Microphone (and wake word) not working

[Benjamo44]

Hi all, this is my first ever Github post, for my first ever Github project. Please forgive me if I am asking this in the wrong section.

I have followed the instructions to build the XiaoZhi AI chatbot, and everything is working perfectly, except the INMP441 microphone.

I have tried the project using three different ESP-32 dev kits - 2 x ESP32-S3 and 1 x ESP-32, and while the LED screen works perfectly and the speaker/amplifier also work perfectly, I cannot get the microphone to recognise wake words or anything at all.

I have been able to switch the chatbot to 'Listening' by connecting GIO5 to Ground, but all the chatbot seems to 'hear' is the word 'I' or 'Yes'. I have been able to create an agent at xiaozi.me in the Console, and the chatbot plays songs and tells me jokes, but I cannot do anything to interact with it.

I am Australian and no doubt have a very strong accent, but is anybody able to help me understand why the chatbot does not recognise me?

Finally, I have tested the 3v3 and ground going in to the microphone and both seem ok. I have attached a photo of my setup with the ESP-32D.

Thank you so much for everything, I am LOVING this project!

****** This issue has been resolved. *********

[Benjamo44]

It was my own fault for being terrible at soldering. I had clearly messed up soldering the microphone bits and pieces. I have been able to get it to work now, and have learned valuable lessons.

Thank you again everyone who has read this, and had anything at all to do with this XiaoZhi-ESP32 project. It is amazing.

Please feel free to delete all of this at the appropriate time.

Ben