I recently received an ESP32 Lyrat board but I’m pulling out my hair trying to get it to work with openwakeword in ESPHome.
The minute the device boots, the logs are filled with this section repeated multiple times. Any idea what’s going on? It looks like no wake word is being detected so it ends the pipeline before I can say anything to trigger the wake word.
[19:25:49][E][voice_assistant:646]: Error: no_wake_word - No wake word detected
[19:25:49][D][voice_assistant:512]: Signaling stop...
[19:25:49][D][voice_assistant:412]: State changed from STARTING_PIPELINE to STOP_MICROPHONE
[19:25:49][D][voice_assistant:418]: Desired state set to IDLE
[19:25:49][D][voice_assistant:412]: State changed from STOP_MICROPHONE to IDLE
[19:25:49][D][voice_assistant:519]: Event Type: 2
[19:25:49][D][voice_assistant:609]: Assist Pipeline ended
[19:25:49][D][voice_assistant:412]: State changed from IDLE to START_PIPELINE
[19:25:49][D][voice_assistant:418]: Desired state set to START_MICROPHONE
[19:25:49][D][voice_assistant:512]: Signaling stop...
[19:25:49][D][voice_assistant:118]: microphone not running
[19:25:49][D][voice_assistant:200]: Requesting start...
[19:25:49][D][voice_assistant:412]: State changed from START_PIPELINE to STARTING_PIPELINE
[19:25:49][D][voice_assistant:519]: Event Type: 1
[19:25:49][D][voice_assistant:522]: Assist Pipeline running
[19:25:49][D][voice_assistant:118]: microphone not running
[19:25:49][D][voice_assistant:519]: Event Type: 9
[19:25:49][D][voice_assistant:118]: microphone not running
[19:25:49][D][voice_assistant:519]: Event Type: 0
[19:25:49][E][voice_assistant:646]: Error: no_wake_word - No wake word detected
[19:25:49][D][voice_assistant:512]: Signaling stop...
[19:25:49][D][voice_assistant:412]: State changed from STARTING_PIPELINE to STOP_MICROPHONE
[19:25:49][D][voice_assistant:418]: Desired state set to IDLE
[19:25:49][D][voice_assistant:412]: State changed from STOP_MICROPHONE to IDLE
[19:25:49][D][voice_assistant:519]: Event Type: 2
[19:25:49][D][voice_assistant:609]: Assist Pipeline ended
[19:25:49][D][voice_assistant:412]: State changed from IDLE to START_PIPELINE
[19:25:49][D][voice_assistant:418]: Desired state set to START_MICROPHONE
[19:25:49][D][switch:016]: 'Use wake word' Turning OFF.
[19:25:49][D][switch:055]: 'Use wake word': Sending state OFF
[19:25:49][D][voice_assistant:412]: State changed from START_PIPELINE to STOP_MICROPHONE
[19:25:49][D][voice_assistant:418]: Desired state set to IDLE
[19:25:49][D][voice_assistant:412]: State changed from STOP_MICROPHONE to IDLE
[19:25:49][D][voice_assistant:512]: Signaling stop...
[19:25:49][D][voice_assistant:519]: Event Type: 1
[19:25:49][D][voice_assistant:522]: Assist Pipeline running
[19:25:49][D][voice_assistant:519]: Event Type: 9
[19:25:49][D][voice_assistant:519]: Event Type: 0
[19:25:49][E][voice_assistant:646]: Error: no_wake_word - No wake word detected
[23:48:13][D][main:262]: An automation to perform when the assist pipeline is started.
[23:48:13][D][voice_assistant:519]: Event Type: 9
[23:48:18][D][voice_assistant:519]: Event Type: 0
[23:48:18][D][voice_assistant:519]: Event Type: 2
[23:48:18][D][voice_assistant:609]: Assist Pipeline ended
[23:48:18][D][voice_assistant:412]: State changed from STREAMING_MICROPHONE to IDLE
[23:48:18][D][voice_assistant:418]: Desired state set to IDLE
[23:48:18][D][voice_assistant:412]: State changed from IDLE to START_PIPELINE
[23:48:18][D][voice_assistant:418]: Desired state set to START_MICROPHONE
I followed your blog post for the device, and having the same results as mentioned. The logs are spammed with ‘No Wake Word’ detected and I cant seem to get any audio out of it when I do a TTS event.
Its the Lyra 4.3 board (confirmed it looks like yours)
Any hints?
