task, "tts_task", 8192, task_param, 5, NULL, 0);
+    xSemaphoreTake(done_semaphore, portMAX_DELAY);
+    free(task_param);
+}
diff --git a/port/drivers/tts/tts_pcm.c.bk b/port/drivers/tts/tts_pcm.c.bk
new file mode 100644
index 00000000..c7e2538a
--- /dev/null
+++ b/port/drivers/tts/tts_pcm.c.bk
@@ -0,0 +1,116 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "esp_system.h"
+#include "freertos/semphr.h"
+#include "esp_tts.h"
+#include "esp_tts_voice_xiaole.h"
+#include "esp_tts_voice_template.h"
+#include "esp_tts_player.h"
+#include "ringbuf.h"
+
+#include "esp_partition.h"
+#include "esp_idf_version.h"
+
+typedef struct {
+    char text[256];   // 存储待转换的文本
+    short* pcm_data;  // 存储 PCM 数据
+    SemaphoreHandle_t done_semaphore;  // 任务完成信号量
+} tts_task_param_t;
+
+static short* pcm_buffer = NULL;
+static int pcm_data_len = 0;
+
+void tts_task(void* param) {
+    tts_task_param_t* task_param = (tts_task_param_t*)param;
+    
+    const esp_partition_t* part = esp_partition_find_first(ESP_PARTITION_TYPE_DATA, ESP_PARTITION_SUBTYPE_ANY, "font");
+    if (part == NULL) {
+        printf("Couldn't find voice data partition!\n");
+        vTaskDelete(NULL);
+        return;
+    }
+
+    const void* voicedata;
+#if ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(5, 0, 0)
+    esp_partition_mmap_handle_t mmap;
+    esp_err_t err = esp_partition_mmap(part, 0, part->size, ESP_PARTITION_MMAP_DATA, (const void**)&voicedata, &mmap);
+#else
+    spi_flash_mmap_handle_t mmap;
+    esp_err_t err = esp_partition_mmap(part, 0, part->size, SPI_FLASH_MMAP_DATA, &voicedata, &mmap);
+#endif
+    if (err != ESP_OK) {
+        printf("Couldn't map voice data partition!\n");
+        vTaskDelete(NULL);
+        return;
+    }
+
+    esp_tts_voice_t* voice = esp_tts_voice_set_init(&esp_tts_voice_template, (int16_t*)voicedata);
+    if (voice == NULL) {
+        printf("TTS voice init failed!\n");
+        return;
+    }
+    esp_tts_handle_t* tts_handle = esp_tts_create(voice);
+    if (!tts_handle || !task_param || !task_param->text) {
+        printf("Invalid TTS handle or text!\n");
+        return;
+    }
+
+    pcm_data_len = 0;
+    if (!pcm_buffer) {
+        pcm_buffer = malloc(16000 * sizeof(short));
+    }
+    if (!pcm_buffer) {
+        printf("Failed to allocate PCM buffer\n");
+        vTaskDelete(NULL);
+        return;
+    }
+    printf("Parsing text: %s\n", task_param->text);
+    printf("TTS handle: %p\n", tts_handle);
+    if (esp_tts_parse_chinese(tts_handle, task_param->text)) {
+        int len;
+        int total_len = 0;
+        printf("Start creat voice\n");
+        do {
+            short* pcm = esp_tts_stream_play(tts_handle, &len, 3);
+            if (len > 0) {
+                //if (total_len + len > 16000) break;  // 防止缓冲区溢出
+                memcpy(pcm_buffer + total_len, pcm, len * sizeof(short));
+                total_len += len;
+            }
+            vTaskDelay(pdMS_TO_TICKS(10));  // 让出 CPU，避免 WDT 复位
+        } while (len > 0);
+        pcm_data_len = total_len;
+    }
+
+    esp_tts_stream_reset(tts_handle);
+
+    // 释放任务
+    xSemaphoreGive(task_param->done_semaphore);
+    vTaskDelete(NULL);
+}
+
+// 供外部调用的函数
+short* text_to_pcm(const char* text, int* out_len) {
+    static SemaphoreHandle_t done_semaphore = NULL;
+    if (done_semaphore == NULL) {
+        done_semaphore = xSemaphoreCreateBinary();
+    }
+
+    tts_task_param_t task_param;
+    strncpy(task_param.text, text, sizeof(task_param.text) - 1);
+    task_param.text[sizeof(task_param.text) - 1] = '\0';
+    task_param.pcm_data = NULL;
+    task_param.done_semaphore = done_semaphore;
+
+    // 启动任务
+    xTaskCreatePinnedToCore(tts_task, "tts_task", 8192, &task_param, 5, NULL, 0);
+
+    // 等待任务完成
+    xSemaphoreTake(done_semaphore, portMAX_DELAY);
+    *out_len = pcm_data_len;
+    return pcm_buffer;
+}
diff --git a/port/drivers/tts/tts_pcm.h b/port/drivers/tts/tts_pcm.h
new file mode 100644
index 00000000..2d0899b0
--- /dev/null
+++ b/port/drivers/tts/tts_pcm.h
@@ -0,0 +1,9 @@
+#pragma once
+#ifdef __cplusplus
+extern "C" {
+#endif
+short* text_to_speech(const char* text);
+#ifdef __cplusplus
+}
+#endif
+
diff --git a/port/drivers/tts/tts_pcm.h.bk b/port/drivers/tts/tts_pcm.h.bk
new file mode 100644
index 00000000..ffc29e7e
--- /dev/null
+++ b/port/drivers/tts/tts_pcm.h.bk
@@ -0,0 +1,9 @@
+#pragma once
+#ifdef __cplusplus
+extern "C" {
+#endif
+short* text_to_pcm(const char* text, int* out_len);
+#ifdef __cplusplus
+}
+#endif
+
diff --git a/port/makeimg.py b/port/makeimg.py
index 3d8e0a84..5e2b9bd8 100755
--- a/port/makeimg.py
+++ b/port/makeimg.py
@@ -104,7 +104,8 @@ def load_partition_table(filename):
 arg_application_bin = "micropython.bin"
 arg_output_bin = bin_name
 arg_output_uf2= uf2_name
-arg_font_bin = "../Noto_Sans_CJK_SC_Light16.bin"
+# arg_font_bin = "../Noto_Sans_CJK_SC_Light16.bin"
+arg_font_bin = "../components/esp-sr/esp-tts/esp_tts_chinese/esp_tts_voice_data_xiaoxin.dat"
 
 # Load required sdkconfig values.
 idf_target = load_sdkconfig_str_value(arg_sdkconfig, "IDF_TARGET", "").upper()