[18:32:29][D][voice_assistant:124]: microphone not running
[18:32:29][I][app:102]: ESPHome version 2023.11.6 compiled on Feb 22 2024, 18:20:54
[18:32:29][C][wifi:559]: WiFi:
[18:32:29][C][wifi:391]: Local MAC: 94:B9:7E:65:C2:B8
[18:32:29][C][wifi:396]: SSID: [redacted]
[18:32:29][C][wifi:397]: IP Address: 192.168.10.120
[18:32:29][C][wifi:399]: BSSID: [redacted]
[18:32:29][C][wifi:400]: Hostname: 'voice-speaker'
[18:32:29][C][wifi:402]: Signal strength: -62 dB ▂▄▆█
[18:32:29][C][wifi:406]: Channel: 4
[18:32:29][C][wifi:407]: Subnet: 255.255.255.0
[18:32:29][C][wifi:408]: Gateway: 192.168.10.254
[18:32:29][C][wifi:409]: DNS1: 192.168.10.254
[18:32:29][C][wifi:410]: DNS2: 90.207.238.97
[18:32:29][D][voice_assistant:529]: Event Type: 9
[18:32:29][D][voice_assistant:124]: microphone not running
[18:32:29][C][logger:416]: Logger:
[18:32:29][C][logger:417]: Level: DEBUG
[18:32:29][C][logger:418]: Log Baud Rate: 115200
[18:32:29][C][logger:420]: Hardware UART: UART0
[18:32:29][D][voice_assistant:529]: Event Type: 0
[18:32:29][E][voice_assistant:656]: Error: no_wake_word - No wake word detected
[18:32:29][D][voice_assistant:522]: Signaling stop...
[18:32:29][D][voice_assistant:422]: State changed from STARTING_PIPELINE to STOP_MICROPHONE
[18:32:29][D][voice_assistant:428]: Desired state set to IDLE
[18:32:30][D][voice_assistant:422]: State changed from STOP_MICROPHONE to IDLE
[18:32:30][C][i2c.arduino:053]: I2C Bus:
[18:32:30][C][i2c.arduino:054]: SDA Pin: GPIO18
[18:32:30][C][i2c.arduino:055]: SCL Pin: GPIO23
[18:32:30][C][i2c.arduino:056]: Frequency: 50000 Hz
[18:32:30][C][i2c.arduino:059]: Recovery: bus successfully recovered
[18:32:30][I][i2c.arduino:069]: Results from i2c bus scan:
[18:32:30][I][i2c.arduino:075]: Found i2c device at address 0x10
[18:32:30][D][voice_assistant:529]: Event Type: 2
[18:32:30][D][voice_assistant:619]: Assist Pipeline ended
[18:32:30][D][voice_assistant:422]: State changed from IDLE to START_PIPELINE
[18:32:30][D][voice_assistant:428]: Desired state set to START_MICROPHONE
[18:32:30][D][voice_assistant:522]: Signaling stop...
[18:32:30][D][voice_assistant:124]: microphone not running
[18:32:30][D][voice_assistant:206]: Requesting start...
[18:32:30][D][voice_assistant:422]: State changed from START_PIPELINE to STARTING_PIPELINE
[18:32:30][D][voice_assistant:529]: Event Type: 1
[18:32:30][D][voice_assistant:532]: Assist Pipeline running
[18:32:30][D][voice_assistant:124]: microphone not running
[18:32:30][C][switch.gpio:068]: GPIO Switch 'AMP Switch'
[18:32:30][C][switch.gpio:091]: Restore Mode: always ON
[18:32:30][C][switch.gpio:031]: Pin: GPIO21
[18:32:30][D][voice_assistant:529]: Event Type: 9
[18:32:30][D][voice_assistant:124]: microphone not running
[18:32:30][C][template.switch:068]: Template Switch 'Use wake word'
[18:32:30][C][template.switch:091]: Restore Mode: restore defaults to OFF
[18:32:30][C][template.switch:057]: Optimistic: YES
[18:32:30][D][voice_assistant:529]: Event Type: 0
[18:32:30][E][voice_assistant:656]: Error: no_wake_word - No wake word detected
[18:32:30][D][voice_assistant:522]: Signaling stop...
[18:32:30][D][voice_assistant:422]: State changed from STARTING_PIPELINE to STOP_MICROPHONE
I never found the root cause of this issue but as @RichardPar said, the simplest workaround is to turn off the use_wake_word switch, wait a second and then turn it back on.
For those who haven’t seen my project, here is the article which includes the esphome yaml config.
I changed the yaml a bit to reset the component… do you know of a method to record the audio on the HomeAssistant side? or a way to verify that audio is getting though?